Why the two get confused
Because they present identically. Revenue is below plan, the pipeline looks thin, and the conversion rate in the CRM is disappointing. Every one of those symptoms is consistent with both diagnoses, and the CRM cannot distinguish between them — it reports on the opportunities it received, and the entire question is about the population it did not.
So the default diagnosis wins by convention rather than by evidence, and the default diagnosis is a sales problem. It is the one with an obvious intervention, an owner, and a vocabulary everyone already shares. The interventions that follow — more reps, more training, a new comp plan, more pipeline — are all expensive, all slow to evaluate, and all useless against a latency problem.
The test, in detail
It needs one quarter of history and no new instrumentation. The point is to compare like with like, so the work is mostly in defining the populations honestly.
- 01Pull every inbound enquiry from last quarter, from every channel. Every channel is doing real work in that sentence: the ones you forget are usually the ones leaking.
- 02For each, timestamp the first GENUINE human response. Not the autoresponder, not the system-generated acknowledgement, not the SDR sequence step one. If a person did not engage, it does not count.
- 03Split the population at one hour. The threshold is arbitrary and that is fine; use whatever is meaningful in your market, and use the same threshold for both groups.
- 04Compare close rate between the two groups, using your own outcomes rather than any benchmark.
- 05Separately, count the enquiries that received exactly one contact and no second. This number is usually the surprise, and it is a different leak from latency.
Now read the result. A large gap with a large slow group is leakage. A small gap is a sales problem — your team is reaching people and not closing them, and that is a commercial issue. A large gap with a tiny slow group is a leakage problem that is already mostly solved and not where your money should go next.
What each diagnosis actually costs to act on
This is where the distinction pays for itself. A sales intervention is a hiring and enablement programme: months to staff, a quarter or more before the signal is readable, and the cost is ongoing headcount. A leakage intervention is a build: weeks, a signal visible in the first week for latency and within one sales cycle for conversion, and the cost is largely one-time. Implementation at this scale typically runs $15,000 to $75,000 across all phases.
Getting the diagnosis wrong in the expensive direction — treating leakage as a sales problem — means adding headcount to a funnel that will lose a similar proportion of whatever the new reps generate. The new demand leaks through the same gaps. That is the specific failure mode worth avoiding, because it is self-concealing: revenue does go up, so the intervention looks like it worked, and the leak is now proportionally larger in absolute dollars.
Why additional marketing spend is the worse mistake
Same logic, sharper. If 65% of your enquiries are not reached inside an hour, then buying more enquiries buys a population of which roughly 65% will be handled the same way. You are paying acquisition cost on demand you have a structural reason to believe you will not convert.
This is the argument for sequencing rather than for choosing. Close the gap, then buy the demand, and the same media budget produces more revenue than it did before without the budget changing. Doing it the other way round works too, eventually, and it costs the difference.
The cases where it genuinely is a sales problem
- Conversion is flat across response speeds and your slow population is small. The operational path is clean and the problem is downstream.
- Deals are reaching late stage and dying there. Late-stage loss is a commercial and competitive problem, not a latency one.
- Win rates differ sharply between reps on comparable territories. That is a people and enablement signal and no amount of routing fixes it.
- Your average deal value is falling while volume holds. That is pricing and positioning.
And the cases where it is both
Frequently, and the sequencing still matters. Where both are true, the leakage side is usually the one to do first for a reason that has nothing to do with which is larger: it is measurable faster and it is cheaper to reverse if you are wrong. A build that closed a gap can be evaluated against a baseline within a quarter. A hiring programme cannot be evaluated inside a year, and it cannot be unwound cheaply.
The broader reason to run the test at all is that it converts an argument into an arithmetic problem. Whether you have a sales problem is a question people hold opinions about; whether your slow-response population converts worse than your fast one is a question your own systems have already answered.