# ZopaAI > ZopaAI combines internal sales data with 230,000+ external sources and surfaces potential and risk at every level of the sales organisation. Every statement is checked against the metadata of 30,000+ scientific publications. **URL:** https://zopaai.com **Contact:** info@zopaai.com --- ## Hero **Rotating headlines:** Potentials. Risks. Forecasts. Recommendations. / Planning. Analyzing. Prioritization. / Scientific. Dependable. Transparent. **Validated claim:** Reliable, Hallucination-Free AI for B2B Sales — Validated Against the Metadata of 30,000+ Scientific Publications ZopaAI places value on your products and customers, by connecting your internal data to intelligence from more than 230,000+ external sources. Your teams see where measurable customer and product potential sits and the reasoning behind every recommendation. **Key figures:** - 230,000+: External data sources - 30,000+: Scientific publications checked via metadata - 0: Recommendations without a source --- ## The name — ZOPA is the zone of possible agreement. Every negotiation has a floor the seller will not go below and a ceiling the buyer will not go above. The range between them is where a deal is still possible, and it is usually guessed at. ZopaAI works it out from evidence: what the market pays, what this account has paid, what the alternatives cost, what the tender window allows. The company is named after it because finding that range is the whole job. --- ## Data Sources **External** (230,000+ sources): - Market Data (Daily) - Trade Data (Daily) - Tenders (Hourly) - News (Live) **Internal data** (Your systems): - CRM-Systems - ERP-Systems - Customer Reports - Procurement **Supported connectors:** - Salesforce - SAP - Oracle - Microsoft Power BI - Microsoft Excel - Icertis --- ## Research Partners and Funders - DFKI, German Research Center for Artificial Intelligence - Saarland University - TUM Venture Labs - Federal Ministry for Economic Affairs and Energy - Co-funded by the European Union - de:hub German Ecosystems --- ## Navigation - [ZopaAI](https://zopaai.com/) - [Industries](https://zopaai.com/#industries) - [Platform](https://zopaai.com/platform/) - [Science](https://zopaai.com/science/) - [Stories](https://zopaai.com/stories/) - [About](https://zopaai.com/about/) - [Join us](https://zopaai.com/join-us/) --- ## How it works — Three steps towards your scientifically validated insights ZopaAI adapts to your current tech stack. Minimal effort required from your side. **01. Connect:** ERP, CRM, quotes and sell-out connected as they are. No migration. **02. Enrich:** 230,000+ external sources added and refreshed, not bought once. **03. Validate:** Every output cross-checked against the metadata of 30,000+ scientific publications. Enabling evidence-based, hallucination-free outputs. **04. Decide:** Potential and risk at every level, with the next best action and its source. --- ## Aggregation — Four levels, one figure Every level is the same measurement, aggregated — not a separate number **01. Customer:** Potentials, risks, forecasts and recommendations per customer. **02. Product:** Distribution, substitution and lifecycle per product. **03. Competitor:** Who is gaining, who is losing, and in which category. **04. Company:** Portfolio, regions, opportunities and risks. One number the board can act on. --- ## Science — Hallucination-free by procedure, not by promise. Every statement is traceable to its inputs. If it is not supported by evidence, we withhold it. **Capabilities:** - Customer scoring - Product scoring - Category scoring - Prioritisation - Potential analysis - White-space analysis - Segmentation - Territory planning - Contact and visit steering - Account planning - New customer search - Lookalike search - Lead qualification - Buying centre mapping - Competitor analysis - Price positioning - Win-loss analysis - Churn risk detection - Substitution risk - Cross-sell and upsell - Next best action - Recommended actions - Opportunity ranking - Tender monitoring - Portfolio steering - Launch prioritisation - Assortment gaps - Distribution gaps - Promotion evaluation - Share development - Target setting - Forecast support **Expert quotes:** > "Reliable AI is a procedure, not a promise. Through formal verification and statistical model checking, we give ZopaAI's systems a scientific foundation." > — Timo P. Gros, Deputy Manager and AI Researcher, DFKI > "A recommendation that cannot name its source is worthless in an enterprise context, regardless of how plausible it sounds." > — Prof. Dr. Benedikt Schnellbächer, Chief Innovation Officer, ZopaAI --- ## Capabilities — Eight modules. Usable on their own, together a steering system. - **Contact and visit steering:** Contact planning by potential instead of a class derived from last year's revenue: who to see, how often, and in which channel. - **Competitor analysis:** Market share, substitution movements and price development by product, region and channel, from sales, pricing and market data. - **New customer search:** Look-alike models on structural, location and environment data find accounts that resemble your most profitable existing customers. - **Prioritisation:** A potential score and a risk score at every level: customer, product, competitor and company. - **Customer scoring:** One profile per account: revenue and margin development, portfolio breadth, ordering rhythm and payment behaviour, against a matched peer group. - **Recommended actions:** The score becomes a task: what to do, with whom, when and with which argument, including rationale and data source. - **Cross-sell and upsell:** Portfolio gaps versus comparable accounts, scored by expected revenue effect, close probability and the right moment. - **Territory planning:** A potential map at postcode, district and country level from customer density, sector structure, investment climate and competitive presence. --- ## Industries — Reliable custom solution for your industry What works in Steel, does not work in FMCG. That is why ZopaAI provides custom solutions for each industry and adapts them to your individual needs and challenges. --- ## Industry: Pharma & Healthcare **URL:** https://zopaai.com/industries/pharma/ CRM data, field activity and market intelligence, enriched with insights from over 230,000 external sources. Turning complex signals into clear priorities for every visit cycle. **At a glance:** - Unit of decision: Pharmacy, clinic, buying group - Typical cycle: Weekly visit, quarterly negotiation - Data refresh: Weekly - Steered by: Sales director, key accounts **What makes pharma and healthcare different** Six channels, one patient. Sell-in is not sell-out. The people who decide are not the people you sell to. A generic sales model misses where the margin actually shifts. - **Six channels, one customer:** Pharmacies, clinics, wholesalers, buying groups, hospitals and public tenders each move on their own clock. Treating them as one segment hides where the margin is really moving. - **Sell-in is not sell-out:** Rebate agreements, listing changes and substitution at the counter mean that what leaves your warehouse is a poor proxy for what reaches the patient. - **Decided outside the sales cycle:** Fee schedules, tender windows and Rx switches are set by regulators and payers, and rarely reach the CRM the day the decision is taken. **Four levels, in pharma and healthcare** - **Customer:** Pharmacy, clinic, wholesaler and buying group scored on order rhythm, share of chest and switching risk. - **Product:** Distribution, substitution and lifecycle by SKU across OTC, Rx and consumer health. - **Competitor:** Who is winning shelf, who is losing it, and in which channel and region. - **Company:** Portfolio across channels, exposure to tender cycles and rebate risk on one page for the board. **Eight building blocks, usable on their own, together a steering system** All modules are usable on their own or interconnected in a steering portfolio. ### Module: Contact and visit steering Contact planned by future potential, rather than historical visit patterns. ZopaAI identifies where a visit is most valuable and steers contact frequency and channel accordingly, from in-person visits to inside sales and digital outreach. - Contact frequency per account instead of rigid segmentation classes - Steering the right contact measure: field visit, inside sales or digital outreach - Impact measurement: sales change after contact, per territory **Sample output:** ”Territory South: 38 of 210 accounts generate 64% of the expected potential. Recommended: 18 with in-person visits, 46 through inside sales, the remaining accounts through digital outreach.” ### Module: Competitive analysis Competitor activity is tracked across individual accounts, products and regions. ZopaAI identifies where competitive activity threatens existing business and where changes in the market create new opportunities for your field teams. - Quantified competitive pressure per account and product - Forecasting the accounts and regions most under pressure - Product, positioning and market comparison across relevant competitors **Sample output:** ”Region West: competitive pressure increased across 27 accounts over the last eight weeks. Competitor X gained share in 16 priority accounts, while market signals indicate the strongest pressure around Product Group 03.” ### Module: New customer discovery Your CRM only contains the accounts you already know. ZopaAI identifies new practices, pharmacies, hospitals and other relevant healthcare accounts, matches them against the profile of your strongest customers, and estimates their potential before the first contact. - New practices, pharmacies and healthcare accounts matched on market fit - Lookalike search built on your best-performing existing accounts - Expected potential per prospect, before first contact **Sample output:** ”Region North: 42 practices and 18 pharmacies outside your CRM match the profile of your highest-potential accounts. 21 show above-average potential based on specialty, location and relevant market signals.” ### Module: Prioritisation ZopaAI structures accounts and opportunities by expected future potential for the coming cycle and shows explicitly where field teams should focus. Resources follow opportunity, rather than static target lists or historical priorities. - One ranked account list per team, territory and visit cycle - Ranking by expected future potential, not historical performance - Priorities continuously updated as new signals emerge **Sample output:** ”Q4, field force South: 34 of 180 accounts represent 61% of the expected potential for the coming cycle. New market and engagement signals move 19 accounts into the priority group and 27 below the cut-off.” ### Module: Account scoring Every practice, pharmacy, hospital and healthcare account carries an explainable score down to the underlying factors. ZopaAI makes account potential comparable while allowing your teams to understand exactly what drives each score. - Potential, engagement, activity and risk per account - Every score decomposes into the variables that produced it - Continuous updates instead of static annual segmentation **Sample output:** ”Account 0902, specialist practice: potential 82, engagement 67, competitive risk high. Contact frequency fell 21% over two quarters while external market signals indicate increasing potential in the surrounding area.” ### Module: Next-best-action recommendations Not a dashboard to interpret, but concrete next steps per account. ZopaAI names the action, the expected effect and the relevant market and engagement signals it is derived from. - Clear next best action for every relevant account - Prioritised by expected effect, not by ease of execution - Withheld where the evidence does not support a reliable recommendation **Sample output:** ”Account 2210: recommend an in-person visit within the next two weeks. Engagement has increased across the last three interactions, while comparable accounts responded strongest to personal contact at this stage of the cycle.” ### Module: Cross-sell & upsell ZopaAI identifies cross-sell and upsell potential within existing accounts, showing which products or product groups offer additional potential and where relevant opportunities remain unrealized. - Product gaps per account, ranked by expected potential - Relevant cross-sell opportunities identified across the existing portfolio - Additional product opportunities identified per account and segment **Sample output:** ”Pharmacy segment: 74 accounts show strong potential for Product Group B while currently engaging only with the core portfolio. Comparable accounts adopting both product groups show 12% higher portfolio penetration.” ### Module: Region overview & territory planning Territories cut by potential, not by static logic that misses important factors. ZopaAI puts the potential and risks of practices, hospitals and other healthcare accounts on one map, enabling active territory planning instead of relying on historic structures. - Potential, coverage and risks per territory, on one map - Always know which regions and accounts require attention - Clear next best actions per territory based on current potential **Sample output:** ”Territory South carries 1.8× the potential of Territory East at comparable field-force capacity. Reallocating 24 priority accounts across the territory boundary balances expected potential while reducing uncovered high-potential accounts.” **What changed at an OTC-manufacturerin 6 months.