Performance, demystified. Most teams measure motion, not money. They count likes, impressions, watch time, and then guess which dollar did the work. The old way was a marketing report on one screen and a bank statement on another, with a person in the middle insisting the two were related. We do it differently. Every dollar gets tied to a result, end to end, with no faith required. Today I walk the whole method out loud: a UTM spine that labels every link, GA4 attribution that reads those labels, and a reach to capture to convert funnel that shows where money actually moves. No hype. No vanity metrics. Just the receipts. Let us start at the very first click, where everything either gets named or gets lost.

First, the spine: UTM parameters. A UTM is just five tags you bolt onto a link. Source is where it came from, like linkedin or newsletter. Medium is the kind of traffic, like social or email. Campaign is the push, like w23-performance. Add content and term when you are testing two versions. The rule that saves you is a naming convention written down in one file, lowercase, no spaces, no synonyms. Pick utm_source equals linkedin and never also write LinkedIn or li. One canonical name per thing. Build the links with a single source of truth, a spreadsheet or a builder, so a human never hand-types a tag. That spine is the difference between attribution that adds up and a dashboard full of unknown traffic you cannot defend.

Now the brain that reads the spine: GA4. Google Analytics 4 ingests every tagged link and files the visit under the right source, medium, and campaign automatically. But traffic is not the point. Conversions are. So you define events that mean something to the business, then mark the real ones as key events: a form submit, a booked call, a checkout. In GA4 you create the event, flip the key-event toggle, and now the platform knows which actions are worth money. Tie a monetary value to each one where you can, even an estimate, so a lead is not treated the same as a sale. From here, every report stops counting clicks and starts counting outcomes. That is the quiet shift that makes the rest of this honest.

Attribution is just the question: which touch gets the credit. GA4 lets you choose a model, and the default is data-driven attribution, which spreads credit across the touches that actually preceded a conversion instead of handing it all to the last click. Last-click is the comfortable lie. It thanks the bottom of the funnel and starves the top that fed it. To audit this yourself, open the runReport in the GA4 Data API, or the Exploration tab in the interface, and pull conversions broken out by sessionSource and sessionMedium. Now you can see the newsletter that started the journey, not just the retargeting ad that closed it. Same data, fairer story. The number you trust is the one you can trace back to a named link.

Picture the funnel as three honest stages: reach, capture, convert. Reach is everyone the work touched, the impressions and visits. Capture is the smaller group that gave you something back, an email, a form, a booked slot, a known identity instead of an anonymous view. Convert is the slice that paid or did the thing that pays. Each stage has one rate that matters: reach to capture tells you if the message earns a hand-raise, and capture to convert tells you if your follow-through closes. In GA4 you build this as a funnel exploration, three steps, and watch where the drop is steepest. The biggest leak is your biggest lever. You do not need more reach if capture is where the floor falls out.

Here is where vanity metrics get fired. A vanity metric is any number that goes up without telling you whether you made money: raw followers, raw impressions, raw views. They feel like progress and budget like a tax. The discipline is simple. For every metric on the dashboard, ask one question: if this number doubled, would revenue move. If the answer is no, it is context, not a target. So we report cost per captured lead, cost per conversion, and revenue per source, all pulled through the UTM spine into GA4. The funnel rates sit next to the spend. Now a campaign is not winning because it is loud. It is winning because the convert stage filled and the cost to fill it went down.

Proof, on ourselves. This long-form, every social cut, the captions, and the spoken track all came from one brief and one topic file. The narration you are hearing is the caption on screen, one track, no second pass. The office that publishes it is live, and the same UTM convention links it back here, so when someone arrives we know the exact source, not a guess. We run a parity gate before anything ships: the rendered video, the spoken words, and the on-screen text must match, or it does not pass. That is attribution applied to our own production, not just our ads. The method is not a slide we sell. It is the wiring under the thing you are watching right now.

So tie every dollar to a result. Lay the UTM spine first, one canonical name per source, built from a single file so nothing is hand-typed. Let GA4 read it, mark your key events, and put a value on each. Choose data-driven attribution so the top of the funnel gets the credit it earned. Watch reach to capture and capture to convert, and fix the steepest drop before you buy more reach. Then report cost per conversion and revenue per source, and let the vanity numbers sit in the corner as context. That is ROAS you can defend in a room, traced link by link. Minus the hype, this is just bookkeeping for attention. Name it, measure it, follow the money. The receipts do the talking.