Specimen, fictional company
Company ABC: commercial diligence
Part 1Specimen, fictional company
Summary
The company and the round
An AI customer service platform with EUR 2.4 million of stated ARR, raising a EUR 5 million Series A at EUR 18 million pre-money.
Company ABC sells an AI customer service platform: agents that handle conversations across chat, messaging, email and voice, and act inside the customer's own helpdesk, CRM, order and billing systems. It sells across retail, travel, utilities and fintech, into one of the most crowded categories in software.
It was founded in 2022 and has 38 customers and EUR 2.4 million of ARR, up from EUR 0.9 million a year earlier. Pricing is outcome-based with guardrails: a platform minimum of at least EUR 1,500 a month, a fee per resolved conversation that falls as a customer's volume rises, and a monthly cap on what a customer can spend.
A venture investor is weighing leading the EUR 5 million Series A at EUR 18 million pre-money, 7.5x stated ARR. This report is what that investor reads before its investment committee decides, and it ends with an example of what the first 100 days after the round could hold.
Our commercial opinion
Real commercial potential, in a narrower market than the deck describes. Would we put our own money in, on the commercial case alone? Yes, with conditions: 6 out of 10.
The category is real and Company ABC has a defensible wedge: buyers who run customer service across several systems at once, where a helpdesk's own agent cannot reach. The deck's headline numbers overstate it. On the company's own data, 44% of conversations are confirmed resolved, not 70%. Recurring run-rate is EUR 2.07 million, not EUR 2.4 million, and growth on that basis is 130%, not 167%. Net revenue retention is 94% outside the 5 largest accounts.
Most of what is wrong is commercial, and can be changed in the first 100 days: how a resolution is defined and priced, how caps behave at peak, and who closes the large deals. One risk is structural: the cost of a resolution rises with volume while the price falls. This is an opinion on commercial potential, not a price. The levers and questions for the negotiation are in part 7.
- Would we invest our own money, on the commercial case alone
- 6 out of 10
- Yes, with conditions
| Commercial dimension | Weight | Score, 1 to 10 | Why |
|---|---|---|---|
| Market and wedge | 25% | 8 | A real wedge in buyers running several systems, with a 58% win rate there. Narrower than the deck implies. |
| Revenue quality and retention | 25% | 6 | 130% growth on recurring run-rate and a floor of platform minimums, but 94% retention outside the 5 largest accounts. |
| Unit economics and pricing | 20% | 4 | A 49% gross margin that falls as volume rises, and caps that serve the summer peak for nothing. |
| The product claim | 15% | 5 | 44% confirmed resolution against the 70% claimed, and 31% in voice. |
| The commercial engine | 15% | 5 | A strong demo, and a motion that still needs the founder to close large deals. |
| Weighted score | 100% | 5.8 | Rounded to 6 |
The scale: at 1 to 3 we would not invest on the commercial case, at 4 to 5 only with the risks priced and conditioned, at 6 to 7 we would invest with conditions, and at 8 to 10 we would invest as it stands. Each dimension is scored on the evidence in parts 2 to 6.
Opportunities
Five opportunities, each backed by the company's own data, and each with a way to capture it.
| Opportunity | Evidence | How to capture it |
|---|---|---|
| A wedge buyers pay for | 58% win rate where the buyer runs 3 or more systems, against 17% with 1 | Point the ideal customer profile and the pipeline at multi-system buyers, the first field in part 8 |
| Growth that survives the rebuild | 130% on recurring run-rate, EUR 0.9 million to EUR 2.07 million | Report on recurring run-rate from the first board meeting, so growth is measured on what recurs |
| A floor under revenue | Platform minimums are 61% of stated ARR and 71% of recurring run-rate, contracted annually | Keep the minimum in every renewal while the caps are redrawn |
| Room in the price where it wins | Multi-system buyers are not choosing on price, and nothing in the price list charges for the systems an agent acts in | Test a price that rises with the number of systems the agent acts in |
| Expansion in the accounts that land | The 5 largest accounts retained 168% | Renew them early on redrawn caps, trading a longer term for certainty of spend |
Risks
Five risks. Two can be addressed in the first 100 days, two only once the first renewals arrive in March 2027, and one is structural.
