How to Evaluate a Precedent Transaction Database Before Buying Access
Test known-deal coverage and source evidence, then size search, reveal and with-multiples allowances before buying a precedent transaction database.
A precedent transaction database is useful when you can reconstruct which deal happened, which value was disclosed, and which financial period produced each multiple. Before choosing a provider, select transactions you already know and ask which checks are possible before purchase; run the full test when access is available. Check how many of those deals become usable records, not only how many appear in search. A large result count cannot make a pending deal completed or turn an undisclosed EBITDA figure into a reported multiple.
This is a buying test for M&A deal data. If you already have a dataset and need to decide which transactions belong in a valuation set, use the separate guide to select precedent transactions.
Test the record, not the headline count
Choose a small group of deals relevant to your work: a completed transaction, an announced deal that has not closed, and one with incomplete public financial disclosure. Search for each, then open the underlying record. The fields below determine whether a comparable can be used without quietly changing its meaning.
| Check | What you need to see | Failure that changes the analysis |
|---|---|---|
| Transaction status | Announced, completed, withdrawn or terminated, with the relevant event date | Announced consideration shown as the price paid at closing |
| Value basis | Equity value and enterprise value distinguished, with the source or calculation method | An offer for shares labelled enterprise value without a cash and debt bridge |
| Financial period | LTM, NTM or another stated period attached to each revenue or earnings denominator | Multiples compared across different periods as though they were the same |
| Disclosure versus calculation | A reported multiple identified separately from one calculated from disclosed inputs | A derived figure presented as if a party disclosed it |
| Currency and terms | Currency, consideration mix and contingent terms recorded where disclosed | Cash, stock or an earnout silently collapsed into one comparable cash price |
| Evidence | A way to identify and verify the public document behind each material fact | A precise-looking number that you cannot reconcile to a dated disclosure |
Do not expect every row to have every field. A blank for an undisclosed term is more useful than an estimate that cannot be separated from a sourced fact. For the distinction between a reported and calculated multiple, check whether the record keeps the inputs, arithmetic and source date together.
Keep a separate evidence column
Do not assume that a search result or revealed record displays a clickable filing or named source link. For each test deal, write down what the product actually shows and what you independently verified. Use the announcement or merger agreement for the stated offer terms, and a separate closing disclosure for completion. Note the document, event date, currency, value type and financial period next to the number you intend to use. If you cannot trace a material input, mark that input unverified; do not fill the gap with an LLM estimate or call the resulting multiple reported.
Check status against the underlying filing
For a US public-company transaction, the SEC’s Form 8-K provides a practical status check. Item 1.01 covers entry into a material definitive agreement; Item 2.01 covers completion of an acquisition or disposition when the form’s requirements apply. The two events may be reported in different filings. Find the company in SEC Search Filings, open the relevant filing and compare its event date with the database’s status and date fields.
A completion filing establishes the reported closing event; it does not by itself establish that every final consideration component or valuation multiple was disclosed. For deals outside US public filing rules, ask the provider which exchange announcement, company release or other public document supports the same status distinction. The announced-versus-completed guide shows how to keep those observations separate in a precedent set.
Test coverage for the mandate you actually have
A database can return many transactions and still miss the deal universe you need. Before buying, define a small known-deal sample independently of the provider: public transactions already identified from company announcements or a filing search, across the countries, industries, target ownership types and announcement years relevant to your mandate. Include a few deals with thin disclosure. Record the public announcement URL and date for each so that a missing search result is distinguishable from a deal you never checked.
Use the product’s available industry, geography and ownership filters to run the same searches you would use in practice. A missing result under one filter may reflect a classification choice; check reasonable alternative filters before calling the deal absent. If you cannot inspect a field until revealing a record, mark that check unknown before reveal, not present. Keep these four counts separate in a worksheet:
| Stage | Count a deal only when… | What the count tells you |
|---|---|---|
| Known candidate | You have independently identified the transaction and its public announcement | The denominator for this specific sample, not a claim about the whole market |
| Found in search | A result can be matched to the target, buyer and transaction event | Whether the search can surface candidates in your chosen segment |
| Revealed and verified | The relevant status, date and value fields can be checked against public documents | Whether the record supports the intended analysis |
| Usable multiple | EV, the denominator, currency and financial period are disclosed or reproducibly calculated on the same basis as the rest of the set | Whether the deal belongs in your multiple calculation |
Record why each candidate stops: no search match, ambiguous identity, no disclosed value, unavailable EV bridge, missing financial period, or a genuine comparability exclusion. For example, a completed deal with an undisclosed EBITDA denominator may be a valid transaction record but cannot become an EV/EBITDA observation. Do not use the share of your sample found as a market-wide coverage estimate: a hand-picked set is neither random nor exhaustive.
Use one audit row per known deal
Download the blank known-deal audit sheet as a CSV and open it in a spreadsheet. Set the target, buyer, announcement URL and intended multiple before searching the database. This keeps the sample independent of whichever provider you are testing. Add one row for each known deal, including deals that never appear in search; otherwise the worksheet will overstate how often the database finds your candidates.
