A preliminary market assessment for an investment committee answers four questions: how big the market is on the definition the deal needs, how fast it grows on that basis, who captures the value, and what would kill the thesis. Spend day one on published figures and their definitions; spend day two on value capture and a two-page memo.
Why does a 48-hour market view decide access to a deal?
274 days: the median time to close a North American private equity deal in the second quarter of 2026, up from the 200 to 220 days typical between 2018 and 2021, on With Intelligence data. The process is long. The decision to enter it is not. Before a fund commits adviser fees, management time and a bid, the investment committee (IC) decides whether the market behind the teaser deserves that spend, and the deal team makes the case on the timetable of the next committee meeting, not the timetable of the process. This means the market view is the first gate in the deal, and a team that clears it late may never see the second.
82%
Competition shortens the clock further. BDO's 2026 survey of 400 US private equity professionals found 82% expecting deal prices to rise over the next 12 months, citing tight competition for too few quality assets, and 48% naming the hiring and keeping of the right people as their biggest constraint on executing deals at speed. The scarce input is analyst hours, and a market view that consumes two associates for a week is a cost the fund pays again on every deal it does not win.
The answer is a narrower task, done properly. A preliminary market assessment is a decision to spend, not a decision to buy, and it should be built for that decision alone.
What does the committee need to see at 48 hours?
Four answers, each with its basis. How big is the market, on the definition that matches the target's revenue? How fast does it grow, on that same definition? Who captures the value as it grows: the target's segment, or a supplier, customer or incumbent beside it? And what single fact, if true, would make the committee pass?
Everything else belongs to diligence. Customer interviews, cohort data and a bottom-up model by channel test a thesis; they take weeks and they cost money. The preliminary view decides whether that spend is justified, and tells the diligence team exactly what to test.
| Question | At 48 hours | After commercial due diligence |
|---|---|---|
| Market size | A band on the target's definition, each source shown with its publisher, date and basis | A bottom-up build tied to the target's customers, volumes and prices |
| Growth | A base-case rate on the same definition, and what would move it | Growth by segment and channel, tested against customer interviews |
| Value capture | Who holds the margin today, and who is moving to take it | Share, pricing power and win-loss evidence |
| Valuation context | Recent multiples, each with its basis: trailing or forward, revenue or EBITDA | Comparable transactions adjusted for the target's growth and margin |
| Kill risk | The one fact that would make the committee pass, written as a test | The test run, with primary evidence |
| Gaps | What is not known, and how diligence will find it | Closed, or priced into the bid |
This means the 48-hour document is short by design. A committee can act on a band, a growth rate and a named risk. It cannot act on forty pages nobody had time to read.
How should the 48 hours be spent?
Front-load the definitions; back-load the writing. The common failure runs the other way: a day building a market model from first principles, a night formatting it, and no time left for the question the GP will ask first.
- Hours 0–2: write the deal's market definition in one sentence (what the target sells, to whom, where) and the four questions the memo must answer.
- Hours 2–12: collect every published size estimate with its publisher, date and basis, and mark each difference as definitional (different things counted) or genuine (the same thing, measured differently).
- Hours 12–20: choose the band whose definition matches the target's revenue line, and take the growth rate from sources that count the same thing.
- Hours 20–30: map who captures the value, from suppliers and customers to incumbents moving into the segment, and look for the strongest evidence against the thesis rather than for it.
- Hours 30–36: pull the valuation context (recent transactions and trading multiples) and state the basis of every multiple.
- Hours 36–44: write two pages in which every header is a conclusion, every figure carries its source, date and basis, and the last section lists what is unknown and how diligence will find it.
- Hours 44–48: have someone outside the deal team read it as the GP would, and answer every question they raise in the text.
The first twelve hours are sourcing, and sourcing is the part that no longer has to take twelve hours. A cited analysis that already sets the published estimates beside each other, with their bases, covers most of steps two and three; a Caspr Market Research Study returns one in under 15 minutes. The deal team's hours then go to steps four and five, where judgment is the input the committee is paying for.
Which market size goes in the memo when the published figures disagree?
The one whose definition matches what the target sells, with the others beside it and the reason they differ. Published estimates for one market routinely sit several times apart, and the gap is usually definitional: one publisher counts software alone, another adds services, a third sizes a wider category the target never sells into. AI clinical documentation shows the pattern at full scale.
| Publisher | What it counts | 2025 | 2026 | Use in a 48-hour view |
|---|---|---|---|---|
| Mordor Intelligence | AI in clinical documentation | $0.98bn | $1.15bn | Benchmark for a pure-play ambient scribe vendor |
| Fortune Business Insights | Generative AI for clinical documentation | $0.79bn | $1.05bn | Narrower cut: 33.3% CAGR to 2034, off a smaller base |
| The Business Research Company (secondary listings) | Clinical documentation improvement, including legacy software, coding and services | $4.01–4.30bn | $4.65–5.16bn | Adjacent market: context, not the target's revenue pool |
The largest figure is four to five times the smallest, and none of the three publishers is wrong; they count different things. Splitting the difference would produce a number that describes no market at all. For a pure-play ambient scribe vendor, the committee needs the narrow band and one sentence on why the larger figure exists. Making that call explicit, source by source, is how analysts reconcile market size estimates that don't match.
A genuine disagreement is different: two sources counting the same thing, for the same year, that still do not agree. It goes in the memo as a range with both publishers named, never resolved quietly in a footnote.
What would make the committee pass?
Name it before the committee does. Every thesis has one fact that would end it, and it is rarely the market's size. It is who captures the value as the market grows: a supplier with pricing power, a customer consolidating its purchasing, or an incumbent turning the target's product into a feature of its own.