| Risk | Evidence | Can it be addressed in the first 100 days? |
|---|---|---|
| The product claim is overstated | 44% confirmed resolution against 70% contained, and 31% in voice | Yes, in how it is defined and priced. The rate itself moves more slowly. |
| Founder dependence in large deals | 8 of the last 10 deals above average contract value closed with the founder on the call | Yes: a deal desk from day 20, with half of such deals closed without the founder by day 100 as the test |
| Retention rests on 5 accounts | 94% retention outside the 5 largest, which hold 32% of ARR | No. The first renewals on current pricing fall in March 2027. |
| Pressure on price at renewal | Customers expect the price per resolution to fall at renewal, from interviews only | No, for the same reason. Graded Medium in the evidence register. |
| Margin falls as volume rises | Gross margin from 64% on the smallest customers to 34% on the largest, and EUR 0.09 per resolution over direct cost at the top tier | In part: caps and top-tier prices can change, the cost of a voice resolution cannot |
What the score does not cover
The score is on the commercial case only. Before investing, the investor needs financial and legal due diligence, and a technical review, from specialists in each.
- Financial due diligenceQuality of earnings, how usage fees and platform minimums are recognized as revenue, cash burn and runway, working capital, and tax.
- Legal due diligenceCustomer contract terms, including caps, liability and termination, ownership of the product's IP and the licences it depends on, data protection and processing agreements, exposure to AI regulation, employment and founder agreements, and the cap table.
- Technical reviewSecurity, architecture, and how dependent the product and its costs are on the model providers it runs on.
What the deck said, and what we found
Each headline claim in the investor deck, set against what the company's own data shows.
- What the deck said
Our AI resolves 70% of conversations end to end.
- What we foundOf all conversations, 70% never reach a human, and only 44% are confirmed resolved: not abandoned, and no repeat contact within 7 days.
- What the deck said
Customers only pay for outcomes.
- What we foundPlatform minimums, paid whatever the outcome, are 61% of ARR. Resolution fees are 31%.
- What the deck said
Net revenue retention is 131%.
- What we foundThe 5 largest accounts retained 168%. The other 17 customers in the same cohort retained 94%.
- What the deck said
We plug into every channel and system.
- What we foundThe company lists 52 integrations. Just 3 of them carry 86% of conversation volume.
- What the deck said
Gross margin reaches 75% at scale.
- What we foundBlended gross margin is 49%, and it falls as volume rises: 64% on the smallest customers, 34% on the largest.
- What the deck said
We win against the big platforms on accuracy.
- What we foundWin rate follows the buyer's stack, not accuracy: 17% where the buyer runs 1 system, 58% where it runs 3 or more.
Part 2Specimen, fictional company
Market and position
What the company's own documents show
Before a single interview, the data room says 2 things: the company plans to hire to integrate, not to sell, and its pricing has changed 3 times in a year.
- Hires in the operating plan, next 12 months
- 14
- 6 of them integration and solutions engineers
- Case studies quoting confirmed resolution
- 0 of 9
- 7 quote containment or deflection instead
- Changes to pricing in 12 months
- 3
- From the price books
The company's operating plan and hiring plan, September 2026 to August 2027, from the data room. No role in finance, pricing or infrastructure cost is planned.
| When | What changed | What it signals |
|---|---|---|
| October 2025 | Moved from annual licences to a price per conversation | Revenue tied to volume, whatever the outcome |
| March 2026 | A platform fee from EUR 1,500 a month, plus a fee per resolution and a monthly cap on spend | A floor under revenue, introduced alongside the outcome price |
| July 2026 | The cap moved to the front of the sales deck and every proposal, sold as predictable | Buyers are asking for certainty of spend, and the cap has become a reason to buy |
The company's price books, sales decks, and a sample of proposals and signed contracts from each period.