** **Results:** - +11 %: revenue uplift at targeted accounts - −26 %: time spent on field sales preparation - +9 pp: listing depth at targeted accounts **Project scope:** - Client: An OTC manufacturer with a self-medication portfolio across several indication areas. Presented anonymised. - Territory: Germany, two sales regions, ~ 2000 pharmacies in the target segment. - Implemented: Pharmacy prioritisation, visit frequency planning and per-store cross-sell recommendations. - Period: Six months, rolled out in waves, each measured against a control region on the previous logic. > "In OTC we saw the category shift weeks before it reached our own sell-out figures." > — Head of Field Sales, OTC manufacturer **Questions we hear first** **Q: Which data do we need to get started?** ZopaAI can run completely on external data. Market, prescription, sales or other internal data can be added where available. ZopaAI connects these sources without requiring you to rebuild your existing data infrastructure. Naturally, the combination of internal and external data provides the best results. **Q: Does this work for both Rx and OTC portfolios?** Yes. ZopaAI can be applied to both Rx and OTC environments. The underlying data and decision logic are adapted to the respective market, customer structure and commercial objectives. **Q: Do we need complete market and prescription data?** No. ZopaAI can start with the data already available to your team and enrich it with relevant external sources. Additional market, prescription or third-party data can be integrated as it becomes available. --- ## Industry: FMCG **URL:** https://zopaai.com/industries/fmcg/ Category data, promotion history and sell-out, combined with intelligence from over 230,000 external sources. Enabling you and your team to win every sales channel. **At a glance:** - Unit of decision: Retailer and banner - Typical cycle: Weekly promo, quarterly range - Data refresh: Weekly - Steered by: Key account manager **What makes FMCG different** The retailer holds the shelf, you hold the range. Distribution and promo depth decide the outcome long before the sell-out data arrives, and the margin usually sits one level up from the SKU on the shelf. - **The retailer holds the shelf:** Distribution, facings and promotion depth decide the outcome long before sell-out data lands. A rep defending an SKU without the category picture is already a step behind. - **Promotions have a memory:** What you gave away last quarter shapes what the buyer will ask for this one, and few CRMs track that back at the SKU or banner level. - **Category beats brand:** The margin usually sits one level up from the SKU your rep is defending. A category-level view is where the negotiation is actually won. **Four levels, in FMCG** - **Customer:** Retailer, banner and store scored on range compliance, promotion payback and order regularity. - **Product:** Distribution, cannibalisation and category role by SKU across ambient, chilled and impulse. - **Competitor:** Share of shelf and share of promotion by category and by retailer, private label included as its own competitor. - **Company:** Portfolio position across banners, promo commitment coming up and margin at risk on one page. **Eight building blocks, usable on their own, together a steering system** All modules are usable on their own or interconnected in a steering portfolio. ### Module: Contact and visit steering Contact planned by short-term and long-term potentials, rather than historical data. ZopaAI enables you to focus on the future-strong customers, from retail stores to restaurants. - Contact frequency per account instead of rigid A/B/C classes - Steering the needed contact measure - Impact measurement: sales change after contact, per territory **Sample output:** “Region south needs attention, due to rising competitive pressure by CompetitorX, emerging revenue decline of -12% and listing depth development of -3%.” ### Module: Competitive analysis Competitor movements and analysis on a granular basis, customer per customer. ZopaAI lets you see which customers pose a threat at which products, customers or regions. - Quantified competitive pressure per account - Forecasting the regions most under pressure - Assortment and price comparison per retailer and category **Sample output:** ”Category 07, drugstore channel: your share fell 1.8 pp over nine weeks, 1.1 pp of it to private label. Leaflet and online pricing data place two competitors 40 cent below your price point since week 31.” ### Module: New customer discovery Retail head offices are finite. The store and wholesale long tail is not. ZopaAI identifies unlisted stores, regional chains or cash-and-carry partners, matches them against the structural profile of your strongest accounts, and estimates expected volume before the first contact. - Unlisted stores and regional chains, matched on category fit - Lookalike search built on your best-performing accounts - Expected volume per prospect, before first contact **Sample output:** ”Region West: 214 outlets carry no listing of your core range. 38 sit in catchments matching your strongest 50 accounts on purchasing power and store format; expected volume 1.4 m € p.a.” ### Module: Prioritisation ZopaAI orders accounts, categories and SKUs by expected margin or volume contribution for the coming cycle, and shows explicitly where the focus should be. - One ranked list per team, region and cycle - Ranking by expected sales volume or margin contribution, not revenue history - Plan-versus-model deviations shown, not smoothed over **Sample output:** ”Q4, field force South: of 120 planned accounts, 31 carry 68 % of the expected margin effect. Regional category growth data move 22 further accounts below the cut-off.” ### Module: Account scoring Every retail head office, store, drugstore and wholesale partner carries a score, that is explainable up to the lowest level. Guaranteeing insightful scoring, whilst allowing you to understand the very fundamentals of this score. - Potential, margin quality, order rhythm, proximity and risk per account - Every score decomposes into the variables that produced it - Continuous update instead of an annual segmentation **Sample output:** ”Account 0902, buying group: potential 62, margin quality 41, churn risk high. Order frequency fell 18 % over two quarters while catchment data show the category growing 3 %.” ### Module: Next-best-action recommendations Not a dashboard to interpret, but one concrete next step per account. ZopaAI names the action, the expected effect and the comparison cases it is derived from. - Recommended action per account, with its reason - Prioritised by expected effect, not by ease of execution - Withheld where the evidence does not carry the needed granularity **Sample output:** ”Store 2210: propose secondary placement for SKU 4120 ahead of week 40, where leaflet data place the category's promotion peak. Outlets comparable by footfall and catchment gained 1.6× uplift with the same mechanic; margin effect 11 k € per quarter.” ### Module: Cross-sell & upsell ZopaAI provides you with detailed cross-sell & upsell potentials, showing the sum that is left out and the potential of pushing this action. - Range gaps per account, ranked by expected margin - Substitution risk flagged before the second SKU is pushed - Pack and size upgrades identified per channel **Sample output:** ”Drugstore channel: 74 accounts stock your core range but not the premium line — all in catchments where purchasing-power data show the premium segment growing. Comparable accounts add 8.2 % category margin with no measurable cannibalisation.” ### Module: Region overview & territory planning Territories cut by potential, not by a static logic, that misses out on important factors. ZopaAI puts potentials and risks on one map, allowing for an active planning, instead of historic practices. - Potential and risks per territory, on one map - Always know where your attention is needed - Clear next best actions per region **Sample output:** ”Territory South carries 2.1× the potential of Territory East at identical headcount, measured on store density and catchment demographics. Re-cutting three postcode areas balances workload to within 6 %.” **What changed at an international drinks manufacturerin 3 months.** **Results:** - +6–9 %: revenue uplift in target stores - ~35 %: less time to prepare for a field force visit - ~4100: stores implemented **Project scope:** - Client: Drinks manufacturer, multiple categories, EU-based. - Territory: Germany, 6 retail chains, broad listings. - Implemented: Competitor analysis, Account prioritisation, field force steering, recommended actions & more. - Period: 3 months (ongoing project). > "Our field sales team finally has future-ready intelligence on its customers. We were able to successfully rework our high priority accounts and gain market share, with ZopaAI." > — Head of Field Force, Drinks manufacturer **Questions we hear first** **Q: Do we have to license panel data for this?** No. Panel data improves the picture where you already have it, but it is not a precondition. The external layer is built from our 230,000+ databases. If you do license panel data, we work inside the terms of your contract. **Q: Does this work for one category, or do we need the whole portfolio?** One category is enough to start. The comparison logic gets stronger with more of the portfolio, and most teams start with the category where current challenges are greatest. **Q: Does this work if we do not get sell-out from every chain?