How to Set Strategy as a First-Time CEO
Most first-time CEOs produce something they call a strategy that is actually a plan: a set of goals, some initiatives, and a revenue target. It is not useless — plans are necessary — but it does not do the thing a strategy is for, which is to tell people what to do when a situation arises that nobody anticipated. The test of a strategy is whether someone three levels down can use it to make a decision you did not know they were making. This page is about writing one that passes.
What’s inside
A plan is not a strategy
The four questions a strategy answers
The one-page format
The anti-bets
Three tests of whether it is real
The cadence
Where Blomma fits
A plan is not a strategy
The distinction is practical rather than academic.
Aplanis a list of things you intend to do, with dates and owners. It answerswhat are we doing this quarter.Necessary, and it goes out of date the moment something unexpected happens — which is continuously.
Astrategyis a set of choices about where you will compete, how you intend to win there, and what you are therefore not doing. It answershow do we decide.It is durable precisely because it is about the logic rather than the activities, and it works when a situation arises that nobody planned for.
The reason this matters is delegation. If your people have a plan and something changes, they have to come to you, because the plan is silent on the new situation. If they have a strategy, they can reason from it. Every founder complaining that their team is not strategic, or that everything escalates, should first check whether they have given anyone anything to reason from.
One consequence: a strategy that produces no clear answer to a novel question is not doing its job, however well-written it is.
The four questions a strategy answers
Strip away the frameworks and there are four. Answering them concretely is most of the work.
Who is this for?Specifically. Not “mid-market companies” — a description precise enough that your team could look at a prospective customer and say whether they are in or out. Most strategies fail here first, because a broad answer feels safer and produces no ability to say no.
What do we do for them that others do not?The honest version, not the marketing version. If the answer is “we do it better,” that is not a strategy; almost nobody wins on being better at the same thing. Look for a structural difference: a different distribution route, a different cost base, a different insight about the customer, a different starting point.
Why will this get harder for competitors to copy over time?What compounds. Distribution, data, switching costs, brand, a network effect, accumulated craft. If nothing compounds, you have a product rather than a business, and the strategy needs to name what you are building toward.
What has to be true for this to work?The assumptions underneath, stated as claims that could be false. This is the most useful and most frequently omitted question, because it converts a strategy from an assertion into something testable.
The one-page format
One page. Longer documents do not get read, and a strategy nobody has read constrains nothing.
Write the four answers, plainly, in the fewest words that are still specific. Then add three things.
The bets.Two or three things you are doing this year because of the above. Not a full initiative list — the handful of choices that follow from the logic.
The anti-bets.What you are explicitly not doing. More below, because this is the section that does the most work.
What would change our mind.For each major bet, one sentence naming the evidence that would cause you to abandon it. This is the line that makes a strategy revisable without every new data point reopening everything.
Then circulate it to the whole company, not just your executives. A strategy your organisation has not read is a document you wrote for yourself, and it will not survive contact with the first unanticipated decision.
The anti-bets
The section founders skip and the one that makes a strategy usable.
A strategy without explicit nos is a list of hopes. Everything sounds worth doing in isolation, so a document that only says yes gives nobody grounds to refuse anything — which means every opportunity gets partially pursued and nothing gets the concentration it needs.
Write them plainly: the segments you are not serving this year, the features you are not building, the channels you are not testing, the deals you will decline. Specific enough that someone can point at one and say no on your behalf without asking you.
Two things follow. Your team gains the ability to decline things, which is most of what makes a strategy operationally real. And you will find out quickly whether you meant it, because within a quarter something attractive will appear on the no list and you will want it. How you handle that moment determines whether the anti-bets have any force — taking one back without explanation teaches everyone the list is decorative.
The test for this section: if an unnamed anti-bet is being quietly attempted somewhere in your company right now, the list is incomplete.
Three tests of whether it is real
The novel-decision test.Take a decision someone made last month that was not in any plan. Could they have reasoned to it from your strategy? If not, the strategy is not doing its job — it is a description rather than a decision aid.
The no test.Name three attractive things you declined in the last quarter because of the strategy. If you cannot, you either have no strategy or you have one you are not applying. A strategy that has never caused you to refuse anything has not cost you anything, which means it has not chosen anything.
The three-people test.Ask three people at different levels what the strategy is. If you get three different answers, it has not been communicated, whatever you have written. Most founders who run this test are surprised, and the usual cause is under-repetition rather than a bad document.
Run all three quarterly. They take twenty minutes and they are considerably more informative than a strategy review.
The cadence
Strategy is not an annual document, and it is also not something you revise every time the market moves.
Annually:write or rewrite the four answers from scratch. From scratch matters — editing last year’s version preserves assumptions you should be re-examining.
Quarterly:check the assumptions. Has anything happened that contradicts a “what has to be true” claim, or triggered a “what would change our mind” condition? Thirty minutes, and it is what stops strategy drift from being discovered a year late.
Continuously:communicate it. Your repetition threshold is roughly a tenth of what the organisation needs. Reference it when you explain decisions, which is what teaches people to reason from it rather than just recall it.
After any significant new information:an explicit ten-minute check on whether the logic still holds, rather than letting everyone privately update in different directions.
And keep the old versions. Comparing what you believed a year ago with what happened is the fastest way to get better at this.
Where Blomma fits
Writing a first strategy is difficult for a specific reason: the four questions require you to commit to answers that might be wrong, in a document your whole company will read, and every instinct pushes toward answers broad enough to be safe. Broad answers are what make a strategy useless.
Blomma is an always-on AI career coach with no view on your market. Use it to pressure-test the second question in particular — whether your answer to “what do we do that others do not” is structural or just a claim to be better. Use it to write the “what has to be true” section, which is where founders discover they have been assuming something significant without noticing. Use it to draft the anti-bets, which is uncomfortable and where most of the value sits. And use it to run the three tests honestly, including the no test, which is the one that reveals whether your strategy is operating.
The second application: strategy only works if the layer below you can reason from it, and reasoning from a strategy is a skill your managers may not have. That gap is what produces escalation..
A strategy is not a more ambitious plan. It is a small set of choices, including explicit refusals, written down so that people who are not you can make decisions you were not there for.