The real competitive set
We win against the big platforms on accuracy.
The deck compares Company ABC with the category it replaced. Buyers compare it with the one it has to beat.
| Category | How the deck describes it |
|---|---|
| Rule-based chatbots | Scripted flows, no reasoning |
| Help center search | Answers questions, takes no action |
| Phone menus | Routes the call, resolves nothing |
The company's investor deck, market slide.
| Category | Why a buyer chooses it | Where Company ABC stands |
|---|---|---|
| The helpdesk incumbents' own AI agents | Already inside the helpdesk contract: one vendor, one bill, no integration | Loses where the buyer runs a single helpdesk |
| CRM suites' agent platforms | Sits on the customer record the rest of the business already uses | Wins when the CRM is one of several systems the service team works in, loses when it is the only one |
| AI-native agent vendors | Funded to win on price and speed of deployment | Meets them in most evaluations. Wins on actions taken across systems, loses on price in chat-only deals |
| Outsourced contact centers adding AI | Sells an outcome with people behind it | Rarely in the same evaluation. Competes for the same budget when the buyer is deciding whether to run service in-house at all |
| In-house builds on the foundation models | Cheap to start and owned outright | Loses at the pilot, where a build looks cheaper. Wins once the build has to act inside 3 or more systems |
Competitor fields on every opportunity closed in the CRM over the last 4 quarters, checked in customer and lost-prospect interviews.
Company ABC wins where a customer runs service across several systems at once. That wedge is real, and narrower than the deck implies.
Where Company ABC wins
Win rate rises with the number of systems in the buyer's stack, from 17% to 58%. Accuracy is not what separates the wins from the losses.
Closed opportunities past discovery, from the CRM, September 2025 to August 2026. Win rate is closed won over closed won plus closed lost.
Part 3Specimen, fictional company
The product claim
Resolution, measured
Our AI resolves 70% of conversations end to end.
Of all conversations, 70% are contained and 44% are confirmed resolved. The 26 points between them are customers who gave up or came back.
- Contained: never reached a human
- 70%
- What the deck calls resolved
- Confirmed resolved
- 44%
- No handoff, not abandoned, no recontact within 7 days
- Gap from abandonment
- 11 points
- The customer left before an answer
- Gap from recontact
- 15 points
- The same customer came back on the same issue
- Contained, the deck's measure
- Confirmed resolved
Every conversation logged from June to August 2026, matched by customer across channels. Contained: closed without a handoff to a human. Confirmed resolved: contained, not abandoned, and no contact from the same customer on the same issue, on any channel, within 7 days. A conversation both abandoned and followed by a recontact counts as abandoned. Channel rates weight by volume to the totals.
Confirmed resolution is highest in chat, at 49%, the channel the product was built for, and lowest in voice, at 31%. Voice is 10% of volume today. Customers told us in interviews that they plan to move more voice volume to automation, which would move more volume to where the product is weakest. That comes from interviews alone, and the evidence register grades it so.
Part 4Specimen, fictional company
Revenue quality
ARR by source, and the recurring run-rate
Customers only pay for outcomes.
Platform minimums are 61% of stated ARR. Take out one-off builds and the summer peak in annualized usage, and recurring run-rate is EUR 2.07 million against EUR 2.4 million stated.
| Source | EUR thousands | Share of ARR | Treatment |
|---|---|---|---|
| Platform minimums | 1,460 | 61% | Recurring, contracted annually, including the licences migrated in March 2026 |
| Resolution fees | 750 | 31% | Usage, annualized by the company from June to August |
| Connector builds | 190 | 8% | One-off services |
| ARR as stated | 2,400 | 100% |
Billing exports and conversation logs, September 2025 to August 2026. The company multiplies June to August resolution fees, EUR 187,500, by 4. The fee per resolution only began in March 2026, so there is no full year of fees to measure. The rebuild annualizes each customer's June to August fees on its own 12 months of conversation volume instead, which the logs hold for every customer. Customers live less than 12 months take the volume profile of their industry. On that basis June to August is 30.7% of a year's fees, not 25%, and annual resolution fees are EUR 610,000.
