The last click problem in a small business and what it costs

4 days ago
3 min read

Updated: 3 days ago

Introduction

Most reporting gives all the credit for a sale to whatever the buyer did immediately before contacting you. Usually that is a search for your business name, or a direct visit. Both look outstanding in any report, and both are simply the final step of a journey that started somewhere else entirely.

The consequence is predictable and expensive. The channel that made somebody aware of you gets no credit, so it looks unproductive, so it gets cut. Three months later the searches for your name decline and nobody connects the two events, because the report showed the branded search performing brilliantly right up until the demand feeding it stopped.

This is worth understanding even in a business with no analytics at all, because the same error happens in conversation.

1. The last click problem in a small business is mostly branded search

The characteristic pattern.

People who already know your name search for it, and that search converts extremely well. It is a confirmation step, not a discovery one. Reading it as acquisition is the whole error.

2. Direct traffic has the same issue

The second version.

Somebody typing your address in already knew it. Direct is the record of a decision made elsewhere. It deserves credit for nothing except being easy to reach.

3. Recommendation hides behind both

The invisible cause.

A neighbour's suggestion produces a branded search, and the search takes the credit. This is the single most under-measured channel in small business. Asking the caller directly is the only way to see it.

4. Awareness channels look worst under last click

The systematic bias.

Anything working at the top of the funnel is by definition rarely the last step. The bias against it is structural rather than a data problem. No configuration fixes it.

5. Watch the volume of branded search itself

The leading indicator.

If people searching your name are falling, something upstream has stopped working, months before revenue shows it. It is one of the few genuinely useful early signals. It costs nothing to check.

6. Ask rather than infer

The practical remedy.

One question on the first call outperforms every model available to a business of this size. It also catches the offline half. Consistency matters more than precision here.

7. Judge awareness spend over quarters

The timescale.

A channel that creates demand cannot be assessed on a month of last-click data, and assessing it that way guarantees it fails. Give it a quarter and compare against the same quarter last year. Anything shorter is noise.

8. Keep first touch where you can capture it

The record.

If the source was recorded at enquiry, keep it attached to the customer through to the invoice. Then revenue can be traced back to origin rather than to the final click. This one field makes most other analysis possible.

9. Be wary of cutting the quiet channel

The decision rule.

The channel with no visible attribution is the one most likely to be feeding everything else. Pause it rather than cutting it, and watch for two months. Reinstating something you cut a year ago is much harder.

Be careful about the reverse error too. Not everything unmeasured is secretly working, and a channel that has produced nothing visible for a year while nobody mentions it in conversation is probably producing nothing at all.

Conclusion

Treat branded search and direct traffic as the end of a journey, not the start of one.

Recognise that recommendation hides inside both, that awareness channels are structurally disadvantaged by last-click reporting, and that the volume of searches for your own name is a useful early warning. Ask every caller how they heard about you, judge demand-creating activity over quarters rather than months, keep the recorded source attached from enquiry through to invoice, and pause a quiet channel rather than cutting it.