** Yes. The model generally works fully without internal / bought data from your side. Extra data can be implemented after an initial trial or on the go. --- ## Industry: Automotive **URL:** https://zopaai.com/industries/automotive/ ZopaAI links your project pipeline, quote history and call-off data to 230,000+ external sources and scores every open RFQ on lifetime contribution, win probability and engineering effort. **At a glance:** - Unit of decision: OEM programme - Typical cycle: Programme calendar, 3 to 7 years - Data refresh: Weekly - Steered by: Programme sales lead **What makes automotive different** Programmes decide, quotes follow. The sales cycle runs on the customer’s programme calendar, not on yours, and the first drift shows up in call-off patterns long before it shows up in revenue. - **Programmes decide, quotes follow:** The sales cycle is set by the customer’s programme calendar. Misreading it loses the RFQ before the quote lands. - **Call-off is a signal, not a schedule:** A three-percent drift in call-offs against forecast is usually the first thing that changes, and it goes unread for a quarter. - **Tier-1 activity is public:** Programme announcements, plant retooling and localisation moves show up in press and filings months before they reach the CRM. **Four levels, in automotive** - **Customer:** OEM, tier-1 and tier-2 scored on programme exposure, call-off consistency and awarded-vs-quoted rate. - **Product:** Programme allocation, substitution risk and lifetime margin by part family. - **Competitor:** Who won which programme, at which plant and on which platform. - **Company:** Programme portfolio, forward call-off exposure and retooling risk on one page. **Eight building blocks, usable on their own, together a steering system** All blocks run on the same data foundation and the same score logic. You can start with one of them without having to rebuild the others later. ### Module: Account scoring and next best action One score per OEM group and per customer plant, recomputed weekly, from lifetime contribution of the running programmes, price-down exposure, call-off deviation against nominated volume, quality cost and your share of that customer's relevant platforms. Next to the score sits the movement since last week, the component that caused it, and one action: prepare a claim, raise at the quarterly review, secure the successor, hold, reduce exposure, or do nothing. Each action is shown with its expected effect and the condition that would change it. Before go-live the model is run over your last two nomination cycles, so you can compare its recommendations against what your team actually decided. - One score per OEM group and per plant, recomputed weekly - Movement shown with the component that caused it - Run over your last two nomination cycles before it is used forwards **Sample output:** “Group 3, plant Zwickau: score down 7 points. Driver: call-offs on programme P41 have run 12 % below the nominated volume for two quarters, so tooling amortisation is short. Recommended: raise at the November review, before the price-down round assumes the nominated volume.” ### Module: Programme and RFQ prioritisation A ranked list of open RFQs by expected lifetime contribution per engineering hour. The win probability behind it comes from your own quote history: who the incumbent is, what you were awarded at that group before, whether your footprint fits the OEM's logistics radius, and your quality record on comparable parts. Each RFQ also shows whether you have contact on the engineering side, because a specification is set before purchasing sends the request. - Open RFQs ranked by expected contribution per engineering hour - Win probability derived from your own awards and losses, not from a generic model - Buying center coverage shown per RFQ, engineering as well as purchasing **Sample output:** “14 open RFQs require 3,100 engineering hours, 1,900 are available. The top 6 by expected contribution per hour use 1,750 hours and carry 78 % of the expected value. Two of the remaining eight have an incumbent on the current platform and no contact on your side outside purchasing.” ### Module: Quote and lifetime pricing A price per part across the full programme life, not only at SOP. The calculation holds the annual price-down path, the amortisation of tooling and development over the nominated volume, and the material clause with its index and lag, and shows contribution year by year. It also gives the volume at which amortisation breaks even, so the downside of the nomination is known before it is signed. - Contribution shown year by year to EOP, not only at SOP - Break-even volume for tooling and development stated per quote - Material clause checked for index, lag, cap and floor before it is signed **Sample output:** “RFQ 2291, SOP 2028, seven years: at a 3 % annual price-down and the nominated 240,000 pieces a year, contribution turns negative in year five. Amortisation breaks even at 71 % of the nominated volume. Comparable awards at this group were signed with a 2 % path.” ### Module: Call-off and volume tracking A rolling forecast of call-offs per programme, set against the volume the nomination was calculated on. It is built from call-off history, the OEM's own production plan and the end market, updated weekly, and shown with the change since the previous week. The cumulative shortfall is carried against the amortisation basis, so an unrecovered tooling position is documented as it builds rather than at EOP. - Rolling call-off forecast per programme against the nominated volume - Cumulative shortfall tracked against the amortisation basis - Weekly change shown, so a drift is visible while it can still be raised **Sample output:** “Programme P41: nominated 240,000 pieces a year, call-offs tracking at 211,000. Cumulative shortfall since SOP 46,000 pieces. Tooling recovery is 19 % behind plan. Evidence prepared for the next commercial review.” ### Module: Carry-over, derivatives and aftermarket The follow-up business attached to programmes you already run: facelifts, derivatives, successor platforms and service parts. Each running programme is shown with the expected sourcing date of its successor, which part families you hold on the current version and which you do not, and whether service part volume for the outgoing version is quoted. This is the part of the pipeline that usually exists as knowledge in one person's head. - Successor sourcing dates tracked per running programme - Part families you hold and do not hold shown side by side - Service part volume on outgoing versions surfaced before EOP **Sample output:** “Platform B, facelift sourcing expected Q3 2027. You supply 2 of 4 relevant part families on the current version. Service part volume on the outgoing version is not quoted today, around 30,000 pieces a year at aftermarket pricing.” ### Module: Programme and customer risk Programme delays, volume cuts, cancellations, localisation moves and financial deterioration at your customers, tracked from announcements, filings, plant investments and payment behaviour. Exposure is aggregated by OEM group, by platform and by powertrain, so a concentration that sits across several programme numbers becomes one figure. - Exposure aggregated by OEM group, platform and powertrain - Programme delays, volume cuts and localisation moves tracked from public signal - Payment behaviour and financial deterioration monitored at tier-1 customers **Sample output:** “38 % of contribution after 2029 sits on three ICE platforms at one OEM group. Two have no announced successor. One tier-1 customer's payment delay is up 9 days over two quarters against open exposure of 2.1 million euros.” ### Module: Competitive and platform analysis Which competitor sits on which platform and which part family, from award announcements, plant investments, localisation moves and your own quote outcomes. Losses are split into three causes that call for different responses: price, technology or qualification, and footprint. - Competitor presence by platform, part family and OEM group - Losses split into price, technology and footprint - Plant investments and localisation moves tracked as they are announced **Sample output:** “Chassis brackets, Group 2: 5 of 9 losses went to a supplier with a plant inside the OEM's logistics radius. The price gap was under 2 % in four of them, so footprint rather than price decided the award.” ### Module: New customer and programme discovery OEM groups, tier-1s and platforms whose part families match what you already make profitably, with the sourcing window for each. Group structures are resolved, so a platform at a group you already supply through another plant appears as an existing relationship rather than as a cold contact. Where you hold no qualification with that customer, the approval effort is estimated. - Matched on part family and process capability, not on industry code - Sourcing window shown per platform, so timing is part of the ranking - Existing relationships inside the group resolved before a contact is called cold **Sample output:** “11 platforms with a matching part family enter sourcing within 18 months. 4 belong to groups you already supply through another plant. In 3 cases the relevant engineering location is within 200 km of your development site.” **What changed at a tier-1 supplierin 14 months.** **Results:** - +16 %: hit rate in nominations, relative, versus control regions - −28 %: engineering hours spent on RFQs that were not awarded - +21 %: contribution from carry-over, derivatives and service parts **Project scope:** - Client: Tier 1 supplier, drivetrain and chassis components, anonymised - Territory: Europe, 6 OEM groups, 74 customer plants, 31 sales staff - Implemented: Programme prioritisation, account scoring, carry-over and aftermarket, competitive analysis - Period: 14 months over two nomination cycles, measured against control regions that kept the previous steering logic at comparable programme mix > "Our pipeline was three years of PowerPoint and no numbers. We now know which programmes we should be quoting for and which ones only cost engineering hours." > — Head of Sales, Tier-1 supplier **Questions we hear first** **Q: Does it work for programmes we don't yet supply, and how is aftermarket handled?** Yes. Sourcing announcements, programme milestones and localisation moves are tracked as external signal from the day they become public, so a platform enters the pipeline before the RFQ arrives. Carry-over business is treated the same way: for every programme you run, the expected sourcing date of the facelift, the derivative and the successor is tracked, together with the service part volume on the outgoing version. **Q: What do we have to provide, and when do we see the first result?** One extract: the project pipeline with awarded and lost nominations, quote history with prices and outcomes, call-off data against nominated volumes, and the running contracts with their price-down and material clauses. Standard SAP or your existing pipeline export works. No warehouse project, no ERP replacement, no change to your CRM. Weeks 1 and 2 are master data cleaning and resolving customer plants to OEM groups. By week 6 you see the model run over your last two nomination cycles, so you can compare its ranking against the RFQs your team actually bid for and what those bids returned. From there it is two or three sales staff, roughly two hours a week. **Q: We already track the pipeline. What is different, and what if a recommendation is wrong?