An independent sponsor runs each deal as its own small company: it raises its own equity, borrows its own debt, and has to produce a lender package and an investor package before the exclusivity window closes. The software that does this work is not one product. It is a stack, assembled from tools built for other buyers, and its price is rarely written down in one place, because the pieces are bought at different times from different vendors and several of them do not publish a price at all.
This article is that list. It names categories, not vendors; it prices each category from public pages and published studies, with the date each price was observed; and it says where each piece stops, because the gaps between the pieces are where a sponsor's time goes.
What a sponsor runs a deal on
Six layers, in the order a deal meets them.
Sourcing and relationships. Brokers, bankers, lenders, co-investors and owners. The institutional answer is a pipeline CRM built for committed funds: relationship intelligence across thousands of contacts, email and calendar sync, sourcing databases, fund-level reporting. A sponsor whose pipeline fits on one board often runs this in a spreadsheet or a general-purpose CRM instead.
The data room. The sell-side room is the banker's. The sponsor's own room appears when the lender and the equity need the package: watermarking, per-viewer permissions, a record of who opened what. A virtual data room (VDR) holds the documents; it does not read them.
The model. The valuation, the debt schedule and the returns live in a spreadsheet. Free in the sense that the licence is already paid for, expensive in the sense that every figure in it is typed by hand.
Diligence tracking. The request list, the open questions, who owns which answer: a checklist in a shared document, a task board, or the diligence module of a larger platform. These track the question; the answer lands in the model or in an email.
The earnings work. A Quality of Earnings (QoE) that restates the seller's EBITDA with the add-backs the evidence supports. Three ways to get one: a CPA engagement, a per-deal AI report, or the sponsor's own workpaper built from the broker's add-back list. On an SBA 7(a) acquisition at a Business Purchase Price of $3 million or more, excluding owner-occupied real estate, the lender orders its own and underwrites on its earnings; our article on SBA 7(a) acquisitions after 1 October 2026 sets out the rule.
The package. The credit memo for the lender and the investment memo for equity, written in a word processor from screenshots of the model and rewritten whenever a figure moves. E-signature, fund administration and the cap table at close sit outside the deal's analysis and are not priced here.
What each piece costs
The prices in this table are ranges observed from public pricing pages and published studies in July 2026, when we last priced this market for our own pricing page, except Acquiror AI's, which are the prices published on this site. A category with no public price in that review says so.
| Layer | What it costs, and when the price was observed | Where it stops |
|---|---|---|
| Pipeline CRM built for committed funds | $10,000 to $50,000 a year and up, quote-only, commonly annual up front (public pages and published benchmark figures, July 2026) | The contact record. No earnings bridge, no DSCR model, no valuation |
| Virtual data room | $5,000 to $15,000 typical for one deal; more than 15% of deals exceed $50,000 (SRS Acquiom study of more than 3,800 deals, read July 2026) | Holds the documents and records who read them; does not read them |
| Spreadsheet model | Not priced in the July 2026 review; the licence is usually already held | Every figure typed by hand; nothing checks it against the source |
| Diligence tracking | Not priced in the July 2026 review; ranges from a shared document to a module of a larger platform | Tracks the question, not the answer |
| Quality of Earnings, CPA engagement | $25,000 to $75,000 typical for a small or lower-middle-market business (July 2026) | One deal, once; the signature is the product |
| Quality of Earnings, per-deal AI report | $2,000 to $5,000 a report (vendors' public pages, July 2026) | One deal, once; no financing model, no valuation, no memo |
| Certified valuation | $3,000 to $15,000 for a small business (July 2026) | Required when the lender or the agreement says so; a calculation does not replace it |
| Buy-side advisor | $50,000 to $100,000 typical retainer (July 2026) | Judgment and process; the sponsor still owns the files |
| Memo in a word processor | Not priced in the July 2026 review | Rebuilt from screenshots each time a figure moves |
| Acquiror AI | $139 to $2,999 a month, published, billed monthly, with a 14-day trial (as published on this site) | The analysis layer between the data room and the memo; not a data room, not a CPA signature |
One number says more than the rest. Of roughly thirty vendors researched in July 2026, two publish self-serve prices. Most of a sponsor's stack is bought on a quote, commonly on an annual contract paid up front.
Where the pieces fail to connect
A live deal produces a CIM, three years of statements, a broker's add-back list, a lender's term sheet and an investor's questions. The answers end up in six files: a valuation spreadsheet, a separate DSCR model, a QoE workpaper, a data room, a memo, and the CRM record with the deal's stage. Every figure is typed twice and every change breaks something downstream.
Three places the gap costs time.
- The retype. Revenue, EBITDA and the balance sheet are read off the CIM and typed into the model. Nothing in the stack checks that the figure typed matches the column it came from, so the check is a second person, or nobody.
