Burn the dashboard

Most of your attribution report is a well-designed lie.

Marketers talk about attribution like it’s a science.

It’s closer to astrology.

I’ve run $20M+ in performance marketing campaigns where every click had to earn its keep. And I’ve run brand campaigns where you’re flying blind and hoping for signs.

The stuff I couldn’t measure often worked the best.

But most marketers still worship dashboards. We’ve built million-dollar strategies off metrics that wouldn’t pass a school science project.

Dashboards don’t lie. They just leave out nearly everything that matters.

The fairytale is simple, track every click and you’ll know what works. If it’s not in the data, it didn’t happen.

Most sharing is invisible.

Dark social accounts for roughly 77% of all content sharing, and none of it shows up in your analytics. iOS 14.5 gutted tracking so badly that the attribution numbers most marketers rely on are essentially made up. Your dashboard shows you what’s easy to count, not what’s true.

And that might be the best thing that’s happened to marketing in a decade.

Because once obsessive attribution dies, you have to ask better questions. Do people remember you? Are they talking about you? Are they coming back on their own, without a discount code dragging them in?

That’s marketing. The dashboard just showed you the part that was easy to measure.

I believe most attribution is fiction.

But pretending you don’t need any numbers is stupid too.

Try walking into a board meeting with no data, just vibes.

Let me know how that goes.

You need numbers.

The mistake is treating them as truth instead of what they are: a rough signal. A hunch dressed up in a spreadsheet.

If you starve your brand just because it’s hard to measure, you’re killing demand. I’ve seen this happen. A campaign is working, people are talking, pipeline is filling up, but the dashboard can’t prove the connection. So someone pulls the budget. Six months later the pipeline dries up and nobody connects the dots.

Ryanair gets this. Their TikTok content is chaotic, unfiltered, and impossible to tie directly to bookings. But they estimate earned media value and compare it against paid spend. The CFO gets a number. The number is imperfect. The brand keeps growing.

Attribution’s a hunch. Treat it like one.

The honest approach is incrementality testing. Instead of asking “which channel gets credit,” you ask “did this channel actually change anything?” Run a channel for two weeks. Pause it for two weeks. Measure the difference. That’s your real impact. It doesn’t need cookies, pixels, or creepy tracking. It just needs patience and a willingness to find out that some of your spend is doing nothing.

Most marketers would rather not know. The dashboard lets you tell a story. Incrementality testing might wreck it.

First-party data is the only data that matters now. If you don’t own the relationship, you don’t own the result. Attribution models built on third-party data are dead. Email lists, direct traffic, customer accounts, branded search. These are your actual signals. They’re messy and incomplete, but at least they’re yours.

Tracking is worse. Budgets are tighter. Every founder I talk to secretly knows that half their attribution report is a well-designed lie.

I still use dashboards. Of course I do. But I’ve stopped treating them like gospel. The problem was never the dashboards. It was how much we believed them.

The best marketing I’ve done couldn’t be measured at the time. It showed up later. Repeat customers. Word of mouth. Deals that closed because someone remembered something they couldn’t trace back to a campaign.

If you need a pixel to prove you did your job, you probably didn’t.

Martin, Chief Ranter at Uncharted

P.S. this week’s track is Enjoy The Silence by Depeche mode