- Growth, ARR as stated
- 167%
- EUR 0.9 million to EUR 2.4 million
- Growth, recurring run-rate
- 130%
- EUR 0.9 million, all annual licences, to EUR 2.07 million
- Platform minimums, share of recurring run-rate
- 71%
Net retention and concentration
Net revenue retention is 131%.
The 131% holds on the company's own definition, and 5 accounts carry it. They retained 168%. The other 17 customers in the cohort retained 94%.
- The 5 largest accounts, share of ARR
- 32%
- The largest account, share of ARR
- 10%
- Customers retained from the cohort
- 21 of 22
- Net revenue retention on recurring run-rate
- 120%
- The whole cohort, EUR 900,000 to EUR 1,080,000
| Cohort | ARR August 2025, EUR thousands | ARR August 2026, EUR thousands | Net revenue retention |
|---|---|---|---|
| The 5 largest | 450 | 757 | 168% |
| The other 17 | 450 | 423 | 94% |
| All 22 customers live in August 2025 | 900 | 1,180 | 131% |
Cohort: customers with ARR on 31 August 2025, measured on 31 August 2026, including expansion, contraction and churn, on the company's own ARR definition so the figures compare with the deck. The 5 largest are ranked by August 2026 ARR and are also the 5 largest accounts in the whole base. The one churned customer is among the other 17. ARR in August 2025 was annual licences only.
The cohort added EUR 280,000 net. The 5 largest accounts added EUR 307,000 and the other 17 lost EUR 27,000, so more than all of the expansion came from 5 customers. The 17 new customers of the last 12 months hold the remaining EUR 1.22 million of ARR and have no retention history yet.
Part 5Specimen, fictional company
Pricing and unit economics
Pricing position
Company ABC is priced for a buyer comparing it with the helpdesk agent already in their contract. That suits the single-helpdesk deals it mostly loses and undersells the multi-system deals it wins.
| Category | How it prices |
|---|---|
| The helpdesk incumbents' own AI agents | Per automated resolution, added to the helpdesk contract |
| CRM suites' agent platforms | Per conversation or by prepaid credits, sold with the CRM seats |
| AI-native agent vendors | Per resolution, with volume discounts |
| Outsourced contact centers adding AI | Per contact handled or per agent hour, blended with people |
| In-house builds on the foundation models | Model usage, plus the team that builds and runs it |
| Company ABC | A platform minimum from EUR 1,500 a month, a fee per resolution from EUR 0.95 falling to EUR 0.46 with volume, and a monthly cap |
The competing proposals customers and lost prospects described in interviews, and competitor notes in the CRM. Company ABC from its own price book, September 2026.
| Measure | March to May 2026 | June to August 2026 |
|---|---|---|
| Customers who reached their monthly cap in at least 1 month | 3 | 11 |
| Resolutions served above the cap, unbilled, as a share of all resolutions | 4% | 18% |
Billing exports and conversation logs. A customer at its cap pays nothing more that month, and every further resolution still carries its full cost.
Where there is room in the price
Where the buyer runs 3 or more systems, the alternative to Company ABC is an integration project, not a cheaper agent, and the win rate is 58%. Those buyers are not choosing on price, and nothing in the price list charges for the number of systems an agent acts in, which is the thing they are buying.
At the top tier there is no room. List price leaves EUR 0.09 per resolution over direct cost, and in the summer the cap serves 18% of resolutions for nothing.
Price and cost per resolution
Gross margin reaches 75% at scale.
The price per resolution falls by more than half from the smallest customers to the largest, while the cost of a resolution rises. Volume narrows the margin.
