The choice that decides the rest: software margins or advisory margins
Everything converges on one choice, and it decides which comparator the next funding round measures this company against.
Everything above converges here.
The evidence that this is an advisory business with software leverage: an advisor is embedded in every engagement and signs off before analysis begins and before any output reaches the client; experts are engaged against specific assumptions; the four phases each carry a human gate; the website's own promise is "growth as an embedded capability", and the client-facing material sells "an embedded Forward Deployed Advisor" as the first element of the offering. Delivery cost is dominated by a person's hours.
The evidence that this could be a software business: the deliverable set is standardized rather than bespoke, the workspace is provisioned identically every time, the analysis is rendered from a graph rather than assembled by hand, and the recomputation loop runs without a person. These are real properties and they are what makes the advisory margin better than a consultancy's.
Palantir is the precedent the material invokes, and it is instructive in both directions. It reached 82.4% gross margin in FY2025 on $4,475M of revenue with a forward-deployed model, so the model can reach software margins. It also took about two decades and enormous capital to do it, and at 3,936 employees its revenue per head is $1.14M, not $10M.
The choice is not cosmetic. It sets what the capital is spent on, since one path funds advisors and the other funds engineers, and it sets which metrics matter, since a services business is judged on utilization and repeat rate and a software business on retention and expansion. It is also the frame an investor applies to everything else in the package, though this analysis carries no figure for what that frame is worth and does not assert one.
The material currently claims the software position while describing the advisory delivery. That is the single most consequential inconsistency in it, and it is the one a diligence process is most certain to find, because the delivery model is described in Innovera's own documents.
What would settle it is a measurement nobody has taken: advisor hours per engagement, recorded across the last ten engagements, and the same figure for the most recent three. If the trend is steeply down, the software story is being earned and can be evidenced. If it is flat, the honest category is advisory-led, and there is a good business to build and fund there.