Funnel Drop Off:How to Fix the Biggest One Fast
Funnel drop off: name the cause first, friction, doubt or mismatch, then change one thing. Most people simplify the wrong page. Here is how to fix it.
To fix a funnel drop off, first name the cause: friction, doubt or mismatch. Each has a different fix. Take the largest loss that is also cheap to address, change one thing, and watch that step. Mismatch is fixed upstream, never on the step itself.
How to fix the biggest funnel drop off without making it worse
You found the leak. Good. Now comes the part where most people waste a month.
They assume the problem is friction. They simplify the page. They cut a field, shorten a paragraph, move a button. The numbers don't move, and they conclude funnels are overrated.
The page wasn't the problem. Nobody believed the claim.
People assume friction because friction is actionable. You can see it, you can touch it, you can fix it by Thursday. That's exactly why it's the wrong default. Before you change anything, work out which of three problems you actually have.
Friction, doubt or mismatch: pick the right one
Every funnel drop off comes from one of three causes. The fixes don't overlap.
| Cause | What it looks like | What fixes it | What does not |
|---|---|---|---|
| Friction | Too hard, too long, too slow. Unclear what to do next | Remove steps. Ask for less | More reassurance copy |
| Doubt | They understand and are not yet convinced. Missing evidence, price anxiety, an unanswered risk | Add the right information at the exact point of hesitation | A shorter form |
| Mismatch, the funnel drop off no page edit fixes | Wrong people, wrong expectation, or the last step promised something this one does not deliver | Fix the message or the targeting upstream | Anything you do on the step itself |
Read the last row twice. Mismatch is the only cause where working on the step is guaranteed to fail. If the previous step promised something this one can't deliver, you can redesign this page forever. People arrive expecting the wrong thing, and no layout changes that.
A sharp drop right after someone commits is the classic sign. After signup. After the trial starts. After the first purchase. That's rarely a usability issue. It's your marketing's promise meeting your product's reality.
Free guide · 20 pagesStop guessing where it leaks.The Funnel Playbook: where you are losing people and how to find out.Get the free guide →Pick the funnel drop off to fix: biggest loss, cheapest fix
Before diagnosing, you need to know which step to diagnose. If you haven't ranked yours yet, how to find where you are losing customers shows how: multiply drop-off by volume, rank by people lost.
Then add a second filter. Effort.
The correct first project is usually the largest loss that is also cheap to address. Not the largest loss overall. A confusing sentence is an evening. Rebuilding onboarding is three weekends. Do the sentence first.
That's not laziness. It's sequencing. The sentence might close most of the gap. If it doesn't, you've lost an evening and you know more than you did. If you go straight for the three weekends and it doesn't work, you've lost your month and your appetite.
Also check the segments one more time. If the problem is one audience rather than the step, you don't have a page issue. You have a targeting issue, and the fix may be to stop sending those people. See how to segment a funnel.
Get the why from people, not percentages
The funnel tells you where. It almost never tells you why.
So spend an hour on people. Watch a handful of them go through the failing step. Or ask a few who abandoned it what stopped them. That generates better hypotheses than a week of staring at a chart.
Use this to place your problem in the table above:
- If they sigh, scroll back up and hunt for the next button, it's friction.
- If they read carefully, hover over the price and then leave, it's doubt.
- If they say "oh, I thought this was something else", it's mismatch.
Then write the hypothesis down before you change a thing. One sentence. "I believe people leave here because they can't see what happens after they pay." If you can't write it, you don't have a hypothesis. You have a hunch with a deadline.
Change one thing
This rule is boring and it's the one people break most.
Change several things at the same step and the result can't be read. If the rate goes up, you won't know which change did it. If it goes down, you won't know which to undo. So you keep the whole bundle, including the part that hurt, and you build the next idea on top of it.
One change. One step. One written hypothesis. Then leave it alone.