- The investor's question. One question changes one add-back. The QoE workpaper changes, so the EBITDA changes, so the valuation changes, so the debt the business can carry changes, so the memo changes. Four files re-tied by hand, in the order they depend on each other, by the person who knows the order.
- Two readers, one business. The lender reads for coverage and the equity reads for value, and in a stack of separate files the two packages are two documents that happen to agree on the day they were sent, and stop agreeing the first time either is revised.
None of this is a fault in any one tool; the gap is between them, and no vendor in the table is paid to close it.
A worked example
Every figure in this section is a worked example built from the dated ranges in the table, and the totals are illustrative; a negotiated discount, proration, taxes and the year's deal count move the real figures. A sponsor screens twelve businesses in a year, takes three to a letter of intent and closes one.
Two lines do not change whichever way the stack is built. The lender wants a CPA-prepared Quality of Earnings on the closed deal, and the sponsor hosts a data room for the lender and the equity:
CPA Quality of Earnings on the closed deal $25,000 to $75,000
Data room for the closed deal $5,000 to $15,000
The stack bought piece by piece:
Per-deal AI QoE report on each of the three LOI deals
3 × $2,000 to 3 × $5,000 $6,000 to $15,000
Spreadsheet model, diligence checklist, memo not priced
Software and reports, piece by piece $16,000 to $25,000
The same year on one record. Financing and the investment memo start on Professional, so that is the plan a sponsor takes; on the annual commitment it is $559 a month, $6,708 for the year, with up to 25 deals held at a time and five seats, which covers the year's twelve deals. The pipeline CRM is not bought, because twelve deals fit the nine-stage board, and the per-deal reports are not bought, because the QoE is built in the workbench; the year of Professional is the whole software line.
Take the year of Professional from the piece-by-piece software line of $16,000 to $25,000, and the difference is what one record saves on software; it is the smallest line in the year either way. Add the two constant lines and the piece-by-piece year runs $46,000 to $115,000, and the one-record year is that total less the software difference. The CPA's engagement and the data room are most of the cost either way, so the software line is judged on what it buys: the three LOI deals were analyzed on the same record the memo was generated from, and the investor's question re-tied one figure instead of four files.
One caution on the example. Solo, at $139 a month on the annual commitment, is not the sponsor's plan: it carries the Quality of Earnings, the valuation and working capital, but not Financing or the investment memo.
Educational guide, not advice. This guide is general information, not investment, legal, tax, accounting, lending or valuation advice. Rules, ratios, prices and worked figures are stated as of the date shown, are drawn from public sources and our own experience, and can be out of date or wrong; they are not a substitute for a lender's, regulator's or program's current rules. Verify any figure with its primary source and a qualified professional before acting on it.
How Acquiror AI does this
Acquiror AI is the analysis layer between the data room and the memo. It replaces four of the six files outright, the valuation spreadsheet, the DSCR model, the QoE workpaper and the memo, and carries the fifth, the CRM record, for a team whose pipeline fits on one board. Upload the CIM and the statements, and extraction reads revenue, EBITDA and the balance sheet with a confidence score on every figure, checked against the column it came from. The Quality of Earnings workbench replaces the workpaper: periods, categorized adjustments with evidence and GL references, and a reviewer lock. The locked EBITDA feeds a three-method valuation and a financing scenario that computes the Debt Service Coverage Ratio (DSCR) on cash available for debt service, stress-tested, with a credit memo for the lender; that replaces the valuation spreadsheet and the DSCR model. The investment memo is generated from the same figures, three sections drafted by AI from approved add-backs only, exported to PDF with the firm's branding; that replaces the memo rebuilt from screenshots. The nine-stage pipeline, contacts, tasks and industry playbooks carry the deal record and the diligence list for a team whose pipeline fits on one board, and Firm and Enterprise connect Claude, ChatGPT, Perplexity and Cursor to the deal record through an MCP server.
What it does not replace. It is not a data room: keep yours for watermarking and per-viewer permissions, because in Acquiror AI every member sees every deal in the workspace, so it is not an investor or seller portal either. It is not a CPA's signature: if the lender requires a CPA-prepared Quality of Earnings, buy it, and attach the workpapers as evidence. It does not ship a transaction database; comparables are the ones you enter. It does not model a combined company, synergies or fund-level returns; the memo's returns are deal-level. It does not write the CIM, and it has no public API on any plan. A fund with a sourcing team and a thousand relationships keeps its CRM and runs Acquiror AI beside it.
The takeaway
Write the stack down with its prices and the date each was observed, because most of it is bought on a quote. Expect the CPA's engagement and the data room to be the largest lines and the software the smallest. Then ask of each piece where it stops, because the cost that appears on no invoice is the retyping between the pieces, and that is the only line a sponsor can remove. Every term in this article is defined in the glossary.