How buyers actually first hear about you, checked against evidence

5 days ago
3 min read

Introduction

Every owner has a working theory about where customers come from, and the theory is generally assembled from the most memorable recent cases rather than from any count. The customer who mentioned the advertisement is remembered; the eleven who came from a recommendation and never said so are not.

The gap between the theory and the evidence is where marketing budgets go astray. Money follows the belief, the belief follows salience, and salience follows whatever was mentioned most recently in conversation. Correcting it requires nothing more sophisticated than asking and writing the answer down, which almost no small business does consistently.

1. How buyers actually first hear about you is answered by asking, not by analytics

The tool most people reach for is the wrong one.

Analytics records the last click before an enquiry, which is frequently a search for your name by somebody who was told about you a month ago. It reports the final step and calls it the source. This is not a fault in the software; it simply cannot observe a conversation in a car park.

2. Ask every enquirer, in the first conversation

The whole method.

"Can I ask how you heard about us?" It takes four seconds, it is asked naturally at the start of a call, and after three months you have better data than most businesses ever collect. The discipline is asking every time rather than when you remember, because selective asking produces a biased sample.

3. Record the answer verbatim

Not into a category.

"My sister used you for her kitchen" and "I saw your van outside the school" are different and both get lost if the form only offers word of mouth. Categorise later, after you have seen the real range.

4. Expect a chain rather than a single source

The most common finding.

Heard about you from a neighbour, then searched your name, then looked at your reviews, then called. Every step was necessary. Attribution to one of them misrepresents how the decision actually formed. Recording the first and last step of the chain, where the enquirer can tell you both, is more honest than picking one.

5. Distinguish the trigger from the source

Two different questions.

Where they heard of you, and what made them act now. The second is frequently a boiler failing or a contract ending, and knowing it tells you when to be visible rather than where.

6. Watch for the channel that gets no credit

Almost always word of mouth.

It generates the enquiry and rarely gets recorded, because the last click was a search. The result is that the channel producing most of the work receives none of the investment or attention.

7. Compare answers against your spending

The point of the exercise.

Put your sources by volume next to your spend by channel. The mismatch is usually stark, and it is the clearest argument for reallocating that you will ever produce.

8. Look at conversion by source, not just volume

The second dimension.

A source producing a third of enquiries that convert at five per cent is worth less than one producing a tenth that convert at sixty. Volume alone routinely sends money to the wrong place.

9. Review it quarterly and act on it

Making it change something.

Twenty minutes with the log, comparing sources by volume and conversion against cost. Data collected and never read is the most common outcome, and it is worse than not collecting it.

Be careful about acting on a small sample. Thirty enquiries will not distinguish a channel producing eight per cent from one producing twelve, and a quarter of data is usually the minimum before reallocating anything.

Conclusion

Ask every enquirer and record their words, because analytics reports the last click rather than the source.

Expect chains rather than single sources, separate where they heard of you from what made them act now, watch for word of mouth being credited to search, put your recorded sources next to your spending to expose the mismatch, weigh sources by conversion rather than volume alone, review the log quarterly, and wait for a reasonable sample before moving any budget.