One Constraint, Not a List: How to Find the Single Thing Holding Your Revenue Back

It is 11pm and you have twelve options. Hire a salesperson. Fix the website. Raise prices. Run ads. Rebuild onboarding. Start a podcast. Every one of them is defensible. Every one of them has a case study behind it. And you know — the way you know things at 11pm — that eleven of them are going to cost you a quarter.

The problem is not that you lack a plan. You have too much plan. What you do not have is permission to ignore eleven things, and nobody sells that, because a list of twelve looks like more value than a sentence naming one.

So this post is the sentence. Here is how to find the one constraint that is actually binding on your revenue, using arithmetic you already have, in about twenty minutes.

What "binding" actually means

A constraint is binding when relieving it increases output and relieving anything else does not. That is a stricter test than "this is bad" — most businesses have five things that are bad and one thing that is binding, and the difference between those two facts is a quarter of wasted effort.

Two consequences fall out of the definition, and both of them are counterintuitive enough to be worth stating plainly.

Improving a non-binding stage produces nothing. If your close rate is 12% and you have more qualified leads than your team can call, doubling lead volume adds zero revenue. It adds unreturned calls. The extra leads queue behind the same constraint and then age out.

Only one thing is binding at a time. When you relieve it, the constraint moves — usually somewhere annoying, like delivery. That is not a failure of the diagnosis. That is the diagnosis working. A business is a sequence of constraints you resolve one at a time, and the skill is knowing which one you are standing in front of right now.

Five places it can live

Every revenue constraint we have found in an operating business sits in one of five disciplines. This is not a taxonomy for its own sake — it is the shortest list that is genuinely exhaustive, which matters because the point of the exercise is elimination.

Discipline	The constraint sounds like	The number that reveals it
Demand	"Not enough of the right people know we exist"	Qualified conversations started per month
Conversion	"They talk to us and then go quiet"	Close rate, and where in the stages it dies
Pricing	"We are busy and it does not show up in the bank"	Average order value against fully-loaded delivery cost
Delivery	"We could sell more than we could deliver"	Utilisation, and lead time to add capacity
Retention	"We sell hard and stay flat"	Monthly churn against monthly new revenue

Notice that "marketing is not working" is not on the list. It is not a constraint; it is a department. The constraint underneath it is either demand (nobody hears you), conversion (they hear you and do not act), or pricing (they act and it does not pay). Which one it is changes the fix completely, and the department name hides that.

The twenty-minute pass

Work these in order. Stop at the first one that is true, because the first true answer is the binding one — that is what "binding" means.

1. Is new revenue outrunning churn?

Compare monthly new revenue against monthly lost revenue. If lost is more than about half of new, stop here: retention is binding and nothing upstream matters yet. This is the most commonly missed constraint in the whole list, because churn arrives as a series of small individually-explainable events and new revenue arrives as a celebration.

Illustrative arithmetic, so the shape is clear rather than the specific numbers: you add $18,000 of new monthly revenue and lose $11,000. You are running a demand engine at full tilt to grow $7,000. Fixing the leak is worth more than doubling the engine, and it is cheaper.

2. Could you deliver what you would win?

Take your current monthly revenue and imagine it doubling next month. Now name what breaks first. If you can name it in under five seconds — a person, a step, a queue — delivery is binding, and your sales team is very likely already throttling itself to protect the fulfilment your ops team cannot expand. Nobody will admit this out loud. It shows up as slow follow-up on good leads.

The test that makes it concrete: what is the lead time to add one unit of capacity? If it is longer than your sales cycle, you are capacity-constrained no matter what your pipeline says.

3. Does the work pay for itself?

Take one typical engagement. Add every hour anyone spends on it — including the sales call, the scoping, the revisions, and the two Slack threads at 9pm. Multiply by a loaded rate. Compare to the invoice.

If the gap is thin, pricing is binding, and this is the only constraint on the list where the fix is free. You do not need more leads to charge more. You need to stop discounting the part of the work that is actually the expertise.

4. Do the conversations close?

Look at the last twenty qualified opportunities and count how many closed. Then find the stage where they died — first call, proposal, silence after the price, the second decision-maker who appeared late.

If they die at one identifiable stage, conversion is binding, and the fix is at that stage specifically. This is the constraint most often mistaken for demand, because "we need more leads" and "we lose most of the leads we get" produce the same feeling and completely different actions.

5. Are there enough of the right conversations?

If none of the above tripped, and you close a reasonable share of what you talk to, and you could deliver more, and the work is profitable, and customers stay — then yes: demand is binding, and now you get to spend money on marketing with an actual expectation that it will show up in revenue.

The order is the point. Demand is last on this list and first in almost everybody's budget.

Write down what you are not doing

Here is the part that people skip, and it is the part that makes the exercise stick.

Once you have named the constraint, write the list of things you are not going to do this quarter. Explicitly. By name. "We are not rebuilding the website. We are not hiring a second rep. We are not starting the podcast."

A diagnosis without a stop-list is not a decision — it is a preference, and preferences lose to whoever walks into your office next with a good idea. The stop-list is what converts a diagnosis into ninety days of focus, and it is the thing you will be tempted to leave out because it feels negative. It is not negative. It is the whole point of having diagnosed anything.

Or have it done in a few minutes

The reason we can write this out for free is that the framework was never the moat. Doing it honestly on your own business is hard for a reason that has nothing to do with knowledge: you already have a favourite answer, and it is usually the one that is most fun to work on.

So we built the version that does not have a favourite. Mary reads your business and names one constraint — plus what not to do. She is free to try, there is no card, and you do not have to sign up to find out whether the answer is any good.

Two things to know before you click, because we would rather you heard them from us.

The first: Mary is free and the six specialists are not. If the constraint lands in a specific discipline — sales, marketing, operations, retention, technology, people — the specialist who owns that discipline is $99/mo for one or $499/mo for all seven. You can see exactly what each one produces, per mode, with the price attached, on the capabilities page. Nothing there is a demo.

The second: we have sold one playbook so far. That number is printed on our own pricing page, and it will stay there and change as it changes. We would rather tell you that than borrow credibility we have not earned — partly because it is the right thing to do, and partly because in a category this full of invented social proof, a number that low is the only claim nobody can copy.

If you are an engineer rather than an operator, the same coaches are callable programmatically and billed per answer, which is documented here. Same diagnosis, no interface.

Either way: one constraint. Not twelve. And an explicit list of what to ignore until it is fixed.