The sheet separates what the revealed record displays from the public documents you check. Put the relevant announcement or closing document URL beside the status event, the value disclosure beside its amount and type, and the financial filing beside the denominator and its period. If you calculate a multiple, enter the formula and dated inputs; if a document reports it, label it reported. Leave an undisclosed input empty rather than filling it from an estimate. A public source URL in your audit sheet records your verification; it does not imply that PrivyLogic displays a source link in the buyer record.
Use the decision and reason columns consistently:
| Decision | Apply it when | Example reason to record |
|---|---|---|
| Usable for the intended multiple | Status, value basis, currency, numerator, denominator and financial period are verified and comparable to your set | “Completed; USD EV and fiscal-year EBITDA checked against dated filings” |
| Pending verification | The deal may fit, but one material field or source still needs checking | “Equity value disclosed; closing-date net debt not yet verified” |
| Exclude from the multiple set | A required input is undisclosed, or the transaction fails a documented comparability rule | “No public EBITDA denominator for the selected period” |
These are illustrative reasons, not statements about an actual transaction. An excluded deal can still supply transaction context. Enter each search request once, even if it returns more than one candidate; use 0 or 1 for each record reveal and with-multiples reveal. A revealed record with a multiple counts in both reveal columns. At the end of the trial, total the three allowance columns separately and compare them with the current plan limits. Report found, verified and usable counts separately for this sample, keeping the public-evidence check distinct from what the product displays.
Match the value and period to your workflow
If you screen by EV/EBITDA, inspect the numerator and denominator before relying on the filter. An equity purchase price is not automatically enterprise value; cash, debt and other components may change the bridge. See the equity-versus-enterprise-value explanation for the reconciliation to request from a data provider.
Then ask which earnings period the multiple uses. A trailing period at announcement and a later forward estimate answer different questions. If the period is absent, leave that transaction out of a like-for-like multiple comparison until you can establish it from the source. A database can still help you find the deal even when its financial fields are incomplete.
Check access and cost against the actual task
The right access plan depends on how your team uses records. For occasional manual searches, test which fields appear in search results, which require a reveal, and what source evidence you can independently reconcile. For a repeatable screen, confirm whether the plan permits API or MCP access. Do not assume that a browser plan includes programmatic access or that a search result includes every record field before a reveal.
Estimate three monthly quantities for your own workflow: searches to find candidates, records you expect to reveal, and how many of those reveals need valuation multiples. A deal with a multiple uses both a record reveal and the smaller “with multiples” allowance within it. Compare all three estimates with the live plan limits, leaving room to reject deals after source review. If you need an API or MCP workflow, remove Lite from the shortlist before comparing allowances.
Size all three allowances with a worked trial
The current plan comparison lists searches, record reveals and a smaller allowance for revealed records that carry a multiple. A reveal with a multiple uses both reveal allowances. Plan for the records you open while testing candidates, including those you later reject after source review; a usable-multiple count alone understates that work.
Here is a hypothetical monthly research plan, not a statement about PrivyLogic’s coverage or the contents of any specific plan. Suppose a team expects four screens, each requiring about eight searches. Across those screens it expects to open 18 candidate records; after opening them, nine turn out to carry a multiple.
- Searches: 32. Four screens × eight separate search requests to identify and check candidates.
- Record reveals: 18. Count every candidate opened, including records later excluded from the precedent set.
- Reveals with multiples: 9 of those 18. Count opened records that carry a multiple, even if later source review makes the multiple unusable. These nine are part of the 18 record reveals, not an additional 9 reveals.
Compare 32 searches, 18 reveals and 9 with-multiples reveals separately with the live monthly allowances. A plan has to fit each estimate; spare searches cannot substitute for missing record reveals. Add a margin based on your own reruns and rejected candidates, then repeat the exercise for another mandate if its countries, disclosure levels or filter choices differ. Before access, the with-multiples figure is only a planning assumption: record your actual mix during the known-deal trial and revise it. A displayed multiple still needs the value, currency, financial period and public-source checks above before it enters a valuation set.
PrivyLogic offers a live browser database of sourced precedent M&A transactions. API and MCP access are available on Core, Pro and Enterprise, and are not included on Lite. Compare the current plans and allowances against the number of searches and record reviews your workflow actually needs. No public record count or completeness claim is implied.
When you can inspect records, run the same known-deal test with any provider you are considering. Keep a short audit sheet with the record’s displayed fields, independently checked public source URL, status, value basis, financial period and any unanswered question. A useful decision is usable, candidate pending verification or exclude from the multiple set. A missing public denominator may still leave a transaction useful for context, but it does not support a calculated EV/EBITDA multiple. The sheet tells you whether the database can support a defensible precedent set for your mandate, rather than only whether it returns a long list of names.
Sources
- Form 8-K — Current Report — U.S. Securities and Exchange Commission; accessed October 6, 2026
- Search Filings — U.S. Securities and Exchange Commission; accessed October 6, 2026
- PrivyLogic plans and monthly allowances — PrivyLogic; accessed October 18, 2026