** A pipeline lists what your team entered. It does not tell you which of fourteen open RFQs your engineering capacity should go to, and it does not know what happened to the platform since the entry was made. ZopaAI adds the external half and turns it into a ranking with a stated reason. Every recommendation carries its evidence and the condition under which it would change, so someone who knows the customer can overrule it. Overruled recommendations are logged and feed back into calibration. --- ## Industry: Mechanical engineering **URL:** https://zopaai.com/industries/mechanical-engineering/ Your ERP knows what you sold. It doesn't know what is happening at the plant today. ZopaAI connects your service history and quote log with external signals such as investment, ownership changes, hiring, expansion, and industry activity to score every account across the machine lifecycle. **At a glance:** - Unit of decision: Site and installed base - Typical cycle: 8 to 15 year install, monthly service - Data refresh: Weekly - Steered by: Regional sales lead **What makes mechanical engineering different** The installed base is the pipeline, and the ten years after a new-machine sale decide the account. Losing an install is losing a decade, so the signals that a switch is coming matter more than the pipeline number. - **The installed base is the pipeline:** Where the next new-machine order comes from is decided by service history and retrofit windows, not by outbound activity. - **Service and spares carry the margin:** A new-machine sale looks bigger, but the ten years of service and spares after it decide whether the account is a good one. - **The buyer moves once a decade:** Losing an install is losing a decade. Early signals that a switch is coming matter more than the current pipeline number. **Four levels, in mechanical engineering** - **Customer:** Site, plant and group scored on installed-base age, service depth and switching risk. - **Product:** Machine family, retrofit kit and spare part scored on lifetime margin and substitution. - **Competitor:** Who is installing where, and which retrofits are being bid against yours. - **Company:** Installed-base map, service coverage and lifecycle exposure across regions. **Eight building blocks, usable on their own, together a steering system** All blocks run on the same data foundation and the same score logic. You can start with one of them without having to rebuild the others later. ### Module: Account scoring & prioritisation A/B/C classes come out of last year's revenue and tell you who bought, not who is about to — and the scored list that replaces them is still too long for the week. ZopaAI scores every account on expected twelve-month potential from fleet, service behaviour and external investment signals, then cuts it into a ranked work list per rep, sized to the capacity actually available. - One score per account, with the reasons shown - Ranked work list per rep, sized to real capacity - Separate scores for end customer, dealer, service partner **Sample output:** “Account 3042, potential 81/100. Drivers: 12 machines past 8 years, capex announcement in Q3, two maintenance roles posted. Rank 4 of 9 on the Central list this month.” ### Module: Retrofit & replacement timing The replacement conversation usually starts when the customer asks — by which point a competitor has already quoted. ZopaAI estimates the replacement window per machine from age, service intensity and the customer's own investment behaviour, and puts it on a calendar your field team can plan against. - Replacement window per machine, by quarter - Rising service cost flagged as a replacement signal - Customer capex cycle matched against your quote timing **Sample output:** “18 machines enter the replacement window within three quarters. 5 of them sit at sites with an announced capacity expansion.” ### Module: Service contract retention A service contract lapses quietly, and by the renewal date a third-party provider has often been on site for a year. ZopaAI flags contracts at risk months ahead, using service frequency, parts drift and signs that another provider is already working the machine, and ranks the effort by contract value at risk. - Risk flag per contract, months before renewal - Third-party presence detected from parts drift - Renewal effort prioritised by contract value at risk **Sample output:** “9 contracts flagged at risk, combined annual value €780k. In 4 cases parts consumption dropped while machine hours stayed flat.” ### Module: New customer discovery Trade fair lists and association directories give you names, not reasons, so the field cannot tell a prospect from a placeholder. ZopaAI identifies plants outside your base with a fleet and process profile matching your best customers, and ranks them by how close they are to a decision rather than by company size. - Look-alike plants, matched on fleet and process profile - Ranked by timing signal, not by company size - Competitor installations flagged where visible **Sample output:** “63 plants in DACH match your top-quartile customer profile. 11 show an active investment signal within the last two quarters.” ### Module: Cross-sell & upsell Spare parts and service sell themselves at the big accounts, while the long tail carries the margin and never gets worked. ZopaAI compares each site's parts and service consumption against comparable fleets, flags what is missing, and attaches an expected order value so the list can be worked top-down. - Consumption gap per site, measured against peer fleets - Expected order value, so the list can be prioritised - Long-tail accounts surfaced without manual review **Sample output:** “47 sites consume below peer level for wear parts. The top 12 carry an expected gap of €410k over twelve months.” ### Module: Region overview & territory planning Territories were cut years ago along postcodes, and nobody knows where the potential actually sits or whether headcount matches it. ZopaAI maps potential, coverage and travel effort per territory, so headcount and budget can be set against potential instead of history — and re-cuts can be modelled before they are implemented. - Potential and coverage per territory, side by side - Travel effort weighed against expected effect - Cut scenarios modelled before they are implemented **Sample output:** “Territory South carries 1.8× the potential of Territory North at the same headcount. Rebalancing three postcodes shifts 340 accounts without adding travel time.” ### Module: Next-best-action recommendations Knowing an account matters is not the same as knowing what to do with it on Monday. ZopaAI turns each scored account into one concrete next step — the occasion, the channel and the timing — so the field opens the conversation with a reason rather than a check-in call. - One concrete next step per account, not a status label - Occasion named: what changed and why now - Channel and timing assigned: visit, inside sales or digital **Sample output:** “Account 3042: quote for a wear-part package before end of Q2. Occasion: two machines past the service interval while output stayed flat. Channel: on-site, combined with the Hanover route.” ### Module: Competitive analysis You find out you are up against a competitor when the quote is already on the table, and you rarely learn where they have quietly taken over part of your base. ZopaAI detects competitor presence per site from parts drift, service gaps and public installation signals, and shows where your share of a customer's fleet is shrinking. - Competitor presence per site, not per market report - Share of fleet trend: where your base is eroding - Displacement candidates ranked by recoverable volume **Sample output:** “At 23 sites competitor equipment has entered the fleet since 2024. In 7 cases your parts volume fell by more than 40 % while machine hours stayed flat.” **What changed at a mid-sized machine builderin 12 months.** **Results:** - +27 %: revenue from the installed base - −22 %: quotes that never became orders - +14 pp: renewal rate for service contracts **Project scope:** - Client: Machine builder, special-purpose machines and series plant (anonymised) - Territory: DACH and Eastern Europe · around 2,400 machines in the field · 18 sales staff - Implemented: Installed-base analysis · prioritisation · cross-sell · territory planning - Period: 12 months, rolled out in two waves, measured against control territories > "We were still pushing new machines into accounts where the service business was already gone. Now we defend the service accounts first. The machine sale follows from that, not the other way round." > — Head of Sales, Machine builder **Questions we hear first** **Q: Do you connect to our service ticketing system?** Yes. Service history is one of the strongest predictors of retrofit and replacement, so it is one of the first sources we connect. Where machine runtime data is available, it goes in as well — but nothing depends on it. **Q: How long until we see the first result?** A first scored territory list after two to four weeks. The software was used for twelve months because it was rolled out across two waves and measured against control territories. **Q: Our master data is a mess. Does this even work?** Yes, and it is usually the starting point rather than a blocker. The external layer does not depend on clean master data, and the first thing you get back is a matching report showing exactly which accounts and sites could be resolved and which could not. Most clients use that report to clean their base, not the other way round. **Q: What does our IT have to do?** One read-only export or a read connection to your ERP, plus service and parts data if you have it. No write access to your systems, no changes to your processes, no data warehouse project. Hosting is in the EU and your data is never used to train other models. --- ## Industry: Steel **URL:** https://zopaai.com/industries/steel/ ZopaAI reads your ERP, quote log and contract book, links them to 230,000+ external sources, and re-scores every account weekly on base price, pass-through, freight and risk. **At a glance:** - Unit of decision: Contract account - Typical cycle: Weekly price, annual framework - Data refresh: Weekly - Steered by: Sales director **What makes steel different** Input costs and freight move on a weekly clock, but the CRM does not. Volume is not margin. A framework signed in one price environment becomes the loss-making one when the environment shifts. - **Price moves faster than the CRM:** Iron ore, energy and freight change weekly. Annual account plans go stale in a month if nothing reads that signal for them. - **Volume is not margin:** Two accounts of the same tonnage can differ by six euros a tonne. One of them is the one worth defending, and it is not always the larger. - **Contracts hide the risk:** A framework signed in one price environment can become the loss-making one when the environment shifts. Forward exposure has to be read on the contract, not on the invoice. **Four levels, in steel** - **Customer:** Account scored on contract exposure, price pass-through and switching risk against import competition. - **Product:** Grade and dimension family scored on contribution margin and forward demand. - **Competitor:** Import and domestic movements by grade and by port. - **Company:** Contract book, forward exposure and margin risk by grade and region. **Eight building blocks, usable on their own, together a steering system** All blocks run on the same data foundation and the same score logic. You can start with one of them without having to rebuild the others later. ### Module: Account scoring and next best action One score per account, recomputed weekly, from realised margin, surcharge pass-through, volume stability, payment behaviour and strategic fit. Next to the score sits the movement since last week and the component that caused it, and one action: reopen the contract, hold, shift to inside sales, secure volume, reduce exposure, or do nothing. Each action is shown with its expected effect and the condition that would change it. Before go-live the model is run over your last two contract rounds, so you can compare its recommendations against what your team actually decided. - Pass-through gap shown per contract, separate from base price - Weekly movement shown with the component that caused it - Run over your last two contract rounds before it is used forwards **Sample output:** “Account 4711 moved down 9 points. Driver: pass-through gap on two contracts widened after the last index step. Payment behaviour unchanged. Recommended: prepare for the November renewal, no action before then.” ### Module: Price and quote guidance ZopaAI gives a price corridor, derived from what comparable grades and dimensions actually realised, adjusted for the current surcharge position, freight and incoterm. Alongside it sits the contribution per hour of the constraining unit, so an order that fills a rolling hour cheaply is visible as such. The corridor is shown with the quotes it was derived from, including the ones that were lost and at what price. - Upper and lower figure per quote, derived from comparable realised business - Base price and surcharge position held separately, never as one number - Contribution per hour of the constraining unit shown next to the price **Sample output:** “Quote 8841, 320 t, spring steel: corridor 40 euros per tonne wide, based on 14 comparable quotes over eight months. Current offer sits at the lower edge. Week 38 is already committed at below-median contribution per hour, so lead time rather than price is the constraint here.” ### Module: Contract round preparation Frameworks are valued at signature and then not again. ZopaAI recalculates the remaining term at today's input costs and checks the surcharge formula for lag, cap and floor. Set against the account's alternatives, meaning competitor presence in that grade, import pressure and lead times, that gives an opening position with its evidence and a walk-away volume fixed before the meeting. - Remaining term valued at today's input costs, not at signature - Surcharge formula checked for lag, cap and floor before it is signed again - Walk-away volume quantified per account before the negotiation starts **Sample output:** “Account 4711, renewal in week 44: the current formula lags the scrap index by one month, costing around 12 euros per tonne in a rising market. Opening position: shorten the lag to two weeks, hold base price. Walk-away volume 900 t, replaceable from the order book within six weeks.” ### Module: Volume and call-off planning A rolling forecast per account of what will actually be called off, set against the committed volume. It is built from call-off history, the customer's own order situation and the end market, updated weekly, and shown with the change since the previous week. A gap between commitment and forecast is therefore visible from the second quarter, not at year end. - Rolling call-off forecast per account, against the committed volume - Weekly change shown, so a drift is visible while it can still be acted on - Aggregated by plant and grade, so the production side sees the same figure **Sample output:** “Account 4711: 12,000 t committed, 9,400 t forecast at year end. The gap opened over the last two quarters, driven by the customer's own order intake. Recommended: address at the November review, before the renewal discussion assumes last year's volume.” ### Module: Early risk detection Payment behaviour, filings, capacity announcements and a customer's own call-off pattern change weeks before receivables do, so a credit limit is usually cut when the tonnage is already in the order book. ZopaAI monitors accounts and the groups behind them for these signals, aggregates exposure to group level rather than per legal entity, and shows where the book is concentrated: one group, one end market, one destination, one quota regime. - Deterioration signals from payment behaviour, filings and capacity announcements - Exposure aggregated to group level, not per legal entity - Concentration shown by end market, destination and quota regime **Sample output:** “Group 3120, four legal entities: payment delay up 11 days over two quarters, one plant closure announced at its largest customer. Open exposure 4,200 t across the next two quarters. Recommended: cap open volume, shorten terms at the next order.” ### Module: Contact and visit steering A contact frequency and a channel per account, derived from expected effect per contact hour with travel time included, instead of an A/B/C class set once a year from tonnage. Each processor, distributor or industrial customer is assigned to a personal visit, to inside sales or to digital outreach. Coverage is also shown per territory, which surfaces groups covered by two representatives and accounts with no assigned owner. - Contact frequency per account instead of rigid A/B/C classes - Channel assignment: visit, inside sales or digital outreach - Duplicate coverage inside groups and unassigned accounts made visible **Sample output:** “Region West: 22 of 145 processors carry 70 % of the expected effect. Recommended: 8 monthly, 14 quarterly, the rest through inside sales. Three accounts visited monthly show no measurable response after four cycles.” ### Module: Competitive analysis Which competitor supplies which grade into which region, from trade flows, capacity announcements and your own quote outcomes. - Competitor presence by grade, region and destination market - Import pressure tracked through customs and trade flow data **Sample output:** “Bearing steel, northern Italy: 4 of 11 losses were on lead time, not price. Average gap 9 days. Price concessions in this segment did not change the win rate.” ### Module: New customer discovery Candidates matched on grade, dimension and application profile against what you already produce profitably: processors, forges, foundries and distributors. Group structures are resolved, so subsidiaries of accounts you already supply appear as well. Existing qualifications are part of the match, with an estimate of the approval effort where one is missing. - Matched on grade and dimension profile, not on industry code - Group structures resolved, so subsidiaries of existing customers surface - Existing qualifications carried into the match, approval effort estimated where missing **Sample output:** “17 candidates with a matching grade profile. 6 are subsidiaries of groups you already supply. 9 existing accounts run an application already approved for your special grade at a sister plant.” **What changed at a steel producerin 12 months.** **Results:** - +2,4 pp: contribution margin, top-quintile accounts, versus control regions - −34 %: write-offs and receivables overdue beyond 90 days - +19 %: share of special grades in total volume **Project scope:** - Client: Steel producer, long and flat products (anonymised) - Territory: Europe · around 900 active accounts · 26 sales staff - Implemented: Margin and pass-through analysis, price guidance, early risk detection, contract preparation - Period: 12 months, two contract rounds, measured against control regions > "We used to negotiate on tonnage. Now we negotiate on where the margin actually is, and we know which accounts to keep and which ones the market will decide for us." > — Sales Director, Steel producer **Questions we hear first** **Q: How do you handle surcharges, and how current is the data?** Scrap, alloy and energy surcharges are modelled per product line exactly as they are contractually agreed, including index reference, lag, cap and floor. Input costs and freight are refreshed at least weekly, and contract exposure is recomputed on every quote and every order-book update. Base price, surcharge and realised contribution are shown separately, so a pass-through gap becomes visible per contract instead of disappearing into an invoice total. **Q: What do we have to provide, and when do we see the first result?** One extract: invoice lines with grade, dimension, plant and incoterm, the surcharge components as billed, the quote log with won and lost outcomes, and the contract book. Standard SAP COPA works. No warehouse project, no ERP replacement, no change to your CRM. Weeks 1 and 2 are master data cleaning, holding consolidation and grade mapping across plants. By week 6 you see the model run backwards over your last two contract rounds, before it recommends anything forwards. From there it is two or three sales staff, roughly two hours a week. **Q: We already have BI dashboards. What is different, and what if a recommendation is wrong?** A dashboard reports what happened inside your own system. It does not know what the market did to your contract since it was signed, and it does not tell you which of 900 accounts to open next week. ZopaAI adds the external half and turns it into a ranked recommendation with a stated reason. Every recommendation carries its evidence and the condition under which it would change, so someone who knows the account can overrule it. Overruled recommendations are logged and feed back into calibration. --- ## Industry: Chemicals **URL:** https://zopaai.com/industries/chemicals/ Application engineering, formulation history and regulatory signal live in different places. ZopaAI joins them to 230,000+ external sources and grades every account on the margin you actually keep. **At a glance:** - Unit of decision: Formulator and application - Typical cycle: Quarterly qualification, annual tender - Data refresh: Weekly - Steered by: Application-led sales **What makes chemicals different** Specialty and commodity are two different businesses in one company. Application engineering is the sales cycle, and regulation decides where growth is possible months before the pipeline shows it. - **Specialty and commodity are two businesses:** A model that works for one hides where the money is in the other. They need separate scores that aggregate cleanly at company level. - **Application engineering is the sales cycle:** Whoever is in the customer’s lab is the one likely to win the next tender. Application time is the resource to steer, not the meeting count. - **Regulation moves the market:** REACH, biocides and chemical restrictions decide where growth is possible, often months before the pipeline shows it. That signal has to reach sales. **Four levels, in chemicals** - **Customer:** Account scored on formulation depth, share of application and switching risk. - **Product:** Specialty grade and commodity line scored on contribution margin and regulatory exposure. - **Competitor:** Who is winning which application, at which formulator and with which grade. - **Company:** Portfolio position across specialty and commodity, and regulatory risk by market. **Eight building blocks, usable on their own, together a steering system** All blocks run on the same data foundation and the same score logic. You can start with one of them without having to rebuild the others later. ### Module: Regulatory & substitution signal A REACH restriction, an SVHC listing or a customer's own substitution programme is the moment a formulation comes under review — and you usually hear about it after the trial has started. ZopaAI tracks regulatory decisions and substitution announcements across your customers' end markets and maps them onto the formulations they buy from you, so the conversation starts while the alternative is still being chosen. - Regulatory changes mapped to affected customer formulations - Substitution pressure flagged in the customer's own end market - Lead time shown: quarters until the change bites **Sample output:** “14 accounts are exposed to the upcoming restriction on two of your additive lines. In 5 cases the end customer has already announced a reformulation programme for 2027.” ### Module: Account scoring & prioritisation A/B/C classes come out of last year's tonnage and tell you who bought the most, not where future margin potential lies. ZopaAI scores every account on expected twelve-month specialty potential from purchase history, application activity and external production signals, then cuts it into a ranked work list per rep, sized to the capacity actually available. - Scored on specialty potential, not on tonnage - Ranked work list per rep, sized to real capacity - Separate scores for formulator, processor and distribution partner **Sample output:** “Account 1184, potential 78/100. Drivers: three commodity grades with specialty equivalents, new production line announced, two formulation roles posted.” ### Module: Application engineering allocation Application engineering is your scarcest resource and it goes to whoever asks loudest. ZopaAI estimates the conversion probability and expected margin of each application request before the hours are committed, so technical time lands on the projects most likely to convert rather than simply the loudest requests. - Conversion probability per application request, before the work starts - Expected margin attached, so requests can be ranked - Hours spent measured against orders won, per region **Sample output:** “Of 38 open application requests, 11 carry a conversion probability above 60 %. Together they account for 71 % of the expected margin in the pipeline.” ### Module: Specialty switch potential Customers sit on standard grades for years because nobody made the case at the right moment. ZopaAI identifies which commodity positions have a viable specialty equivalent, and ranks them by the signals that make a switch realistic now — a new line, a regulatory trigger, a change in the customer's own product spec. - Commodity positions with a viable specialty equivalent - Ranked by switch trigger, not by account size - Expected margin uplift per switch **Sample output:** “62 accounts hold commodity positions with a specialty equivalent. 