What Culture Diligence Can Observe Before the Term Sheet

Culture work before a term sheet is not the work most guides describe. Confidential surveys, focus groups, skip-level interviews: that program needs a signed deal and a sponsor inside the company. Earlier, a buyer has documents, a few management meetings, public signals and whatever people data the seller shares in aggregate. Thin, but not empty. Enough for a preliminary view and a list of hypotheses. Not enough for a verdict. The distance between those two is the discipline.

Before the Term Sheet, Culture Diligence Sees Artifacts and Stated Values

Edgar Schein's levels of culture still sort what is visible. The surface holds artifacts: office layout, manner of dress, visible behavior patterns, public documents such as charters and employee orientation materials. Underneath sit the basic assumptions Schein treats as the core. His warning about the surface is the one a pre-deal review should carry: that level of analysis is tricky, because the data are easy to obtain but hard to interpret.

So collect the artifacts, and read them as evidence about intent, not behavior.

• What the company publishes about itself. The values page, the careers page, the onboarding pack, the handbook, the all-hands deck if it circulates outside.
• What governs people decisions. Promotion criteria, the performance cycle, who signs off on a hire, a raise, a title change.
• Where the artifacts contradict each other. A stated value of autonomy next to an approval chain where one person clears every exception is not a culture finding yet. It is a question with a name.
• Job postings over time. What the company asks for, what it stops asking for, which teams it staffs ahead of the plan.

This is the raw material for culture due diligence at this stage, and it is where human capital due diligence begins while access is still being negotiated.

Management Meetings Are Self-Description, So Test Them for Consistency

Most of what a buyer hears about culture before a term sheet is the leadership team describing itself. Advisory work on people risk in transactions names the problem plainly: deal teams have historically relied on leaders to describe their own style and the broader organizational culture, which invites bias, since leaders may omit or misrepresent key details, on purpose or not.

Treat each answer as a claim with an owner, then test it.

Ask the same question of several leaders separately and compare the versions. Ask for events, not adjectives: the last decision that split the leadership team, who made the final call, how long it took. Ask who left the leadership team most recently and what they said on the way out. Ask for the calendar rather than a description of how the company meets. Where a claim can be checked against an artifact, check it.

Practitioners writing on culture clashes in deals put the ceiling honestly: pre-deal work may rest on more anecdotal evidence because access to individuals is limited, but a preliminary view can usually be formed. That is the expectation to set with an investment committee. The formal management team assessment comes later, with real access. What the chart claims about authority is its own question: see the org chart versus reality.

Public Employee Reviews Are Evidence With a Known Skew

Public employer reviews are the only unmediated employee voice available before signing. They also carry a measured bias: a study of employer review data concluded that voluntary employer reviews often suffer from selection bias, because people holding extreme opinions are more motivated to write than people holding moderate ones.

That does not make them useless. It makes the reading method matter.

• Themes, not scores. A composite number over a skewed sample is noise. Recurring language is not.
• Cut by role and by time. A complaint that shows up in engineering and in support, across several years, is a pattern. One angry review is one person.
• Look at the advice-to-management field. That is where employees describe decision rights.
• Distrust a sudden burst of praise. Review platforms prohibit coached reviews, which tells you coaching happens. Positive volume clustered in one month says something about the company's posture, not its culture.
• Compare with what management said. Where the employee account and the self-description diverge, you have a hypothesis worth a confirmatory test. Operating data versus the management presentation is the same move in another workstream.

The People Data Worth Requesting Before Exclusivity

Ask early, ask in writing, and ask for definitions alongside the numbers. Every people metric is a definition before it is a measurement, and a definition you did not read is a finding you cannot use.

• Attrition cut by team, tenure band and manager, voluntary separated from involuntary, with the leaver categories defined. Research using employee reviews and employment data found corporate culture a more reliable indicator of industry-adjusted attrition than how employees assess their compensation. This cut comes before pay benchmarking.
• Regretted and unregretted leavers, and who decides which is which.
• Span of control and layer counts from the HR system, not from the chart drawn for the sale process.
• Promotion and internal-move history for the roles the value creation plan depends on.
• Hiring funnel evidence: time to fill, offer acceptance, which roles stayed open longest.
• The engagement survey history: the instrument, the question wording, participation, and whether results went back to staff. A company that ran a survey and buried it has told you something.

Expect aggregates. Where the buyer competes with the company, that is not evasion. Competition regulators are explicit that premerger exchange must be managed: share the least information needed, mask identities, aggregate competitive data, and establish clean teams and third-party consultants for sensitive material. Employee records carry data protection obligations too. What applies to a given deal belongs with deal counsel early, since it shapes what the request list can ask for.

Aggregates still show concentration. Attrition inside one team, a single manager's span, one unfilled role on the critical path: that is the input to key person risk and to a talent risk matrix.

What Culture Diligence Cannot Conclude From Outside the Company

Name the ceiling in the memo, in the same register as the findings.

• Underlying assumptions. Artifacts and leader interviews do not reach what Schein calls the core. You will not learn why the group behaves as it does.
• Subcultures. A plant, a sales region and an engineering group can differ more from each other than the two companies differ overall. One leadership sample hides that.
• Behavior under pressure the company has not met. Nothing seen in a benign quarter establishes how decisions get made in a bad one.
• Whether the people you met stay. Intentions stated to an incoming owner are not retention.
• Any forecast of returns. A culture observation is an execution risk to test, never an outcome to predict.

Each of those is a reason to label confidence on the finding. None is a reason to skip the work.

Judge Culture Against the Investment Thesis, Not a Generic Ideal

No culture is good in general. A meta-analysis of cultural differences in mergers and acquisitions found those differences affecting sociocultural integration, synergy realization and shareholder value in different and sometimes opposing ways, with effects varying by how related the businesses are and which dimension of culture separates them. Differences that wreck one deal are the reason for another.

So read culture against the plan.

If the thesis leaves the business autonomous, distance from the sponsor matters less than internal coherence. If the thesis consolidates functions, merges a sales motion or moves work between sites, the specific dimensions that differ are the ones that decide how hard team integration becomes. If the thesis depends on faster decisions, then decision rights and escalation behavior are the culture questions, and they sit inside the operational due diligence scope, not a separate people chapter.

Turn Each Observation Into a Test for Confirmatory Diligence

Nothing from this stage should leave it as a verdict. Write each item as four lines: the observation, the evidence behind it, the competing explanation, and the test that would settle it.

• Observation: review language and job postings both point at churn under one manager. Test: attrition by manager, plus references from leavers in that team.
• Observation: leaders describe consensus decisions, the approval policy shows a single signer. Test: walk three recent decisions end to end with the people who executed them.
• Observation: no promotion history above a certain level. Test: internal-move data, and conversations with the senior people hired from outside.

Those tests need access: named interviews, skip-levels, HR extracts, formal evaluation of the leadership team. All of it arrives after a term sheet or letter of intent, in the window between exclusivity and signing, where the leadership work that needs cooperation belongs.

Write the hypotheses down before that window opens. Due diligence that enters exclusivity with a culture question list tests it. Diligence that enters without one asks management how things are going, and writes down the answer.