Innovera: Business and Growth Analysis
An AI advisory business, sized against its delivery model
The brief asked for conditions rather than a call, and two of them govern the rest: advisor hours per engagement trending down, and the recurring component priced as its own line. Both are weeks of instrumentation on a delivery model already returning 63% at the prices buyers are paying.
Executive Summary
Innovera has built something that works and has early evidence that people will pay for it. A price that rose from free to $30-50K per case across roughly seven months, while volume rose with it, is a stronger demand signal than most companies at this stage can show. The delivery method is written down, the deliverable set is standardized, and one account has converted to an annual engagement.
The business described to investors is not the business the delivery model describes, and the gap is large enough to be found in a first diligence call.
Gross margin. Loading an advisor's whole cost into delivery the way published benchmarks do, one initiative costs $17,536: 29.9% against the $25K list price, and 41.5% to 64.9% against the $30-50K recently charged. What investors ask of an AI company in 2026 is a direction rather than a level: an observed average near 52%, an expectation of margin trending above 60%, and a company at 55% with a credible path to 65-70% described as more fundable than one claiming an inflated 75%.
That band should be applied here with care. It was relaxed because inference costs land in cost of revenue, and Innovera's inference is $250 of a $17,536 delivery cost: its cost of revenue is people. What stands without a services comparator this analysis does not hold is the arithmetic itself, and the fact that nobody has measured whether it is improving.
The 75% margin claim does not survive at either price. At $25K it allows 20.8 advisor hours for a complete business case with fourteen deliverables; at $40K, 46.7. The work takes about 80.
Market size. The $18.75B opportunity multiplies $750K per business unit by 25,000 business units. The unit count is defensible, since the Global 5000 is a real database of companies above about $1.1B of revenue. The price is not, being three times Innovera's own $250K list price on another slide. Built from that universe the addressable market is about $1.1B, on a range from $375M to $2.4B, so the claim is 16.7 times the middle.
The trajectory. Reaching $265M by 2029 at list price needs 1,060 business units under contract, 23.6% of the middle build, about 400 advisors and $104.0M of advisor payroll. At the $200K these contracts are more likely to realize at, it needs 1,325 units and about 500 advisors. Growing from $6M to $34M is 467% in a year against a top-quartile benchmark near 70%.
Innovera is a services-led business with unusually high software leverage, selling into strategy consulting spend of $63.7-68B. The nearest existing software category is $0.9-3.6B, so the $65B software category claim rests on software displacing services spend at close to parity, which has not happened in any comparable category.
Two things decide what happens next.
Whether the advisor can come out of the loop without destroying what clients pay for. Every capability that makes the output trustworthy runs through a person, and nothing in the inputs settles whether that is changing.
Whether pilots convert to annual contracts, and at what value. Converting 15% of a pool of forty puts the business near $1.2M of recurring revenue a year out. That rate is three times the one conversion observed and is not aggressive against outside evidence: enterprise platforms convert trials at about 18.6%. The pool is the weaker assumption: it holds only if the stated 80-company pipeline excludes the 20 pilots already run, and if it includes them recurring revenue is $0.6M and the funding gap is roughly six times rather than three.
On the raise. Against a Series A, the gap is $2.3M of recurring revenue on the AI-specific bar of $3.5M, though the same source says $2M with strong growth still gets funded. Against a seed, the picture reverses: the 2026 median seed is $3.1M and $4.6M for AI companies, so $5-10M is a large seed rather than a normal one, and seed investors are reported to look for $300-500K of ARR, which Innovera already clears. Which round this is therefore changes the answer more than any operating decision in this memo, and the inputs never say.
The judgment underneath all of this: the risk is not that the round fails. A real product, a rising price ladder, a named pipeline and this team will raise money. The risk is raising it on a software story that the next round tests against services economics, which is a much harder conversation twelve months from now. Everything expensive here follows from that mismatch, and the two measurements that would settle it cost days of internal work.
What follows sets out the market, the competition, the options open to Innovera, what delivery costs, and the questions an investor will ask. It renders no verdict, because the brief asked it not to, and states the conditions instead.
Units, terms and how figures are stated
Every dollar in this memo is stated on a named basis, so any two figures here can be compared directly rather than assumed to match.
This case states dollars on five different bases, and the summary above moves between three of them within a few sentences. Every figure in this memo carries its basis where it appears.
