Core Values

At ten people, nobody needs these sentences. When something ambiguous comes up you ask the founder, and everyone has watched the founder decide often enough to guess what the answer will be.

At fifty, that breaks. The founder can't sit in every decision, and the same situation now gets different answers from different people. Sales, with a large contract in reach, wants to grant the exception a customer asked for. Engineering refuses, because an exception made once has to be offered to everyone else. Both are arguing for the company and pointing in opposite directions. The meeting burns two hours, ends without a decision, and when a similar request arrives the following week the argument starts over from scratch.

Worse is when the answer varies by team. One team grants the exception, another blocks it, and from the customer's side the treatment they get depends on who picked up their ticket. Internally nobody is confident about which behavior the company actually wants.

Core values are that argument, answered once in advance. They don't decide individual cases, but they fix what carries more weight. When the number of people making judgment calls goes from one to fifty, this is the device that keeps those fifty leaning the same way — the founder's reasoning, written out instead of absorbed by sitting nearby.

A company settles on three or four sentences and holds decisions up against them when the call could go either way. The count stays small because more than ten sentences is more than anyone remembers, and a sentence nobody remembers never gets raised in a meeting.

Writing them takes three steps.

1. Look backwards at decisions already made. Pull out ten hard calls from the past two years and write down which way each one went. Values are less about describing who you want to become and more about putting words to judgments you are already making.
2. Write down what each sentence costs. Next to every candidate, answer in one line: what do we give up by holding to this? No answer means it's a pleasantry, not a value.
3. Wire it into a process. Make the sentence show up in interview questions, promotion criteria, and quarterly retrospectives. A sentence that lives only in a document and hooks into no process is forgotten within six months.

Step 2 is the test that separates a working value from wall decoration. Check whether the opposite sentence makes sense. The opposite of "we put customers first" is "we put customers last," which no company would claim, so the sentence can't tip any decision. "We don't hand over customer data, even when revenue depends on it" has companies on the other side of it, and holding to it costs a specific contract. Only a company that has actually lost one reaches for that sentence in the next decision.

Hiring is the most frequent setting. A culture interview — a separate round for checking whether someone's working style and beliefs line up with the company's — draws its questions straight from the value statements. With a value about customer data, the question becomes "tell me about a time a customer asked for something outside the rules," and the answer reveals what that person has been willing to give up.

Performance reviews use them too. Look only at numbers and whoever beat their target always comes out ahead. If there is no agreement on how to treat someone who hit 120 percent of target while cutting against a value, the team learns within one quarter that only the number counts.

Product decisions pull them in as well. Whether to run ads for free users isn't something an experiment answers. Ads lift revenue this quarter and push some free users away, and no metric decides which of those matters more. The sentence that opens that meeting is a value.

They're often written alongside a mission and a vision. The mission is why the company exists, the vision is what it's trying to become, and core values are what it refuses to trade away on the way there. The first two set direction; this one draws the line you don't cross walking it.

"This runs straight into our second value."

A brake: the proposal crosses a line the company committed to, so it goes back for another look.

"It's in the values but it isn't in the review criteria."

The sentence exists, but nothing in any process depends on it, so it isn't really operating.

"This quarter's values award goes to Theo."

Picking someone publicly makes concrete which behavior the company is actually asking for.

One of Pagedeck's values is "customer data comes before revenue." It landed in a meeting when an enterprise account Sofia was closing asked to export document contents into an outside analytics tool. The contract was worth $45,000 a year, and Theo said the export would take two days to build.

Grace declined it in the room. Not because the feature was dangerous, but because bending the sentence once leaves nothing to hold on to when a bigger contract arrives next quarter. Sofia lost that deal, and since then she answers the same request on the spot instead of calling a meeting.

The most common mistake is assembling a list of nice words and calling it a set of values. Honesty, growth, teamwork — they fit any company and nobody objects to them, so they look like safe picks. But a sentence nobody would argue against can't take a side when a decision splits. Values built that way end up on the wall and the careers page while the meetings that actually divide never mention them. A year later most of the team can't name a single one.

Believing that writing values down makes people follow them is the other frequent error. It was announced, it's in the onboarding doc, so it feels covered. But people learn from which behavior gets rewarded, not from which sentence got published. Watch someone who cut against a value get promoted on results, and that scene becomes the standard instead of the sentence. For values to work, what the person who held to one gained, and what the person who ignored it lost, both have to be visible.