17 show an active switch trigger, worth an estimated €890k in additional margin.” ### Module: Cross-sell & upsell Your top accounts buy across the portfolio. The rest buy the two products they started with, and nobody has time to find out what is missing. ZopaAI compares each account's product mix against comparable formulators and flags the gaps, with an expected order value attached. - Portfolio gap per account, measured against peer formulators - Expected order value, so the list can be worked top-down - Mid-sized accounts surfaced without manual review **Sample output:** “53 accounts buy below peer level across two product families. The top 15 carry an expected gap of €620k over twelve months.” ### Module: New customer discovery Trade fair lists and industry directories give you company names, not application fit. ZopaAI identifies formulators and processors outside your base whose application and production profile matches your best customers, and ranks them by how close they are to a formulation decision. - Look-alike formulators, matched on application and process profile - Ranked by timing signal, not by company size - Competitor supply positions flagged where visible **Sample output:** “71 formulators in Europe match your top-quartile application profile. 9 have announced a new line or a reformulation within the last two quarters.” ### Module: Distribution partner performance More than half your revenue runs through distributors, and their reports tell you what sold, never what was available and missed. ZopaAI compares each partner's realised revenue against the potential in their territory, so the conversation moves from volume targets to coverage gaps. - Realised revenue against territory potential, per partner - Coverage gaps named: which accounts were never approached - Specialty share per partner, benchmarked across the network **Sample output:** “Partner Benelux realised 61 % of the potential in their territory. 34 scored accounts show no recorded contact in the last twelve months.” ### Module: Region overview & territory planning Regions were cut years ago along country borders, and nobody knows whether application engineering capacity sits where the potential is. ZopaAI maps potential, coverage and technical capacity per region, so headcount and AE hours can be set against potential instead of history. - Potential and coverage per region, side by side - Application engineering capacity weighed against expected effect - Cut scenarios modelled before they are implemented **Sample output:** “Region South carries 1.6× the specialty potential of Region North at the same headcount. Two application engineers cover 41 % of the scored base.” **What changed at a specialty chemicals producerin 12 months.** **Results:** - +21 %: revenue from specialties - +11 pp: application projects that converted - +13 pp: switch rate from standard to specialty **Project scope:** - Client: Specialty chemicals producer, additives and formulations · around 800 employees - Market: Europe · around 1,300 active accounts · roughly 55 % of revenue through distribution - Team: 24 field sales and application engineers - Blocks in use: Application engineering allocation · Account scoring · Specialty switch potential · Distribution partner performance - Period: 12 months, three waves · measured against control regions - Data connected: ERP customer master · application project log · quote log · regulatory sources > "We now spend application engineering time where it will pay back, not where the loudest sales rep asks for it." > — Head of Sales, Specialty chemicals **Questions we hear first** **Q: How do you handle formulation confidentiality?** Internal formulation data stays internal. ZopaAI reads it in place and returns scores, never content. Nothing leaves your environment, nothing is used to train models, and the external layer works entirely on public and licensed sources. **Q: Our application project data sits in spreadsheets, not in a system. Does that work?** Yes, and it is the normal starting point. We take the project log in whatever form it exists, including exports and spreadsheets. The first thing you get back is a matching report showing which projects could be resolved against accounts and which could not — most clients use that to clean up the log, not the other way round. **Q: Most of our revenue runs through distributors. Does that work?** That is the part with the weakest internal data, so it is where external signals contribute most. Formulator, processor and distribution partner are scored separately, so you can see where a partner is underperforming against the potential in their territory rather than against last year's numbers. **Q: What regulatory sources are covered?** REACH restrictions and the SVHC candidate list, national and EU authorisation decisions, and publicly announced substitution and reformulation programmes in your customers' end markets. Sources are listed per project, so you can see exactly what a score is built on. --- ## Industry: Electrical engineering **URL:** https://zopaai.com/industries/electrical-engineering/ Products are specified by planners and panel builders long before the order reaches wholesale. ZopaAI joins your ERP, quote log and wholesale sell-out to 230,000+ external sources and scores every account, project and channel partner. **At a glance:** - Unit of decision: Account, project and channel partner - Typical cycle: Specification months ahead of the order - Data refresh: Weekly - Steered by: Key account and channel management **What makes electrical engineering different** The decision is made twice: once when a planner or panel builder writes your product into a specification, and again when wholesale fills the order. Most sales data only sees the second. - **Specified early, ordered late:** Planners, integrators and panel builders decide the brand months before the purchase order. By the time the order shows in the ERP, the account was won or lost elsewhere. - **Wholesale hides the end customer:** Much of the volume runs through distribution, so sell-in looks stable while sell-through to the actual installer or OEM is shifting. - **Raw material moves the margin:** Copper and energy prices change the margin on framework contracts between two price rounds. Which accounts carry that risk is rarely visible in time. **Four levels, in electrical engineering** - **Customer:** OEM, panel builder, installer and wholesale branch scored on potential, specification share and switching risk. - **Product:** Component, drive and system scored on margin, substitution and raw-material exposure. - **Competitor:** Who is being specified where, and in which projects and product families. - **Company:** Project pipeline, channel mix and margin exposure across regions. **Eight building blocks, usable on their own, together a steering system** All blocks run on the same data foundation and the same score logic. You can start with one of them without having to rebuild the others later. ### Module: Project & specification pipeline By the time a wholesaler asks for a price, the specification has been written and your chance to influence it has passed. ZopaAI tracks building and infrastructure projects from planning permission through tender, matches them against your product categories, and shows which ones are still open to specification. - Projects matched to your categories, from permit to tender - Stage shown: pre-spec, in spec, tendered, awarded - Weeks of influence left, per project **Sample output:** “47 projects in your categories are currently pre-tender. 9 are still open to specification, with an estimated 11 weeks of influence remaining.” ### Module: Specifier influence mapping The order comes from a wholesaler, the decision came from a planning office you have never visited. ZopaAI reconstructs which planners, consultants and panel builders stand behind your won and lost projects, and ranks them by how much volume their specifications actually move. - Planners and panel builders ranked by specified volume - Win rate per specifier, not per customer - Competitor specifications flagged where visible **Sample output:** “12 planning offices specify 58 % of the volume in your regions. In 4 of them your win rate is below 20 % despite regular contact.” ### Module: Account scoring & prioritisation A/B/C classes come out of last year's invoiced revenue, which in this industry tells you what the wholesaler moved, not who decided. ZopaAI scores every account on expected twelve-month potential from sell-out data, project involvement and external building activity, then cuts it into a ranked work list per rep. - Scored on decision influence, not on invoiced volume - Ranked work list per rep, sized to real capacity - Separate scores for planner, panel builder, installer and branch **Sample output:** “Panel builder 0412, potential 84/100. Drivers: three pre-tender projects in spec, sell-out up 22 % at two branches, competitor line discontinued.” ### Module: Next-best-action recommendations Knowing an account matters is not the same as knowing what to do with it on Monday. ZopaAI turns each scored account into one concrete next step — the occasion, the channel and the timing — so the field opens with a project reference rather than a check-in call. - One concrete next step per account, with the occasion named - Timed against the project stage, not the visit cycle - Channel assigned: visit, inside sales or digital outreach **Sample output:** “Planning office Hanover: bring the updated switchgear spec sheet before end of month. Occasion: two projects entered detailed planning in April.” ### Module: Channel & sell-out transparency Wholesale sell-in tells you what you shipped into the channel. It does not tell you which installer took it out, or which branch is sitting on stock while another runs dry. ZopaAI combines sell-out data where it exists with external structural data where it does not, and rebuilds the path from branch to end customer. - Sell-out per branch reconciled against sell-in - End-customer view where sell-out data allows it - Gaps labelled as estimates, never presented as reported data **Sample output:** “Of 180 branches, 64 report