The revenue trajectory, tested
What the 2029 plan requires in contracted units and advisor capacity, set against the conversion evidence to date.
The stated plan:
| Year | 2025 | 2026 | 2027 | 2028 | 2029 |
|---|---|---|---|---|---|
| Recurring revenue | under $1M | $6M | $34M | $124M | $265M |
| Year-on-year multiple | 5.7x | 3.6x | 2.1x |
The 2026-to-2027 step is 467% growth. The top quartile for companies at $5-10M of recurring revenue is roughly 70%, and the median across more than a thousand private B2B software companies is 25-26%. The plan asks for 6.7 times the top-quartile rate in that year, then settles onto a still-exceptional path.
What 2029 requires, expressed in what the business sells:
| Requirement | Figure |
|---|---|
| Business units under contract at $250K | 1,060 |
| Share of the addressable buying units on the middle build | 23.6% |
| Advisors needed at 2.65 business units each | 400 |
| Advisor payroll alone | $104.0M |
| Revenue per advisor, before engineering, sales or overhead | $662K |
The 2.65 business units per advisor discounts a 2,080-hour scheduled year for 70% billable utilization, which is the middle of the 60-75% range professional services firms run at.
A 23.6% share of the addressable buying units on the middle build is not a share a new entrant takes in four years in a market with entrenched alternatives, and on the low build it is 70.7%, which nobody takes. On the high build it is 10.9%, which is demanding rather than impossible. The plan's reachability therefore depends on which market build is accepted, which is the fork the conditions section returns to. Hiring 400 senior practitioners in four years is its own constraint whichever build holds, and each one is a $260K commitment before they bill anything.
What the evidence supports instead
Three scenarios, each built as one coherent world rather than by moving a single input. A world in which few pilots convert is also a world in which the contracts that close are smaller, so conversion and realized contract value move together.
The scenarios describe where recurring revenue stands twelve months out, once the pilots already run have had time to convert and a share of the stated pipeline has landed. Stating the horizon matters because the comparison later in this section is against a dated plan figure.
The pool is the 20 pilots run plus a quarter of the stated 80-company pipeline, which is 40. Two cautions about that number, and both push the same way.
The quarter is derived from the company's own history: 20 paid pilots against an 80-company pipeline is a 25% pipeline-to-paid rate already achieved. That derivation only works if the 80 excludes the 20 it is compared against. No input says whether it does. If the stated pipeline already contains the pilots already run, the pool is 20 rather than 40, and the middle case falls from 6 contracts and $1.2M to 3 contracts and $0.6M. That reading would remove the only scenario below in which recurring revenue clears the Series A median.
The same 20 is also the denominator this memo elsewhere argues is overstated, since the deck's own ladder shows a first tranche that was free and a last tranche still in process. Using 20 where it enlarges the pool while calling it overstated where it depresses the conversion rate would be having it both ways, so it is used consistently as stated and flagged in both places.
- Conversion
- 5%
- Units
- 2
- Value per unit
- $120,000
- Recurring revenue
- $0.24M
- Likelihood
- The observed rate held flat, with contracts at two to three initiatives rather than a full license. Pessimistic, because the pilot cohort is too young to have converted
- Conversion
- 15%
- Units
- 6
- Value per unit
- $200,000
- Recurring revenue
- $1.2M
- Likelihood
- Most likely. Three times the observed rate, which is below the roughly 18.6% at which enterprise platforms convert trials and well below the 60-80% reported for structured proofs of concept, and at five initiatives priced at what the market has actually paid per case
- Conversion
- 30%
- Units
- 12
- Value per unit
- $250,000
- Recurring revenue
- $3.0M
- Likelihood
- A strong land-and-expand motion with every contract at published list. It is the only scenario clearing the Series A median, and it is also the only one pricing at a level no input evidences as signed, so it should be read as the ceiling rather than as an outcome
The realized contract value is varied rather than assumed, because no input evidences a signed contract at the $250K list price, and because this memo states elsewhere that the buyer of an annual license is not the buyer of a pilot. Pricing every conversion at list would join two things the analysis says do not join.
The 2027 plan of $34M is 28.3 times the middle case a year out, or 57 times it on the reading where the pipeline already contains the pilots. That is the figure to reconcile, and reconciling it does not require abandoning ambition. It requires stating the path from six contracts to a hundred and seventy.
Margin in the world called most likely here is worth stating, since every margin figure above is computed at list price or at pilot prices. A business unit at $200,000 covering five initiatives returns 52.2% on the narrow basis and 42.8% fully absorbed. The revenue conclusions and the margin conclusions in this memo otherwise live in different worlds.
Two cautions about the conversion rate itself. One conversion out of the pilots run is what the material states, and it should be read as a floor rather than a rate, because most of those pilots ran within the last few months and conversion takes time. Reading it as a rate understates the business; reading it as a proven motion overstates it.
And the denominator is not firm. The deck states "20+ paid pilots" and "20 pilots in 6 months", while its own ladder chart shows roughly 23 engagements across seven months, of which the first tranche were free and the last are marked in process. Paid, closed pilots are fewer than twenty, which would make the observed conversion rate higher than stated. The company should reconcile its own two figures; this memo uses the stated 20 as the denominator and flags that it is probably conservative.