A leaky funnel loses most people at one stage, not everywhere

2 days ago
3 min read

Updated: 1 day ago

Introduction

The phrase describes a real and common situation: enquiries arrive, revenue does not follow, and nobody can say where the people went. The instinctive response is to fix everything at once — new website, new follow-up, new advertising — which is expensive and makes the cause impossible to identify.

Leaks are almost never even. In practice one stage loses far more than the rest, and the arithmetic that finds it takes about an hour with your own records. Everything after that is a decision about whether the fix is worth making.

What follows assumes you can count enquiries, quotes and jobs. If you cannot, that is the first leak.

1. A leaky funnel usually has one dominant hole

The organising insight.

Losses concentrate. One stage will be losing multiples of what the others lose, and it is rarely the stage people assume. Find it before changing anything. Guessing costs more than counting.

2. Count what enters and leaves each stage

The method.

Two numbers per stage, for the same period. The ratio between them is that stage's loss rate. Nothing more sophisticated is needed. A spreadsheet column each will do.

3. Rank by people lost, not by percentage

The prioritisation rule.

A stage losing 30% of 200 people matters more than one losing 70% of 20. Multiply the rate by the volume reaching it. This ordering is arithmetic, not judgement. The largest number is where you start.

4. Check elapsed time as well as counts

The hidden leak.

A stage that loses nobody but takes eleven days is leaking into the competitor who answered the same day. Slow is a form of loss that ratios do not show. Measure the gaps. Use the median days between stages.

5. The commonest leak is between enquiry and reply

The prediction worth testing.

In most small businesses more is lost to unanswered and slowly answered enquiries than to anything on the website. Test your own response time before assuming otherwise. It is also the cheapest thing to fix. Send yourself an enquiry and time the reply.

6. The second commonest is after the quote

The other usual suspect.

Prices issued and never followed up account for a large share of lost revenue in service businesses. Count how many quotes received a second contact. The answer is often close to none. One follow-up is usually worth more than any new channel.

7. Distinguish a leak from bad qualification

The diagnostic caution.

If people who were never going to buy are entering the funnel, the loss is at the top and it is not really a leak. Look at where enquiries come from before redesigning the middle.

8. Fix one stage and measure

The discipline.

Changing three things at once means you learn nothing about any of them. One change, one measurement period, then the next. The list will still be there.

9. Expect a new dominant leak afterwards

The realistic outcome.

Fix the biggest and another becomes the biggest — that is success, not failure. The exercise repeats. Each round is faster than the last.

Be careful about accepting a leak that is really a capacity limit. If enquiries stop advancing because nobody has time to serve them, no process change helps, and the honest answer is fewer enquiries or more people.

Conclusion

Find the one stage doing most of the damage before changing anything.

Count what enters and leaves each stage over the same period, rank the stages by number of people lost rather than by percentage, measure elapsed time as well because slow stages leak invisibly, check your response time and your quote follow-up first since those are usually the largest and cheapest, separate genuine leaks from poor qualification at the top, fix one stage at a time, and expect a new largest leak to appear once the first is closed.