sell-out. For the remaining 116, structural estimates cover 71 % of the volume — flagged as estimated throughout.” ### Module: Cross-sell & upsell Installers and panel builders buy the two product families they started with, and nobody has time to find out what is missing. ZopaAI compares each account's product mix against comparable accounts in the same segment and flags the gaps, with an expected order value attached. - Portfolio gap per account, against peer accounts - Expected order value, so the list can be worked top-down - Mid-sized accounts surfaced without manual review **Sample output:** “68 panel builders buy below peer level across two families. The top 15 carry an expected gap of €540k over twelve months.” ### Module: Framework contract & pricing position Framework contracts get renewed on last year's terms because nobody has time to check whether the volume assumption still holds. ZopaAI compares committed against realised volume per contract, flags where terms no longer match the business, and shows which renewals sit in a strong or weak position before the negotiation starts. - Committed against realised volume, per contract - Renewal position rated before the negotiation - Contracts flagged where rebate terms outgrew the volume **Sample output:** “23 framework contracts come up for renewal within two quarters. In 7, realised volume sits more than 30 % below the committed level.” ### Module: Region overview & territory planning Territories were cut along postcodes years ago, and project activity has moved since. ZopaAI maps project volume, coverage and channel presence per region, so headcount and travel budget can be set against where construction is actually happening. - Project volume and coverage per region, side by side - Channel density weighed against field capacity - Cut scenarios modelled before they are implemented **Sample output:** “Region West carries 1.7× the project volume of Region East at the same headcount. Sell-out coverage in the West sits at 44 %.” **What changed at an electrical components manufacturerin 12 months.** **Results:** - +18 %: specification share in target projects - +14 pp: projects reached before tender - +9 pp: win rate on specified projects **Project scope:** - Client: Manufacturer of electrical components, drives and switchgear · around 700 employees · anonymised at their request - Market: DACH and Benelux · around 1,800 active accounts - Channel: Roughly 70 % of revenue through wholesale · 180 branches in scope - Team: 22 field sales and application engineers - Blocks in use: Project & specification pipeline · Specifier influence mapping · Account scoring & prioritisation · Framework contract position - First output: Scored project pipeline after 5 weeks - Measurement period: 12 months, two contract rounds · measured against control regions - Verification: Figures from the client's own project controlling · reference call available under NDA **Questions we hear first** **Q: Can you see through wholesale to the end customer?** Where sell-out data from distribution partners exists, it is connected as it is. Where it does not, external structural and project data fill the gap and are labelled as such — you always see which part of a view is reported and which is estimated. In practice most clients start with partial sell-out coverage and expand it as partners come on board. **Q: How are projects handled next to run-rate business?** Separately, because they behave differently. Run-rate accounts are scored on consumption and portfolio gaps; projects are scored on stage, specification influence and remaining time to tender. Both feed the same work list, so the field sees one prioritised view rather than two systems. **Q: Most of our revenue runs through wholesale. Does that work?** That is the part with the weakest internal data, so it is where external signals contribute most. Planner, panel builder, installer and wholesale branch are scored separately, so you can see where a branch is underperforming against the project volume in its catchment rather than against last year's numbers. **Q: Our project data sits in the CRM, half-maintained. Does that work?** Yes, and it is the normal starting point. We take the project log as it is and match it against external tender and permit data. The first thing you get back is a matching report showing which projects could be resolved and which could not — most clients use that to clean the log, not the other way round. --- ## Press and Expert Quotes > "ZopaAI stands for reliable, scientifically verified AI for use in companies, in order to avoid the problem of AI hallucinations." > — Frankfurter Allgemeine Zeitung, Newspaper, Frankfurt > "ZopaAI shows that Germany can not only regulate but also shape. Scientifically grounded AI from the heart of Europe is exactly what we need to stay competitive in the global market." > — Matthias Mieves, Committee on Digital Transformation, German Bundestag > "European AI research doesn't just produce scientific excellence — it also produces entrepreneurial courage. ZopaAI is a convincing example of that." > — Prof. Dr. Ludger Santen, President, Universität des Saarlandes > "Reliable AI isn't a promise, it's a proof. Through formal verification and statistical model checking, we give ZopaAI's systems a scientific foundation that can be scrutinized, reproduced, and trusted." > — Timo P. Gros, Deputy Manager & AI Researcher, DFKI > "Most AI systems are only as good as the data they're trained on. We go further by integrating domain knowledge directly into our models — we deliver reliable, interpretable insights even where data is limited." > — Prof. Dr. Verena Wolf, Professor of Algorithmic Intelligence · Scientific Partner to ZopaAI, DFKI --- ## Voices — Voices About Us **Frankfurter Allgemeine Zeitung** (Press) > "ZopaAI stands for reliable, scientifically verified AI for use in companies, free from AI hallucinations." > — Frankfurter Allgemeine Zeitung, Newspaper, Frankfurt **Saarstahl** (Customer) > "ZopaAI is an essential step in making our sales more forward-looking and effective." > — Dr. Nadine Artelt, Sales Director, Saarstahl **TUM Venture Labs** (Research) > "European startups have an edge over Silicon Valley that they play too rarely: scientific credibility. ZopaAI makes exactly that the foundation of its business model." > — Dr. Philipp Gerbert, CEO, TUM Venture Labs **German Bundestag** (Politics) > "ZopaAI shows that Germany can not only regulate but also shape. Scientifically grounded AI from the heart of Europe is exactly what we need." > — Matthias Mieves, MdB, Committee on Digital Transformation, German Bundestag **DFKI** (Research) > "Reliable AI is not a promise, it is a proof. Through formal verification and statistical model checking, we give ZopaAI's systems a scientific foundation." > — Timo P. Gros, Deputy Manager and AI Researcher, DFKI **Universität des Saarlandes** (Research) > "European AI research doesn't just produce scientific excellence, it also produces entrepreneurial courage. ZopaAI is a convincing example of that." > — Prof. Dr. Ludger Santen, President, Universität des Saarlandes **DFKI** (Research) > "Most AI systems are only as good as the data they're trained on. We go further by integrating domain knowledge directly into our models: we deliver reliable, interpretable insights even where data is limited." > — Prof. Dr. Verena Wolf, Professor of Algorithmic Intelligence · Scientific Partner to ZopaAI, DFKI --- ## Company — Who is behind ZopaAI From research with DFKI to customers in 10 industries, in our first year. **Facts:** - Founded: June 2025 - Revenue in year 1: 7-digit - Industries served: 10 **Mission:** AI enablement in sales: translate the data you already have into decisions your team can work with on Monday morning. **Vision:** Know today what will be an opportunity and what will be a risk tomorrow. Evidenced, not estimated. --- ## FAQ **Q: Is hallucination-free AI possible?** Yes. ZopaAI is hallucination-free by procedure: every output is cross-checked against the metadata of 30,000+ scientific publications, and anything the evidence does not support is withheld. What reaches you comes with its source. **Q: Do we have to migrate our data?** No. ZopaAI adapts to your tech stack, ensuring minimal effort from your side, with maximum customization. **Q: Is there one model for every industry?** Yes and no. There is one model for every industry, however every customer solution is a one-off, tailor-made system. **Q: What happens in a demo?** We will present ZopaAI to you and define the most-relevant use-cases for you to start using ZopaAI. --- ## Book a demo Not another sales pitch, but an honest discussion focusing on your needs. 30 minutes, no sensitive data, no risk, no cost. **What to expect:** - Industry lead, not a generalist - Valuable discussion, honest assessment. - Focused on discovering potentials. **Book:** https://zopaai.com#demo --- ## Platform **URL:** https://zopaai.com/platform Every sales decision, backed by data and science. One platform to see your market, prioritize your customers and plan your next move. Built on internal and external data, validated by science, ready for every level from field rep to CSO. **How it works (four steps):** - **Connect:** ERP, CRM, quotes and sell out data are read live, where they already are. No migration, no duplicate records, no system change. Up and running with minimal effort from your team. - **Discover:** Your data shows your side of the relationship. We show you the whole customer: their market, their growth, their influences and their risks, drawn from 230,000+ external sources. - **Validate:** Every output is cross-checked against the metadata of 30,000+ scientific publications before it reaches you. The result is hallucination-free, transparent and verified business intelligence. - **Decide:** Potential, risk and forecasts at every level, from company to individual account. With the evidence behind every recommendation, your team plans with confidence and negotiates from a position of strength. **What your procurement and IT want to know** - **Servers located in Germany:** All data is held and processed exclusively in Germany. Operated as a container-based environment in our secure German data centers. - **Separate data silos per client:** Every client gets its own sealed data silo. There is no shared tenant layer and no mixing of client data in analysis or model training. - **ISO-certified information security:** Operations follow an information security management system in line with ISO 27001, with defined processes for access control, logging, incident management and regular audits. - **On-site hosting available:** On request, ZopaAI runs inside your own infrastructure. For companies whose policies do not permit external processing of sales data, deployment in your own data centre is possible. **No standard model across all industries, one algorithm per industry and per company.