Now the part that stings. You probably don't have the traffic to detect a small improvement in a useful time. A tweak that moves a step by a hair will disappear in ordinary week-to-week wobble.
That leaves two honest options. Make the change big enough that the effect is obvious. Or accept you're making a reasoned judgment, not running an experiment. Both are fine. Pretending a sample of forty proved something is not. It's how people end up defending changes that did nothing. How to read the result properly is in how to know if a funnel change worked.
A worked example, as an archetype
Say you sell a templated service to freelancers. Your biggest loss sits between "booked a call" and "paid". Plenty of people book, few pay.
Your instinct says friction: shorten the proposal. But you watch a few people read it, and they stop at the price. That's doubt. They understand the offer and aren't sure it's worth it. The fix isn't a shorter proposal. It's the missing evidence, placed where they hesitate: what the work includes, what happens if it doesn't suit them, what the next step is after paying.
Now picture a different result. You ask a few people who booked and didn't pay. They say, "I thought it was a course." That's mismatch. Your page promised a thing the call then contradicted. No edit to the proposal helps. You go upstream and fix what the earlier page says.
Same drop-off. Two different causes. Two fixes that would each have failed in the other's place.
That's why diagnosing comes before touching anything. It's cheap, it takes an hour, and it stops you spending a weekend fixing the wrong thing.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →After the fix, look for where the constraint moved
A funnel is a chain. Fix the weakest link and the weakest link moves.
That's success, not a setback. It also means this is a loop, not a project. A few weeks later, look at the step again. Some gains decay. And occasionally a fix just shifts the loss to the next step, because the people you saved were never going to convert.
That last result is worth knowing about. It usually points to a mismatch your fix didn't touch.
Then rank again and pick the next largest cheap loss. Monthly is a sensible rhythm. Don't try to run this daily. You'll mistake noise for progress.
What to write down before you start
Four lines. That's all.
- The step and its starting rate. To the decimal, with the window.
- The cause you believe. Friction, doubt or mismatch, and the evidence from real people.
- The one change. Not a list.
- Whether you can read the result. Honestly. Do you have the volume, or is this a judgment call?
Writing the fourth line is the one that stings, and it is the one that keeps you honest. Half of the value of this loop is that you stop fooling yourself. A funnel that can't detect a small gain should make you bolder in what you try, not more certain in what you claim.
Free guide · 20 pagesStop guessing where it leaks.The Funnel Playbook: where you are losing people and how to find out.Get the free guide →When the fix is not in the funnel at all
Two situations, both common, both annoying.
First: you fixed the step and nothing moved because the offer is the problem. If every step is mediocre and none is catastrophic, suspect the proposition. Percentage gains on a number that was never going to work are motion, not progress. How to tell a funnel problem from a traffic problem covers the split.
Second: the funnel is fine and the totals are small. Then you have a volume problem. Go and get more traffic, now that you know it won't leak.
Fix the leak first. Then carry more water. Every funnel drop off you close makes the next batch of visitors worth more.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →Frequently asked questions
What is the most common reason people drop off a funnel?
There are three causes: friction, doubt and mismatch. People default to friction because it is the easiest to act on, but doubt and mismatch are often the real cause. Work out which one you have before changing a page.
Should I fix the worst percentage or the biggest loss?
The biggest loss, weighted by effort. Ten per cent of a small number is a smaller prize than a modest gain on a large one. Pick the largest loss that is also cheap to address, then move to the next.
Can I change several things at once to save time?
You can, but you won't know which change worked, and you may keep the one that hurt. Change one thing at the step, write the hypothesis first, and give it time before judging.
How do I know a funnel drop off is a mismatch?
Look for a sharp drop right after commitment, and for people who say they expected something else. If the previous step promised what this one doesn't deliver, fix the message or the targeting upstream.
The full method, with the checklists, is in The Funnel Playbook, a free 20-page playbook.
How trapped are you, really?
Take the Corporate Suffocation Index, the free solopreneur readiness assessment.