- List price per resolution
- Cost per resolution
| EUR per resolution | Under 2,000 | 2,000 to 6,000 | 6,000 to 15,000 | Over 15,000 |
|---|---|---|---|---|
| List price per resolution | EUR 0.95 | EUR 0.78 | EUR 0.60 | EUR 0.46 |
| Cost per resolution | EUR 0.31 | EUR 0.30 | EUR 0.33 | EUR 0.37 |
List prices from the company's price book, September 2026. Cost per resolution is model inference, telephony and hosting, allocated by conversation, March to August 2026. Cost rises with volume because the largest customers run more of their conversations on voice, where telephony and speech inference cost most.
Gross margin by volume tier
Blended gross margin is 49%. It is 64% on the smallest customers and 34% on the largest, which hold 30% of revenue.
| Monthly resolutions | Customers | Share of revenue | Gross margin |
|---|---|---|---|
| Under 2,000 | 16 | 20% | 64% |
| 2,000 to 6,000 | 11 | 25% | 60% |
| 6,000 to 15,000 | 7 | 25% | 44% |
| Over 15,000 | 4 | 30% | 34% |
| Blended | 38 | 100% | 49% |
Customers grouped by their average monthly resolutions since the current pricing began in March 2026. Revenue includes platform minimums. Gross margin is revenue less model inference, hosting, telephony and messaging fees, and support and implementation staff. Resolutions above a customer's monthly cap are unbilled and carry their full cost. Tier margins weight by revenue to the blended figure.
Part 6Specimen, fictional company
The commercial engine
Founder dependence
Large deals close when the founder is on the call. Of the last 10 deals above average contract value, 8 did, and 1 of the 4 account executives has closed one without the founder.
- Deals above average contract value with the founder on the call, last 8 quarters
- 8 of 10
- Average contract value
- EUR 63,000
- First-year value of new customers won in the last 8 quarters
- Account executives who have closed above average without the founder
- 1 of 4
This is what the next stage of growth is being paid for. The company is hiring 2 more account executives, and the motion they would join has not yet been shown to close its largest deals without the founder.
Live sales calls, by funnel stage
The demo is strong and carries the wedge. Discovery does not qualify on it, and negotiation gives away the cap and the price before the buyer asks.
| Stage | What we look for | What we saw |
|---|---|---|
| Discovery | Qualifies on the systems in the buyer's stack | Asked on some calls, never used to disqualify |
| Discovery | Establishes what a contact costs the buyer today | Rarely asked. The business case is left for the buyer to build after the demo |
| Discovery | Names who signs, and what they are measured on | Named on the calls the founder led, left open on the others |
| Demo | Shows an action taken across systems, not only an answer | The strongest moment in the motion, and the wedge |
| Demo | Answers how a resolution is counted | Not raised unless the buyer asks, and answered differently by different sellers |
| Demo | Handles the helpdesk's own agent as the alternative | Answered with accuracy claims, not with the multi-system case where Company ABC actually wins |
| Negotiation | Holds the price per resolution | Discounted at the first objection on deals below average contract value |
| Negotiation | Uses the cap to close, or gives it away | Given away: offered at contracted volume before the buyer raises spend |
| Negotiation | Who has to be on the call for it to close | The founder, whenever the deal is above average contract value |
Observed live across discovery, demo and negotiation, with the seller's agreement, and read against the CRM rather than on its own.
Part 7Specimen, fictional company
Negotiation and questions
Levers for the negotiation
Six levers, each tied to a finding in this report, with what to ask for. Which to use, and how hard, is the investor's call.