For a medical device distributor, the threat sits upstream. Caspr's Study of European medtech distribution records Intuitive acquiring the businesses that distributed its da Vinci and Ion systems in Italy, Spain, Portugal, Malta and San Marino, announced in January 2025 and completed in March 2026, with more than 470 da Vinci systems installed across Italy, Spain and Portugal at the end of 2025. Once a supplier's installed base reaches that density, owning the channel pays. A distributor whose largest line can be taken in-house carries that risk in every year of the hold, and the preliminary view should say so on page one.
Valuation context follows the same discipline: every multiple carries its basis. Caspr's Study of vertical SaaS valuations anchored private SaaS M&A at Software Equity Group's Q2 2026 median of about 4.0x EV/TTM revenue and kept forward multiples separate throughout; its own illustration puts one company at 6.0x trailing and 4.5x forward revenue, which is not the same price as 4.5x trailing. A memo that blends NTM and TTM revenue multiples misprices the entry before diligence starts.
Write the kill question as a test diligence can run. “Does the target keep its customers where the incumbent's own feature is live?” is a test. “Competitive risk” is a heading.
How should the two pages read?
As conclusions, in the committee's vocabulary. Each header states a finding and what it means for the return, so a partner who reads only the headers still has the view. “The pure-play market is about $1 billion in 2026; the $4 billion-plus figures count a wider category” is a header. “Market overview” is not.
Under the headers sits a source table: every figure with its publisher, date, basis and what it counts. When the GP asks where a number came from, the answer is already on the page, and the same table becomes the first draft of the market section of the investment memo if the committee says proceed.
Close with three lines: what is known, what is not, and what diligence must establish before a final bid. Written-down gaps are the strongest way to defend a market size estimate in the room, because the hardest questions a committee asks are about what the team did not know.
Two pages, four answers, one named risk. That is a preliminary market assessment a committee can act on in the meeting it was written for.
Worked example
AI clinical documentation: the 48-hour view on a pure-play scribe vendor
Take a teaser for a US ambient scribe vendor. Three published market sizes for 2026 are within easy reach: $1.15 billion from Mordor Intelligence (AI in clinical documentation), $1.05 billion from Fortune Business Insights (generative AI for clinical documentation) and $4.65–5.16 billion attributed to The Business Research Company (clinical documentation improvement). Caspr's Study of the market, with evidence as of August 2026, classified the gap as definitional: the larger figure counts legacy documentation-improvement software, coding and services, not ambient AI revenue. For screening pure-play vendors, its conclusion was a 2026 band of roughly $1.0–1.5 billion and a base case of growth in the low 20s percent a year (Mordor: 21.46% a year to 2031), with upside only where coding or chart-review revenue is already evidenced. The decisive finding was not the size. Epic rolled out AI Charting on February 4, 2026, drafting notes and queuing orders inside the electronic health record (EHR) clinicians already use, and reported its chart-summary feature in use more than 16 million times a month, nearly three times its November 2025 level. The Study concluded that pure ambient transcription is unlikely to remain a durable premium category as EHR owners and lower-cost entrants compress differentiation. GAO added the evidence caveat on July 16, 2026: there are few independent studies of these tools' accuracy. The committee's view fits in three sentences. The pure-play pool is about $1 billion and growing in the low 20s. The EHR incumbents are drawing it into their own product. Proceed only if diligence shows the target keeping customers where Epic's feature is live.
The published figures
| Publisher | Figure | Period | What it counts | Published | Last verified |
|---|---|---|---|---|---|
| Mordor Intelligence | 1.15 (USD billion) | 2026 | AI in clinical documentation, global market value | 2026-04-27 (page last updated) | October 17, 2026 |
| Fortune Business Insights | 1.05 (USD billion) | 2026 | Generative AI for clinical documentation, global market value | 2026-09-28 (page last updated) | October 17, 2026 |
| The Business Research Company, in secondary listings read by the Caspr Study | 4.65–5.16 (USD billion) | 2026 | Clinical documentation improvement, including legacy software, coding and services | Undated secondary listings; Caspr Study evidence as of August 2026 | October 17, 2026 |
Why the figures differ
- definitional Mordor Intelligence ($1.15bn, 2026) and Fortune Business Insights ($1.05bn, 2026) Fortune sizes generative AI for clinical documentation only, a subset of the AI category Mordor counts; the near-identical totals hide different scopes and growth rates.
- definitional Mordor Intelligence and Fortune Business Insights (about $1bn, 2026) and The Business Research Company ($4.65–5.16bn, 2026) The larger figure sizes clinical documentation improvement, which includes legacy software, coding and services; it is an adjacent market, not the revenue pool of an ambient scribe vendor.
Frequently asked
How long should a preliminary market view for an investment committee be?
An investment committee's preliminary market view should run two pages, plus a source table. The pages carry four conclusions: size and growth on the target's definition, who captures the value, and the fact that would make the committee pass. Add a short list of what diligence must establish. Anything longer is diligence done early, at judgment's expense.
What is the difference between a preliminary market view and commercial due diligence?
A preliminary market view tells the fund whether to spend money on a deal; commercial due diligence tests the thesis once it has. The market view runs on published evidence, takes days and ends with a named risk. Due diligence adds customer interviews, primary data and a bottom-up model, and runs for weeks inside the process.
How do you move faster on commercial due diligence in a competitive auction?
Speed up commercial due diligence by doing the market work before the process asks for it. If your preliminary view already lines up the published figures with their bases, diligence starts from a defined thesis and one named risk, not a blank page. The time saved goes where it decides the outcome: testing value capture with customers.
Which market size should go in an IC memo when the sources disagree?
In an IC memo, use the market size whose definition matches the target's revenue, shown beside the others with what each counts. If the gap is definitional, say so in one sentence. If two sources count the same thing and still disagree, show both as a range, publishers named. Never present a blended figure no publisher measured.