Recurring revenue (ARR). Contracted annual revenue that renews. Used only for annual engagements and enterprise licenses. Pilot fees are not recurring revenue.
Pilot fee. A one-time price for a single initiative analysis, charged per case. Innovera's ladder ran free, $10K, $15K, $25K, then $30-50K.
Annual contract value per business unit. The list price of the enterprise tier, $250K per business unit per year. Where this memo says "business unit" it means a buying unit inside a customer: a corporate strategy, new-ventures or divisional growth function with its own budget.
Value delivered. What a client saves or gains. The material asserts $150,000 per initiative. This is a claim about the client's economics, not Innovera's revenue, and the two are never added in this memo.
Category size. Total third-party spend in a defined market in a year. Consulting categories and software categories are sized separately throughout and never summed, because one is the spend the other would have to displace.
Two terms of art. Gross margin here means revenue less the direct cost of delivering it — advisor time, expert network fees and inference, and excludes sales, product engineering and general overhead. Forward Deployed Advisor (FDA) is Innovera's term for the senior practitioner embedded in each engagement; this memo calls that person the advisor.
Where a proportion appears, it carries what it is a proportion of, in the same sentence.
Section added · This case states dollars on five different bases — pilot fee, annual contract value, recurring revenue, value delivered to a client, and category size. The summary collides several of them within a few sentences
The business as it stands today
A price that rose from free to $30-50K per case in about seven months while volume rose with it is a stronger demand signal than most companies at this stage can show.
- 20+, in about six months from v1.0
- 80+
- free, then $10K, $15K, $25K, $30-50K per case
- one named (LG Innotek, expanded to five growth areas)
- four: LG Innotek, Mitsubishi Corporation Americas (carbon), a wellbeing real-estate entry, a renewables-plus-storage case
- under $1M
- ten named leaders across two decks; no headcount stated
Innovera sells the evaluation and de-risking of enterprise growth initiatives. A client brings a question of the form "should we enter this market, and how", and Innovera returns a structured, evidence-traced business case with ranked risks and a sequenced set of next actions.
The delivery has four parts. A software workspace, the Studio, provisioned identically for every initiative across nine sections in three clusters. An analytical layer, the RQA Engine, running six pillars from framing through analysis, a quality gate, risk ranking, action sequencing by value of information, and output rendering, all writing into a claims graph from which every deliverable is a rendered view. A knowledge base combining client documents with retrieved market intelligence. And a human layer: an embedded advisor who directs the work, plus domain experts pulled in against specific unvalidated assumptions.
The engagement runs four gated phases — Initiation, Market Analysis, Business Analysis, Navigate — producing fourteen defined deliverables, with a decision point at the end of each phase and advisor sign-off before anything reaches the client.
Two pricing tiers are published: $25K per initiative per year at the initiative level, and an enterprise license starting at $250K per business unit per year at the portfolio level.
What can be established about performance:
What cannot be established from anything provided: cash, burn, runway, total headcount, customer acquisition cost, gross or net retention, the number of contracts currently signed, the value of the one conversion, the basis of the $150K value claim, and the round's stage, terms or valuation.
The value of the one conversion is not stated either, but it can be bounded rather than declared unknown: the account expanded to five growth areas, and five initiatives priced at what the market has paid per case is $200,000, against a $250,000 list price. Somewhere in that range is a better working assumption than an absence, and confirming it is one phone call.
Three further things are imprecise rather than absent, and this memo uses all three, so it states them as what they are. Recurring revenue for 2025 is given as under $1M, which is a bound and is used as one. Total headcount is not stated, but ten leaders are named across the two decks, which is a floor and is used as one. The pilot sales cycle can be bounded from the ladder chart, which shows three to five engagements closing per month, while the annual-license cycle genuinely cannot be established from anything provided.
That list matters on its own account. Ten of the figures an investor works from at this stage are absent from the material prepared for investors. The analysis below works around them with estimates that are labeled as estimates and derived in the open, but no external analysis can substitute for the company stating its own numbers.
The rising price ladder is the single most valuable piece of evidence Innovera has. A buyer paying $40K for the fifth iteration of something that was free at the first is telling you the product improved and that the value is real. It deserves more prominence than the market-size arithmetic currently gets.