| Lever | The finding behind it | What to ask for |
|---|---|---|
| The ARR base | EUR 330,000 of stated ARR is one-off builds or a summer peak | Any multiple applied to recurring run-rate, EUR 2.07 million, not stated ARR |
| The definition of resolved | 44% confirmed resolution against the 70% claimed | A contractual definition of a resolved conversation in every new customer agreement and renewal |
| The caps | 18% of summer resolutions served above the cap, unbilled, at full cost | Caps set on annual rather than monthly volume, for new contracts now and for every renewal |
| Concentration | The 5 largest accounts hold 32% of ARR and more than all of the cohort's net expansion | A renewal commitment from the 2 largest, which hold EUR 410,000 of ARR, before close, or their expansion treated as upside rather than base |
| Founder dependence | 8 of the last 10 deals above average contract value closed with the founder on the call | A deal desk from day 20, and half of deals above EUR 63,000 closed without the founder by day 100, as a milestone the board tracks |
| Use of the round | 14 planned hires, none in pricing, finance or infrastructure cost | A named owner for pricing and for the cost of a resolution, funded from the round |
The price arithmetic, as a tool
The ask is 7.5x stated ARR and 8.7x recurring run-rate. What follows is arithmetic to take into the negotiation, not a recommended price.
| Private B2B SaaS, EUR 2 to 5 million ARR | Pre-money, multiple of ARR |
|---|---|
| Growth under 50% | 3x to 5x |
| Growth 50% to 100% | 5x to 7x |
| Growth over 100%, with net revenue retention over 110% | 7x to 10x |
The lead investor's own ranges, set by its investment committee for rounds at this scale. They assume a software gross margin of 70% or more.
| Measure | As the deck states it | As rebuilt here |
|---|---|---|
| Growth, last 12 months | 167%, on ARR as stated | 130%, on recurring run-rate |
| Net revenue retention | 131% | 120% on recurring run-rate, and 94% outside the 5 largest accounts |
| Gross margin | 75% at scale | 49% blended, falling as volume rises |
| Range | Over 100% growth: 7x to 10x | Over 100% growth, at its bottom edge |
Parts 4 and 5 of this report.
| Measure | At the ask | 7.5x on recurring run-rate |
|---|---|---|
| ARR base, EUR million | 2.40 | 2.07 |
| Multiple of that base | 7.5x | 7.5x |
| Pre-money, EUR million | 18.0 | 15.5 |
| Investment, EUR million | 5.0 | 5.0 |
| Post-money, EUR million | 23.0 | 20.5 |
| Investor's stake after the round | 21.7% | 24.4% |
Pre-money is the multiple times the ARR base, rounded to EUR 0.1 million. Post-money is pre-money plus the investment. Stake is the investment over post-money, before any option pool.
The rebuilt figures keep the company in the top range, but only just. Growth stays above 100% on recurring run-rate, and retention clears 110% only because of the 5 largest accounts. A 49% gross margin sits well below the 70% the ranges assume.
Every 1x of multiple applied to stated rather than recurring ARR pays EUR 330,000 for revenue that is one-off or a summer peak: EUR 2.5 million at the ask's 7.5x. Where in the range to settle is the investor's decision.
Questions to ask before investing
Eight questions this report could not close, who at the company can answer each, and what evidence would settle it.
| Question | Who answers it | What would settle it |
|---|---|---|
| When do the 5 largest accounts renew, and on what terms? | Founder, finance lead | Renewal dates and the draft renewal terms for each |
| Do the largest customers measure resolution the way this report does? | The customers themselves | Their own resolution reporting, set against what they are invoiced for |
| What does a resolution cost if voice volume doubles? | Platform lead, finance lead | A cost model by channel at twice today's voice volume |
| How exposed is margin to a change in model inference prices? | Finance lead | Inference contracts, committed spend and the price terms with the model providers |
| Which account executives can close a deal above EUR 63,000 without the founder? | Head of sales | Named deals in the current pipeline, each with the account executive who owns it |
| How much of the pipeline is single-helpdesk buyers? | Head of sales | The pipeline tagged by the number of systems in the buyer's stack |
| Why did the one customer lost from the cohort leave? | Customer success, the former customer | An exit interview and the account's last 6 months of conversation data |
| What will the company offer a customer that hits its cap every summer? | Founder | The renewal proposal for one such customer, in writing |
What would move the score
Three things would move it: 2 up, 1 down.
- Up: resolution closing the gap in messaging and voiceConfirmed resolution in messaging, 45%, and voice, 31%, moving toward chat's 49% would lift the blended rate the product can be priced on.
