The growth constraint is rarely where the symptom shows up

Almost every growth conversation starts in the same place: something that used to work has stopped working. Cost per acquisition drifted up. The pipeline is thinner. A channel that carried the business for two years now needs twice the budget for the same result. The instinct is to treat that channel as the problem, because that is where the number moved.

It is usually not the problem. It is the measurement surface. Acquisition sits at the front of the business, so it is the first place any weakness further back becomes visible as a number. A weaker offer, a slower sales follow-up, a fulfilment delay that produces worse reviews, a support queue that stopped retaining customers: all of them show up first as acquisition getting more expensive, because acquisition is what you are paying for and everything else is what determines whether that payment converts.

Symptom, cause, and the cost of confusing them

Spending against a symptom is expensive in a specific way. It works, briefly. More budget, a new creative round, a different agency: each of these produces a short lift, because you can always buy your way past a conversion problem for a while. Then the underlying constraint reasserts itself and you are at the same efficiency with a higher fixed cost. Several rounds of that and the business has an expensive marketing function attached to an unresolved problem somewhere else.

The useful question is not "why did this channel get worse" but "what has to be true for a customer acquired here to be profitable, and which of those things stopped being true."

Walking the chain backwards

The practical method is to start at the money and work backwards through every step a customer passes, asking at each one whether the step is doing what it was doing when the economics worked.

• Retention and repeat: has the second purchase rate or renewal rate moved? If lifetime value fell, acquisition did not get more expensive in absolute terms, it got more expensive relative to what a customer is now worth.
• Delivery and support: are you fulfilling as fast and as consistently as before? Delivery quality shows up in reviews, refunds and repeat rate months before it shows up in a dashboard anyone is watching.
• Sales conversion: is the same proportion of qualified interest closing? A drop here is often process drift, not demand: slower first response, weaker qualification, a step someone quietly stopped doing.
• Offer and pricing: has a competitor changed what the market considers standard? An offer is only strong relative to alternatives, and alternatives move without warning.
• Targeting and message: only now is it reasonable to ask whether the audience or the creative is genuinely tired.

In practice the chain usually breaks somewhere in the middle, and the break is boring. Response time to inbound enquiries went from an hour to a day because the person who owned it took on something else. A supplier lead time stretched by a week and nobody updated the delivery promise on the site. These are not strategic failures. They are unowned steps, and they are cheap to fix once you know which one moved.

One constraint at a time

Once you find a binding constraint, the temptation is to fix everything you noticed on the way. Resist that. A business under pressure has limited attention, and running five improvement projects at once means none of them finish and you cannot attribute the result. Fix the binding constraint, let the system run long enough to read a clean signal, then look again. The next constraint will have moved somewhere else, which is exactly what should happen.