What if the five-year plan is the thing holding you back?
The strongest operators we interview rarely followed a plan. They followed a short list of rules about what they would and would not do — and let the path assemble itself.
| Format | News, with a verdict |
|---|---|
| Prompted by | Molly Graham, 'What if you're not supposed to have a long-term plan?', Lenny's Newsletter, 4 Aug 2026 |
| Verdict | Replace the five-year plan with three or four rules you will not break |
| Applies to | Operators, department heads and anyone hiring them |
Ask an operations manager where they will be in five years and you will usually get a job title. Ask them how they got to the job they have now and you will get a story with at least two accidents in it. Both answers are honest. Only one of them is useful.
Molly Graham made a version of this argument last week in Lenny's Newsletter, and it is worth reading in full at the link in the sourcing note. Her case is that careers behave less like projects and more like emergent systems: a few simple local rules, repeated, produce structure nobody designed. We want to take that idea somewhere slightly different — not to careers, but to the way the companies we cover decide what to build.
The plan is a forecast wearing a costume
A five-year plan is a forecast that has been promoted to a commitment. That promotion is the problem. A forecast can be wrong without embarrassing anyone; a commitment cannot, so it gets defended long after the evidence has moved. We see the same pattern in deployment after deployment. The plant that committed to a three-year automation roadmap in 2023 is still explaining why phase two slipped, while the plant that committed only to "we will not automate a process we cannot already measure" quietly finished four cells.
“The rule survived contact with reality. The roadmap did not.”
That second sentence is a rule, not a plan. It does not say what will be built, in what order, by when. It says what will disqualify an idea. Rules of that shape are cheap to hold and expensive to break, which is exactly backwards from a roadmap — cheap to break, expensive to hold.
What a good rule looks like
The useful ones are specific enough to reject something. "We will be data-driven" rejects nothing. "We will not start a project whose baseline we are not already logging daily" rejects most of the pilots that fail. In the plants and retailers we have documented, the rules that actually did work were almost boringly concrete: never deploy a model that cannot physically act on its output; never buy a system whose failure mode is a person quietly ignoring an alert; never fund a pilot without naming the person who owns retraining.
Notice that each of those is a constraint, not an ambition. Constraints compose. Ambitions compete. Three ambitions in one department produce a prioritisation meeting; three constraints produce a shorter list of viable options, which is what you actually wanted.
Where the emergent argument gets abused
It would be convenient to read all this as permission to stop planning. It is not. Emergence produces structure only when the local rules are good and the feedback is fast. Remove either and you do not get a murmuration, you get drift with better branding. The companies that improvise well are the ones measuring weekly. The ones that improvise badly are usually measuring annually and calling it agility.
So the honest version is narrower than "throw away the plan". It is: hold the destination loosely, hold the rules tightly, and shorten the interval at which reality is allowed to argue with you.
The verdict
If you run a department and you are being asked for a five-year plan, write it — that is organisational politics and it is not going away. Then write the three rules underneath it, circulate those instead, and judge every proposal against them. Within a year the rules will have shaped more decisions than the plan did, and the plan will have been revised twice without anybody minding.
For hiring: stop asking candidates where they see themselves in five years. Ask what they have refused to do, and why. The answer tells you whether they have any rules at all, which is the thing you are actually trying to find out.