- Up: the 5 largest renewing on current pricesRenewals would turn the concentration from a risk into evidence that the wedge holds at scale.
- Down: the largest account renewing at a lower price per resolutionIt is 10% of ARR, and customer interviews suggest buyers expect the price per resolution to fall at renewal. A lower price there would reach the top tier's margin first.
Part 8Specimen, fictional company
The first 100 days
How we would work after the round
As the investor's commercial adviser, working with the company's key executives in focused sessions. The investor keeps a clear view of what is moving, and the team does the work.
After the round, the investor has an adviser who has read the business down to its source data. We work with the founder, the head of sales and the finance lead in focused working sessions, bring the findings of this report into each one, and report back to the investor on what moved and what did not.
What follows is an example of what the first 100 days could hold for Company ABC, not the plan itself. In an engagement, the plan is built with the team from the findings, and the fields below are where it usually starts.
What the first 100 days could include
Nine fields where the findings turn into work with the team, from the ideal customer profile to AI in the daily sales workflow.
| Field | What we would do with the team | What the investor gets |
|---|---|---|
| Ideal customer profile | Rebuild it around buyers who run several systems, and score every open opportunity against it | A pipeline weighted to where the company actually wins |
| GTM plan | Set the segment, the motion and the sequence for the next 4 quarters against the targets agreed at the round | A plan the board can hold the team to |
| Pitch deck and sales narrative | Challenge every claim in the deck against the evidence, starting with resolution | A story that survives the buyer's own diligence |
| Battlecards and competitive overview | Guide the battlecard for each competitor category, built from what lost prospects actually said | Sellers who win on the multi-system case rather than on accuracy claims |
| Sales documentation | Review the discovery guide, the business case template, proposals and pricing guardrails | Fewer discounts given at the first objection |
| CRM stage gates | Review the exit criteria a deal must meet at each stage, and where deals stall | A forecast the board can read |
| Tools, tracking and accountability | Review the sales stack, the operating dashboard and who owns which number | The metrics agreed at the round, in one view, each with an owner |
| AI in the sales workflow | Set up AI with the team for outreach, follow-ups, CRM updates and support during calls | More selling time per seller, and every call leaving a trace in the CRM |
| Pricing and packaging | Define resolution in the contract and redraw the caps on annual volume | Revenue and the product claim measuring the same thing |
An example, sequenced
One way the fields could run for Company ABC: pricing and the dashboard in place before the first board meeting, and the deal desk running before the largest renewals come up.
| Workstream | Days | First milestone | Owner at the company | By day 100 |
|---|---|---|---|---|
| Ideal customer profile and pipeline scoring | 0 to 30 | Every open opportunity scored on the systems in the buyer's stack, and single-helpdesk prospects moved out of the forecast | Head of sales | Single-helpdesk prospects under 20% of forecast value |
| Sales narrative and pitch deck, challenged against the evidence | 0 to 30 | The deck's resolution claim restated on the confirmed rate, with the multi-system case leading | Founder | One narrative used on every first call |
| CRM stage gates and forecast discipline | 0 to 45 | Stage exit criteria agreed with the sellers before the first forecast call | Head of sales | The day-45 forecast within 15% of revenue closed by day 100 |
| Pricing and packaging: resolution defined in contract, caps redrawn | 10 to 60 | A resolved conversation defined in the contract template, and caps set on annual rather than monthly volume | Founder and finance lead | Every new contract and renewal on the new terms |
| Battlecards by competitor category | 15 to 45 | The helpdesk incumbents' agents first, from lost-prospect interviews | Head of sales | A battlecard in use for each of the 5 categories |
| Deal desk, so large deals close without the founder | 20 to 100 | Every deal above EUR 63,000 reviewed by the deal desk before the proposal, with the founder joining by exception | Head of sales | Half of deals above EUR 63,000 closed without the founder |
| Operating dashboard and accountability | 30 to 60 | Confirmed resolution, recurring run-rate, net retention outside the 5 largest and margin by volume tier, on this report's definitions, each with an owner | Finance lead | In use at the first board meeting after the round |
| AI in the sales workflow: outreach, follow-ups, call support | 30 to 75 | Follow-ups drafted from each call's notes, and call summaries written to the CRM | Head of sales | Every sales call summarized in the CRM within a day |
An example for Company ABC, not a fixed plan. Days from the close of the round. Hover over or focus a workstream to see its first milestone on the chart; every milestone is also in the table.
