What a Strategy Actually Is
Diagnosis, guiding policy, coherent action. Everything else on the slide is ambition.
Your VP opens the quarterly review with a slide titled Strategy. It says: grow revenue 40 percent, win mid market, become the platform customers trust. Nobody in the room objects, because there is nothing to object to. Six weeks later two teams are building overlapping features and a third is quietly rewriting the billing service, and every one of them can point at that slide and say they are on strategy.
That slide is a set of goals. Goals are things you want. A strategy is how you intend to get them given that the world is not cooperating. Richard Rumelt calls the difference the kernel, and once you have seen it you cannot unsee how many documents called strategy are missing two thirds of it.
The three parts, and what happens without each
The kernel of a strategy contains three elements: a diagnosis, a guiding policy, and coherent action.
The kernel
- Diagnosis
- A claim about what is actually going on. It names the critical obstacle and throws away most of the noise. A good diagnosis is falsifiable: someone could look at the data and say no, that is wrong.
- Guiding policy
- The approach you will take to deal with that obstacle. It rules things out. If it does not exclude an option that a reasonable person would have picked, it is not guiding anything.
- Coherent action
- A set of moves that support each other rather than compete for the same people and budget. Coherence is the hard part, because each action taken alone can look sensible while the set pulls apart.
Diagnosis is where most teams skip. It feels like analysis paralysis to spend two weeks arguing about what is wrong when leadership has already told you the number to hit. But the diagnosis is the load bearing piece. If you get it wrong, your guiding policy is a well organised way of solving the wrong problem, and your actions will be coherent in service of nothing.
Notice also that the diagnosis has to be singular enough to act on. Saying activation: The moment a new user first gets the value the product promised. is weak, churn: The rate at which customers stop paying or stop using the product. is high, and sales cycles are long is a list of symptoms. A diagnosis says which of those is causing the others.
Goals dressed as strategy
- Become the leading platform for mid market finance teams
- Improve NPS by 15 points
- Ship AI across the product
- Double enterprise ARR
- Every team gets a slice of the ambition
A kernel
- Diagnosis: we win the trial and lose the rollout, because the second and third user in an account never get set up
- Guiding policy: make multi user setup the product's strongest moment, at the cost of single user polish
- Actions: shared workspace defaults, invite flow in onboarding, admin reporting, sales demo rebuilt around a team of five
- What we are not doing this year: the mobile app, the public API
Worked example
Hypothetical: Ledgerly, a bookkeeping tool
Ledgerly's board wants 40 percent growth. The team digs into the funnel and finds that signups are fine and trial to paid is fine, but accounts with a single active user churn heavily in month four while accounts with three or more active users almost never do. That is the diagnosis: the product is bought by one person and never spreads, so it stays a personal tool and gets cancelled when that person gets busy. The guiding policy follows: treat account expansion inside the company as the growth engine rather than new logo acquisition. Now the actions pick themselves, and so do the refusals. The requested dark mode does not survive contact with that policy. Neither does the integration a large prospect asked for, unless it happens to pull a second user in.
One test before you circulate a strategy document. Hand it to someone on the team and ask them to name a plausible, popular, well argued project that this strategy kills. If they cannot, you have written a mission statement. Strategy is a set of bets, and a bet you cannot lose is not a bet.
The second test is coherence. Lay the actions side by side and ask whether the output of one makes the next one easier. In the Ledgerly example, shared workspace defaults make the invite flow worth building, and the invite flow makes admin reporting meaningful. That mutual reinforcement is what separates a strategy from a to do list that happens to share a theme.
Quick check
Why is the diagnosis the load bearing part of a strategy?
The takeaway
A strategy is a diagnosis of the real obstacle, a policy that rules options out, and actions that reinforce each other.
Try this tomorrow
Take your current roadmap and write the one sentence diagnosis it implies. Then ask two teammates to write theirs from the same roadmap, and compare.
Answer the check above, then bank the day.
Where this comes from
- Good Strategy Bad Strategy, Richard Rumelt
- Escaping the Build Trap, Melissa Perri
Strategy is one of six tracks. These lessons summarise and build on the work above, they do not reproduce it. Buy the books, they are better.