There is a particular feeling that comes from reading a list of sixty creative growth ideas and recognising that your competitors read the same list this morning. The ideas are good. Some of them worked spectacularly for the companies that first tried them. And the moment they were written down and circulated, they stopped being an edge and became a floor, the set of things every reasonably informed team now does, producing reasonably similar results. Growth tactics have always decayed. What has changed is the speed, because the distribution of tactics is itself now instant, and a move that took years to spread a decade ago spreads in a week.

The instinctive response is to generate more ideas, on the theory that the bottleneck is supply. It is not. Most teams have more growth ideas than they could ever run, and the lists make sure of it. What they lack is taste: the ability to look at forty possible moves and pick the three that suit this product, this audience and this team’s particular strengths, and then to execute those three with a care that a competitor copying the tactic will not bother with. When the tactics are shared, the tactic is not the advantage. The selection and the execution are, and both of those are craft.

Why identical playbooks produce identical results

A tactic that everyone runs is, from the customer’s side, noise. The tenth referral programme, the twentieth waitlist with a share-to-skip mechanic, the hundredth founder posting a thread about lessons learned: each of these once stood out and now blends in, because the customer has seen it and knows what it is for. The arbitrage that made the tactic work was the gap between what customers expected and what the tactic did, and the gap closes the moment enough companies have done it. Running the tactic anyway is not wrong. It just produces the average result, which is what everyone else gets.

The decay is also faster now than it used to be because the write-ups have become the product. A tactic that works is written about within weeks, summarised within days of that, and circulated to every founder who follows the same handful of newsletters. The half-life of an edge is measured in months at most, and anyone who plans a year around a tactic they read about last quarter is planning around something that has already been priced in.

This is why more ideas rarely help. Adding ideas to a list that is already too long does nothing about the real problem, which is that the team is choosing from the list on the basis of what worked elsewhere rather than what would work here. The lists are useful as raw material. They are useless as a plan, and teams that treat them as a plan end up running a slightly random sample of other companies’ strategies and wondering why the results look like a slightly random sample of other companies’ results.

A filter that does the work

The craft is in the filter, and a good filter asks three questions of every candidate move. Is it reversible: if this fails, can we stop it cleanly, or does it leave a mess in the product, the brand or the customer relationship. Does it have story value: would a customer who encountered it mention it to someone, because a move people talk about distributes itself and a move nobody mentions has to be paid for every time. And does it compound: does running it this month make next month’s version easier or more effective, or is it a one-off that has to be rebuilt from nothing each time.

When the tactics are shared, the tactic is not the advantage. The selection and the execution are, and both of those are craft.

Most items on any growth list fail at least one of these, and a surprising number fail all three. The ones that pass are usually fewer than five, and they are usually the ones that a competitor with the same list would not pick, because they depend on something specific to this company: an unusual audience, an odd product capability, a founder with a particular voice. That specificity is the point. A move that could only be run here is a move that cannot be copied, and uncopyable moves are the only ones that hold their value.

The ninety-day test

Taste without execution is opinion, so the second half of the craft is a small, disciplined test. Pick two or three moves that pass the filter, commit to running them properly for ninety days, and refuse everything else for the duration. Properly means with the care that the tactic’s originators applied and its copiers skipped: the referral programme where the reward is actually something people want, the content that is actually good rather than merely present, the partnership where both sides actually benefit. Ninety days is long enough to see whether something compounds and short enough that a wrong bet is cheap.

Care is the variable that the lists cannot transmit. Two companies can run the same referral programme and get entirely different results, because one of them thought about what the referred person actually experiences in the first minute and the other copied the mechanic and moved on. The tactic was identical. The craft was not, and the craft is where the result came from.

The test does not require a growth team. It requires one person who owns it, a weekly look at whether the move is doing what the filter predicted, and the willingness to stop at the end of the period if it is not. Most companies never run this test because they never stop running everything else long enough to see what any one thing is doing. The refusal is the hard part. Doing three things well is uncomfortable when the list has sixty, and the discomfort is exactly what a competitor copying the list does not feel.

What to kill

The final act of taste is subtraction. Every team is running tactics it keeps running because a competitor does, or because it worked once, or because stopping would feel like admitting the earlier decision was wrong. These are the moves that fail the filter in retrospect: not reversible because they have become habits, no story value because everyone has one, not compounding because they were never designed to. Killing them frees the attention that the chosen moves need, and attention, not ideas, is the resource that growth actually runs on.

Subtraction is also where the taste shows most clearly, because anyone can add a tactic and only a team with a view of its own can decline one. The list of things a company refuses to do is as revealing as the list of things it does, and it is usually the better predictor of whether the chosen moves will get the care they need.

The playbook everyone runs will keep being published, and it will keep being useful as a source. But the companies that grow in a market where everyone has the same playbook are the ones that treat distribution as a craft rather than a checklist: choose few, choose specific, execute with care, and stop doing the things that only exist because someone else did them first. That is not a tactic. It is the thing that makes tactics work.