** - **Industry logic:** Every industry has its own rhythm. What signals growth in one market means little in the next, and the moments that decide a deal differ just as much. ZopaAI knows which signals matter where and when they matter, from pharma and FMCG to steel and automotive. Each industry model comes with its own data sources, metrics and cycles, so your team works with the logic of your market from day one. - **Company logic:** Your company, your rules. Behind every sales organisation sits its own set of decisions: which products matter most, which customers come first, which partners carry the business and how the team is set up to reach them. ZopaAI is configured around exactly that structure, down to the metrics you steer by. The result is a system that fits the way you sell, not one you have to adapt to. --- ## Science **URL:** https://zopaai.com/science Built with DFKI, next door to CISPA. ZopaAI's headquarters is on the Saarland University campus, next to the German Research Center for Artificial Intelligence and CISPA. **The process enabling verified outputs** Every output is validated through these four steps before it leaves the system. - **Structured analysis:** Random forest models analyse sales and market data. They are applied because their decisions can be traced and explained. - **Industry-specific context:** Custom models based on each client's business and industry. - **Matching against evidence:** Every statement is checked against empirical data and the metadata of more than 30,000 peer-reviewed publications. - **Output with source, or not at all:** If a statement cannot be supported by evidence, it is not returned. Every recommendation comes with its source. **What “not returned” means** Every candidate statement has to name at least one admissible source before it can reach you. Four came up here. Three could. One was modelled internally but could not, so it was withheld rather than softened. **Research and funding** Funded by European institutions, supported by the German Research Center for Artificial Intelligence and leading universities. - **DFKI (Research partner):** Our core research partner. Together, we develop the hybrid AI solution and continuously validate our models. - **Saarland University (Company seat):** Company headquarters on the university campus. The roots of our scientific basis. - **CISPA Helmholtz Center (Professional exchange):** Professional exchange on trustworthy European handling of corporate data. - **TUM Venture Labs (Venture partner):** Supports ZopaAI's growth as part of the Technical University of Munich's venture network. --- ## About **URL:** https://zopaai.com/about Born in one of Europe's leading AI clusters. Our roots are in one of Europe's leading AI research clusters. Developed together with DFKI, our platform brings scientific rigor to enterprise sales. Today, a team of 20+ is shaping the future of reliable B2B-Sales AI. **From cutting edge research to measurable sales impact.** Our journey from scientific research to a proven solution, used by international market leaders in 10 industries. **Origin timeline:** - **Scientific base:** Research on digital transformation and AI at Saarland University lays the scientific foundation for what will later become ZopaAI. - **2025:** ZopaAI is founded in June 2025. Together with the German Research Center for Artificial Intelligence (DFKI), the team develops its AI-driven sales software, builds a strong network of partners from science and industry. - **Progressing:** ZopaAI strengthens its network of scientific collaborations and receives funding from German and European governmental bodies and public institutions. - **Today:** 7-digit revenue in year 1 and customers in 10 industries. ZopaAI keeps expanding its team across Germany, and partners and customers publicly endorse ZopaAI. **Value of our ecosystem.** What research partners, industry and shared expertise bring to the way we build. **Backed by European research and public institutions.** The methodology is developed with the German Research Center for Artificial Intelligence and validated at Saarland University. Funding comes from European institutions. --- ## Stories **URL:** https://zopaai.com/stories Impact you can see ZopaAI in the press: independent coverage, awards and milestones from industry and innovation. - **[PRESS] IHK Magazine Features ZopaAI’s Success Story** (2026-10): IHK Saarland featured ZopaAI in its latest magazine, taking a closer look at where we come from, what we’ve achieved since our commercial launch in June 2025 and how we’re growing across Europe. The article shows how we replace gut feeling with a mix of internal and external data, from forecasting to recommended actions and new customer acquisition, for industries like pharma, steel, mechanical engineering, FMCG and automotive. - **[COMPANY] State Secretary of Economic Affairs Interviews ZopaAI** (2026-10): State Secretary Elena Yorgova-Ramanauskas of Saarland’s economic ministry sat down with our founders Benedikt Schnellbächer and Alexander Schöneseiffen to talk about AI, scaling a business and the future of entrepreneurship in Saarland. In this 20-minute discussion, they share how ZopaAI is actively shaping the future of B2B sales. - **[COMPANY] Winner at Harvard University’s AI Venture Competition** (2026-10): ZopaAI is an award winner at the AI Venture Competition 2026 at Harvard University. It was the only startup from outside the Harvard community and still was able to succeed. Reliable, scientifically validated AI for B2B Sales, now award-winning at Harvard. - **[COMPANY] Rohde & Schwarz and ZopaAI at Bits & Pretzels 2026: Reliable AI at the Core of Strategic Sales** (2026-09): At Rohde & Schwarz, reliable AI is part of how the company plans and steers its sales. On stage in Munich, the two companies showed what that looks like in practice. Rohde & Schwarz has been building technology for more than 90 years, driven by innovation and curiosity. That spirit extends to its sales organization, made up of highly specialized engineers around the world. With ZopaAI, these teams base strategic sales decisions on verified intelligence. - **[COMPANY] ZopaAI Meets Saarland’s Minister for Economic Affairs** (2026-09): ZopaAI discussed three questions with Minister Jürgen Barke at the Ministry for Economic Affairs, Innovation, Digital and Energy: how regional companies strengthen their sales with ZopaAI, how ZopaAI drives its European expansion, and how Europe positions itself in the global AI race. - **[AWARD] ZopaAI Wins the L-Booster** (2026-09): ZopaAI is winner of the southwestX L-Booster, selected by the jury for showing industrial companies where sales investments actually pay off. - **[PRESS] Vertriebszeitung: Europas KI-Chance im B2B – Warum Verlässlichkeit wichtiger wird als Modellgröße** (2026-08): Die Vertriebszeitung zeigt, warum Europas KI-Rückstand im B2B-Vertrieb kein Nachteil sein muss. Prof. Benedikt Schnellbächer erklärt, dass große US- und China-Modelle zwar kreativ sind, aber im B2B zu wenig Faktentreue und Branchentiefe bieten – und warum wissenschaftlich geprüfte, spezialisierte KI wie ZopaAI genau hier den Unterschied macht. - **[PRESS] FAZ Features ZopaAI in Debate on AI Accountability** (2026-07): Frankfurter Allgemeine Zeitung spotlights ZopaAI in a feature on AI companies and corporate accountability, with CIO Benedikt Schnellbächer explaining how scientific verification eliminates AI hallucinations in B2B decision-making. - **[PRESS] Saarbrücker Zeitung: How ZopaAI Makes Million-Euro Deals More Likely** (2026-07): Saarbrücker Zeitung profiles CEO Alexander Schöneseiffen and ZopaAI’s growth from two founders to 20+ people in one year, drawing on 230,000+ public industry databases to power strategic sales decisions for customers like Saarstahl AG. - **[COMPANY] One Year of ZopaAI** (2026-06): In just twelve months, ZopaAI grew from two founders to more than 20 people, opened a second office in Cologne, and crossed €1 million in ARR, driven by the trust of its customers and the support of partners including DFKI, TUM Venture Labs, SouthwestX, and the European Union. - **[COMPANY] ZopaAI at the Sales Managers Congress** (2026-06): ZopaAI presented at the Sales Managers Congress of the Bundesverband der Vertriebsmanager e.V., with one clear takeaway: reliable, scientifically validated AI for B2B sales is more in demand than ever. - **[COMPANY] White Paper: The End of Gut-Feeling Selling** (2026-06): ZopaAI publishes its latest white paper on how AI and external data make value-based sales operationally real in B2B specialty industries, from instinct to intelligence, across the full account portfolio. - **[PRESS] Rheingau Magazine Spotlights ZopaAI** (2026-05): Rheingau Magazine features ZopaAI founder Alexander Schöneseiffen and his ambition to build one of Europe’s leading independent sales-AI platforms. - **[AWARD] ZopaAI Is Awarded the InnoBonus** (2026-04): The start-up ZopaAI has been awarded the Saarland InnoBonus for its outstanding technology transfer to the industry. - **[COMPANY] ZopaAI Expands to Cologne** (2026-04): ZopaAI opens a new office in Cologne to accelerate collaboration with industry partners and deepen its presence in one of Germany’s most dynamic innovation hubs. - **[CUSTOMER] Saarstahl: Advancing Strategy with AI** (2026-03): Saarstahl, one of Europe’s leading steel manufacturers, has deployed ZopaAI’s hybrid AI solution to strengthen market intelligence, customer insights, and strategic sales planning. - **[PRESS] ZopaAI Brings AI Intelligence to the B2B Market** (2026-02): A detailed look at how ZopaAI uses hybrid AI, scientific verification, and 120,000 data sources to transform B2B sales. - **[PRESS] How AI Enables Better Decisions** (2026-01): An interview exploring ZopaAI’s origins, scientific foundation, and the vision behind building a European AI alternative. --- ## Join Us **URL:** https://zopaai.com/join-us Build the future of business relations with us ZopaAI is growing, and every new colleague shapes our direction. If you want to help shape the next generation of data-driven business relationships, we would love to meet you. **Shape what comes next** People join us because they want to solve real industrial challenges and create an impact. - **Real industrial challenges:** Our customers range from industrial leaders to fast-growing innovators, serving customers in 10 industries. - **Scientifically validated KI:** We combine scientifically validated AI with one of Europe’s most comprehensive industry data ecosystems. - **Insights sales teams trust:** Together they deliver insights sales executives can trust. **Open roles:** - **Business Development Associate** (Commercial) — Saarbrücken, Cologne, Copenhagen, Hamburg, Berlin, Mannheim, Paris, Munich: Drive the growth of our customer base by building and managing strategic relationships with top companies worldwide, shaping how our AI solution creates value and evolves in the market. - **Market & Data Analytics Associate** (Commercial) — Saarbrücken, Cologne, Hamburg, Mannheim: Transform complex industrial data into clear insights that guide product, sales and strategy. You analyse markets, identify patterns and help shape how our AI solution delivers measurable value to customers. - **Marketing & Brand Communications** (Commercial) — Cologne: Boost ZopaAI’s visibility by creating clear, engaging communication. You shape our brand voice, craft impactful content and turn complex AI topics into messages that attract customers. - **Software Developer (Backend)** (Engineering) — Saarbrücken, Mannheim: Design and build the core systems that power our AI platform. You develop reliable, scalable backend services, shape our technical architecture and turn complex data and product requirements into clean, high-performance solutions. - **Software Engineer (Backend, Business Logic)** (Engineering) — Saarbrücken: Build the core of our platform by developing scalable backend systems and business logic that turn data into actionable insights for global industry leaders. - **AI / ML Engineer** (Engineering) — Saarbrücken: Advance our AI core by developing and deploying LLM-based models and agent systems that transform complex data into intelligent, high-impact insights. - **DevOps Engineer** (Engineering) — Saarbrücken: Operate and scale the infrastructure behind our AI platform by automating deployments, optimising reliability and ensuring smooth, secure operations as we grow. --- *Generated from src/data/site.ts, industry/*.ts, pages.ts, cms/ (stories) and careers.ts. Last build: 2026-10-25.*