Ricardo Semler once served on the jury of a best-workplace award, and the experience, as he has told it, cured him of the genre. The prize went to a bank where the tellers were not allowed to sit down. The runner-up served kombucha. Somewhere between the standing tellers and the fermented tea, the jury had lost track of the only question that mattered: what is it actually like to work here when nobody is measuring?
The story is funny until you examine the business model behind it. Then it is just economics.
The customer is the subject
Take the most established badge in the market. Great Place To Work's own certification FAQ is admirably plain about the mechanics: the company pays a program fee that 'covers the cost of the survey and the data analysis needed to produce a Great Place To Work Certification profile.' The survey is administered to the company's employees, the threshold is that roughly 7 out of 10 respondents report a consistently positive experience, and the resulting certification 'is valid for 12 months,' after which it must be renewed.
Each element is defensible alone. Surveys cost money to run; someone must pay; an independent administrator is better than a self-run poll. But assemble the elements and a structural fact emerges: the entity being evaluated is the paying customer, the evaluation happens because the customer initiated it, and the deliverable is a marketing asset with a subscription cycle. In any other assurance market this arrangement has a name: an unresolved conflict of interest. Financial auditing spent a century building rules against exactly this, with mixed success, and auditing at least involves adversarial standards and legal liability. Culture certification involves neither.
The selection effect does the quiet work. Companies choose whether to enter, when to enter (rarely mid-layoff), and whether to publicize the result. The universe of certified workplaces is therefore not the universe of good workplaces. It is the universe of workplaces that expected to pass, paid to try, and liked the answer. A badge drawn from that pool can be perfectly accurate on its own terms and still tell an outside observer very little.
What a seven-out-of-ten proves
Even taking the survey at face value, consider what it measures: sentiment, at a moment, among current employees, who know the company is being scored. It does not measure what former employees would say, what the lowest-paid tercile would say anonymously about their manager specifically, or what happens to the person who gives the wrong answer to the wrong question in a meeting. Sentiment surveys are real instruments; the peer-reviewed literature on workplace culture uses them too. The difference is that research designs exist to catch distortion, and marketing designs exist to survive it.
The deeper problem is Goodhart's old trap. Once the badge matters for recruiting, the rational company manages the badge: timed pulse campaigns, reminders from managers, a communications push in survey week. None of this is fraud. All of it converts the measurement from a thermometer into a thermostat, an instrument the organization adjusts rather than reads.
The tell that costs nothing
Semler's own company never had a plaque, as he tells it; it had a hammock in the middle of the office. The contrast is the point. A plaque is a claim addressed to outsiders. A hammock (or open books, or the freedom to disagree with a boss in public) is a condition experienced by insiders, visible to any visitor, unpurchasable by the marketing department alone.
That suggests the practical test for candidates and journalists reading employer badges: treat the badge as evidence of budget, then look for the unpurchasable signals. Ask what employees can veto. Ask what information the company shows its people that it does not show the press. Ask the receptionist how long she has worked there, and watch her face. Culture is the sum of what happens when no jury is visiting.
The award market will keep growing, because it solves a real problem for its actual customers, and its actual customers are employers with hiring targets, not workers choosing where to spend their lives. Workers get to use a cheaper instrument: the company as it behaves when it is not for sale. The bank with the standing tellers had the trophy. The tellers had the data.
Sources
- Great Place To Work (accessed 2026). Certification FAQ. Certification is a paid program (fee covers survey and analysis); threshold is ~7 of 10 employees reporting a consistently positive experience; the badge is valid for 12 months. https://www.greatplacetowork.com/solutions/certification-faq