A couple of weeks ago I wrote about how positioning drifts without you noticing. Buyers describe their problem in words you stopped using eighteen months ago and the website talks past the sales call.

‍

A few of you replied saying the harder problem was one step upstream of that. "We know things need updating. We just can't agree on which channels to fix it in first." 

So this week: attribution. 

‍

Specifically, the mid-year moment where founders and CFOs ask what's working, get a spreadsheet back, and cut the wrong thing.

The mid-year cut that makes things worse

Half-year review time is when this happens most reliably.

‍

Board pressure building. A push to meet targets and tighten spend at the same time before Q3. The natural question: what's performing? Marketing pulls the data. Paid search looks strong. Paid social is shaky. The newsletter, the founder's LinkedIn, the content archive - they have cost-per-click numbers that make the CFO wince. Nothing trackable attached.

‍

So you cut the content. You put more behind paid search.

‍

Then pipeline drops. CAC goes up. The paid search that was converting so well starts needing more budget to hold its numbers. Nobody can quite explain why.

‍

Here's what is likely to have happened. The content, the newsletter, the LinkedIn presence, were warming buyers up before they ever searched. When they finally did search, they already knew the brand. They clicked. They converted. The last click went to paid search. The six months of upstream work went unrecorded.

‍

You removed the warmth and then wondered why the conversion cooled.

What attribution models miss

Standard attribution records the journey from the first trackable touch. It doesn't record anything before that. The podcast episode someone listened to on the commute. The LinkedIn post that made them forward your newsletter to their head of growth. The webinar they watched six months before they were ever in-market.

‍

This is what's been called the dark funnel. It sounds mysterious bu it isn't at all. Buyers just don't announce the early part of their journey because there's no form to fill in, no UTM parameter, no session to log. They were already half-sold before they showed up in your data.

‍

The tell is in the sales call. Ask any rep (or even better, analyse the sales transcripts!): where do your warmest leads come from? What did they already know before the call started? Almost always they've been following the content, reading the newsletter, watching the founder's posts. None of it shows anywhere in the attribution report.

‍

So the report says cut it. And the founders who follow that advice are usually back wondering why conversion is down within a quarter.

A story from this spring

A business I work with ran a quarterly attribution review in Q1. On paper, their content marketing had zero revenue attached to it. The newsletter was a cost line. The founder's LinkedIn was harder still to quantify.

‍

The team ran a quick pass on their last twenty closed deals. We listened to the discovery calls and read the onboarding survey responses. One question: what had they engaged with before the first sales conversation?

‍

Seventeen of the twenty buyers mentioned the newsletter, a LinkedIn post, or a piece of content. Several mentioned something specific. A few had been subscribed for over a year before they got in touch.

‍

The attribution model had credited none of it. The paid channels had been credited with almost everything.

‍

They didn't cut the content. They cut two paid channels that nobody could find in those twenty conversations at all.

‍

Pipeline held. CAC fell. The content budget stayed exactly where it was.

What actually tells you a channel is working

The mid-year question most teams ask is: what does the data say is working?

‍

The more useful question is: what do closed customers say influenced them?

‍

Those two questions should produce similar answers. When they don't, the gap is worth investigating before you make any budget calls. Because the channels in the gap are almost certainly doing work that your attribution model can't see.

‍

We have a short set of questions we run at onboarding with every new client. Five questions, asked in the first two weeks, that tell us more about which channels are actually pulling weight than most platforms do after six months of tracking. The pattern across a dozen or so businesses is always the same: the channels the data undervalues are the ones the buyers keep mentioning.

The one thing to do this week

Before any mid-year budget conversation, do this first.

‍

Pull your last ten to fifteen closed deals. Go to the discovery call recordings or the onboarding survey responses. Write down, in plain terms, what each buyer had already seen or read before the first sales conversation.

‍

Then hold that list next to your attribution report.

‍

If there's a channel that keeps coming up in the conversations but barely registers in the data, that's not a channel that isn't working. It's a channel your attribution model can't see. Cutting it is one of the fastest ways to make the channels you can measure perform worse for no obvious reason.

Both pictures matter. Most mid-year reviews only look at one of them.

‍

Drop me a line if you're heading into a budget review soon and want a second set of eyes on what you're looking at. I'm happy to take a look.