Culture as Product

The proposition that every company builds two products simultaneously: one for customers and one for its team. Culture is the latter. Coined and operationalised by Dharmesh Shah at HubSpot.

Core claim

Treating culture as a product changes what you do:

You iterate it, not preserve it. Freezing culture code is as absurd as freezing product code; the needs of employees (the “customers”) change as the company scales.
You measure it. Run NPS on employees quarterly — the same survey sent to product users. Publish all results.
You file bug reports. Treat employee feedback as bugs: some get fixed (“committed”), some get marked “works as designed” with explanation.
You have core values vs. configurable norms. Core values = federal law (inviolable, e.g. transparency). Other norms = state law (configurable per org, e.g. meeting hours, role constraints).

The aspirational liner note

Culture documents can include items that are not yet true — but must be explicitly flagged as aspirational. Dharmesh added “liner notes” to HubSpot’s Culture Code deck marking what was aspiration vs. current reality. Result: new hires read aspirations as facts and behaved accordingly. Aspirations became self-fulfilling.

This sidesteps the failure mode Airbnb experienced: they had “simplify” as a core value, discovered it wasn’t actually true, and removed it. If it had been marked aspirational from the start, they could have worked towards it rather than discarding it.

Operating cadence

Culture NPS survey every quarter
Results published company-wide (anonymised responses)
Items categorised and addressed at next all-minds meeting (all-hands)
Some fixed; some “works as designed” with explicit rationale
Culture deck iterated over time — not frozen

Relationship to company operating systems

Culture as Product is one expression of the broader idea that a company has an explicit operating system (Organisational Kayfabe and Claire Hughes Johnson on Scaling People and the Company Operating System). The product frame is distinctive in that it imports a quality-management vocabulary — NPS, bugs, iterations, product requirements — that most culture work lacks.

Where mainstream views differ

The conventional view is that founders must preserve culture as companies scale. The culture-as-product view inverts this: the job is to iterate culture, not protect it. Preservation instincts produce culture lag — the organisation’s norms fall behind the needs of the people it employs.

A second contested point is whether culture can be intentionally designed at all. HubSpot employees initially reacted negatively to Dharmesh’s culture survey on the grounds that explicit culture-building was inauthentic. His response: culture already exists in every company; articulating it is not imposing it.