Finding the leak in your funnel before spending on more traffic

5 days ago
3 min read

Updated: 3 days ago

Introduction

When sales are below where they should be, the reflex is to generate more enquiries. It is the most expensive possible response and usually the least effective, because pouring more into a container that leaks produces more leakage rather than more customers.

The alternative takes an afternoon and costs nothing. Count how many people reach each step, work out where the largest proportion disappears, and look at that step specifically. Owners are wrong about the location surprisingly often — the step everybody blames is rarely the one losing the most — and being wrong about it is what makes the marketing budget feel like it disappears.

1. Finding the leak in your funnel means comparing rates, not totals

The distinction that matters.

Losing four hundred people at the top and forty in the middle looks like a top-of-funnel problem. If the top converts at eighty per cent and the middle at twenty, the middle is the leak. Always compare proportions.

2. Count the last ninety days from your own records

Enough data, recent enough to be true.

Enquiries received, quotes issued, orders won, from your diary and your invoices. Three numbers and a period. Precision matters less than consistency of definition.

3. Find the biggest proportional drop

The one number the exercise produces.

Whichever stage loses the largest share is where the attention goes. Everything else waits, because fixing a second-order leak while the largest one runs is how effort disappears without a visible result.

4. Check the drop against what is normal for you

Context before conclusions.

Some drops are structural. Losing eighty per cent between enquiry and quote is alarming in one business and entirely normal in another where most enquiries are outside the service area. Compare against your own history rather than a benchmark.

5. Look at time, not only at count

The second kind of leak.

A stage where nobody is lost but everybody waits eleven days is leaking too, because delay is where buyers go elsewhere. Median elapsed time per stage frequently reveals more than the conversion rate.

6. Read the ones that stopped

Where the reason comes from.

Twenty enquiries that did not convert, read in full: the original message, what you sent, what happened next. The pattern is usually obvious after fifteen and it is never what the numbers alone suggested.

7. Suspect your own response before their intent

The uncomfortable first hypothesis.

Slow replies, quotes that took a week, follow-ups that never happened, an enquiry routed to someone on holiday. A large share of apparent buyer disinterest is your own process, and it is the cheapest thing to fix.

8. Fix one thing and measure it

Sequencing.

Change a single step, keep everything else constant, and compare the same rate for the next ninety days. Changing four things at once means you learn nothing about which of them worked.

9. Expect the leak to move

Not a permanent finding.

Fix the enquiry-to-quote rate and the bottleneck relocates to quote-to-order. This is progress rather than failure, and it means the exercise is a quarterly habit rather than a one-off project.

Be careful about a leak that is really a qualification problem. If most enquiries were never going to buy, the middle-stage conversion rate is low for a good reason, and the fix is at the top rather than in the middle.

Conclusion

Compare conversion rates between stages rather than raw numbers, because the largest loss is rarely where it appears.

Count the last ninety days from records you already keep, find the biggest proportional drop and work only on that, judge it against your own history rather than a published benchmark, look at elapsed time as a second form of leak, read twenty non-converting enquiries in full to find the reason, suspect your own response speed before assuming buyers were not serious, change one thing at a time and measure it, and expect to repeat the exercise once the bottleneck moves.