Traffic has to be re-bought every month.
Conversion applies to every visitor.
For founders who have traffic and are not sure where they lose it. This playbook is where you are actually losing people, and how to find out instead of guessing.
// 30 PAGES · FIVE PARTS · THREE CHECKLISTS · FREE
Seven things to know about your funnel.
Each is a mechanism you can plan around, and the guide is plain about the limit: the funnel is only one part of the picture.
Every improvement compounds against all future acquisition
This is the structural point and it is worth sitting with. If you improve one step, the gain applies to every visitor who ever arrives afterwards, through every channel, including the ones you have not started yet.
It localises the problem, which ends the unproductive argument
"Growth is slow" is not a diagnosis and cannot be acted on, so it generates a meeting in which everyone proposes their own department's solution.
It tells you what to build
Product roadmaps at early-stage companies are usually assembled from customer requests, competitor features and internal conviction.
It exposes the gap between what you promise and what you deliver
A sharp drop immediately after someone commits — after signup, after the trial starts, after the first purchase — is usually not a usability problem.
It is the diagnosis layer underneath every other marketing decision
Attribution tells you which channel produced a customer. The funnel tells you what happens to the people that channel sends, and the two together are considerably more powerful than either alone — because a channel with poor conversion may be sending the wrong people rather than being a bad channel, and you cannot distinguish those without both views.
It converts a slow feedback loop into a fast one
Waiting to see whether a change affected revenue can take a full sales cycle. Watching whether it affected the step it was meant to affect is often possible within days.
And now the honest limit
A funnel measures how well you convert the interest you already have. It cannot manufacture interest, and it cannot rescue a proposition people do not want.
Five parts, and a last page.
Parts one to five are the thinking and the doing, worth your time even if you never touch a piece of software. The last pages show how to keep the funnel counting correctly.
Why the funnel is the cheapest lever
Define it before you measure it
Instrument it simply
Read it and fix it
The checklists and the mistakes
Do it all in one place
Nine mistakes, all of them avoidable.
None is about traffic or luck. They are failures of definition and measurement, which is good news, because both can be fixed in advance.
Buying more traffic to fix a conversion problem
The intuitive response and usually the most expensive one — it pours more volume into a process that loses most of what enters, at a cost that only rises.
Counting steps independently instead of as a cohort
It produces a chart shaped like a funnel with percentages that mean nothing, and it is the single most common invalidating defect in funnel dashboards.
Building the funnel from the org chart
Three handed-off funnels mean nobody owns the seams, and the seams are usually the worst-performing steps precisely because of it.
Optimising a blended rate
The average hides both the segment worth doubling down on and the one dragging the number down. Segment before you touch anything.
Chasing the worst percentage
Ten percent of a small number is a smaller prize than a modest improvement on a large one. Rank by absolute loss, weighted by effort.
Assuming every drop-off is friction
Simplifying a page will not fix doubt, and nothing done at the step will fix a mismatch created upstream.
Changing several things at once
The result is uninterpretable and you will keep the wrong one, then build on it.
Declaring victory on a sample too small to read
Most early funnels cannot detect small improvements. Say "this is a reasoned judgment" rather than pretending it was an experiment.
Optimising a funnel when the offer is wrong
A uniformly mediocre funnel with no catastrophic step is usually a positioning problem, and percentage gains on a number that cannot work are motion, not progress.
Thirty pages.
Yours in one click.
Five parts, three checklists and the nine mistakes. One email with the link, no drip sequence, and we do not sell the list.