AppendixSpecimen, fictional company
Evidence and definitions
Confidence levels
Every finding carries one of 3 levels, set by the classes of source behind it rather than by how strongly it is worded.
| Level | What it requires |
|---|---|
| High | Measured in the company's own data, and consistent with an independent class of source |
| Medium | Measured in the company's data, or consistent across independent sources, but not both |
| Indicative | One class of source, or a pattern too narrow to generalize from |
Register of findings
Each finding in this report, with its confidence level, the classes of source behind it, and the evidence a reader can check.
| Finding | Confidence | Source classes | Evidence |
|---|---|---|---|
| Confirmed resolution is 44%, against 70% contained | High | Conversation logs, customer interviews | Rebuilt from every conversation logged June to August 2026, matched across channels |
| 61% of ARR is platform minimums, not outcome fees | High | Billing exports, contracts | Every active contract reconciled to invoices |
| Recurring run-rate is EUR 2.07 million, against EUR 2.4 million stated | High | Billing exports, conversation logs | Connector builds excluded, and resolution fees annualized on each customer's 12-month volume profile |
| Net revenue retention is 94% outside the 5 largest accounts | High | Billing exports, contracts | The August 2025 cohort, every customer, on the company's own ARR definition |
| Of the 52 listed integrations, 3 carry 86% of conversation volume | High | Platform usage data, the company's integration catalogue | Volume by connector, trailing 12 months |
| Gross margin falls from 64% to 34% as customer volume rises | High | Billing exports, cost ledgers | Costs allocated by conversation, March to August 2026, reconciled to the monthly accounts |
| Large deals depend on the founder | High | CRM, observed sales calls | 8 of 10 deals above average contract value in the last 8 quarters had the founder on the call, and the calls observed agree |
| Losses cluster where the buyer runs a single helpdesk | High | CRM, lost-prospect interviews | Win rate 17% with 1 system against 58% with 3 or more, and lost prospects named the helpdesk's own agent as the alternative they chose |
| Buyers expect the price per resolution to fall at renewal | Medium | Customer interviews, former staff | Raised without prompting. Not yet visible in any contract, since no customer on the current pricing has renewed |
| Customers plan to move more voice volume to automation | Indicative | Customer interviews | One class of source, and not yet visible in conversation volume |
Definitions
Every metric in this report, as it is measured here. Where the company's definition differs, both are given.
| Term | Definition |
|---|---|
| Contained | Closed without a handoff to a human. What the company calls resolved. |
| Confirmed resolved | Contained, not abandoned, and no contact from the same customer on the same issue, on any channel, within 7 days. |
| ARR as stated | The company's figure: platform minimums, resolution fees annualized by multiplying the latest 3 months by 4, and connector builds. |
| Recurring run-rate | Platform minimums, plus resolution fees annualized on each customer's own 12 months of conversation volume. One-off services excluded. |
| Net revenue retention | ARR on 31 August 2026 from customers with ARR on 31 August 2025, including expansion, contraction and churn, divided by their ARR on 31 August 2025. |
| Gross margin | Revenue less model inference, hosting, telephony and messaging fees, and support and implementation staff. |
| Cost per resolution | Model inference, telephony and hosting, allocated by conversation. |
| Win rate | Closed won over closed won plus closed lost, for opportunities past discovery. |
| Average contract value | First-year contract value of new customers won in the last 8 quarters. |
| Multiple | Pre-money valuation divided by the ARR base named. |