The Dark Funnel: Why Most of Your Best Leads Look Like “Direct Traffic”

Here’s a journey your analytics tool will get completely wrong. A homeowner sees your van outside a neighbour’s house. A week later their boiler starts making a noise. They mention it at work; a colleague says “we used someone decent last winter” — it’s you. They check your Google reviews on the bus, look at your Facebook page that evening, and three days later type your business name into Google and ring you.

Your analytics records: one direct visit, one phone call. Source: direct / none. Every part of the journey that actually won the job — the van, the word of mouth, the reviews, the Facebook presence — is invisible. Marketers call this the dark funnel, and for local businesses it’s not a corner case. It’s most of your good leads.

Why this costs you real money

The danger isn’t the mislabelled data — it’s the decisions made from it. The classic mistake runs like this: the report says Facebook ads produced “only three leads” last month, so the budget gets cut. Two months later enquiries are down across the board, including the “direct” ones — because those Facebook ads were feeding the invisible journey. The ad didn’t get the click, but it put your name in the neighbourhood’s head.

What you can actually do about it

1. Just ask

The most underrated attribution tool is the question “How did you hear about us?” asked on every enquiry, with the answer written down somewhere. It’s imperfect — people compress “saw your ad twice, then my mate mentioned you” into “my mate mentioned you” — but it catches what software can’t see.

2. Track calls properly

For trades, 70% or more of enquiries are phone calls. If calls aren’t tracked back to their source, your data is mostly fiction with a confident font. Call tracking numbers fix this and cost a few pounds a month.

3. Judge channels on the right timescale

Search ads should justify themselves within weeks — they catch people already looking. Awareness channels (Meta, vans, sponsorships, leaflets) work on months. Cutting an awareness channel after four bad weeks is like digging up seeds to check if they’re growing.

4. Watch the total, not just the labels

The cleanest signal for a local business is brutally simple: total enquiries per month against total marketing cost per month. If you switch a channel off and the total dips six weeks later, the channel was working — whatever the attribution report said.

The takeaway

Don’t aim for perfect attribution; for a local business it doesn’t exist. Aim for not being fooled: ask every caller how they found you, track your calls, and trust the total enquiry number over any per-channel label. That’s enough to stop cutting the things that are quietly winning you work.