If you can't trace a sale back to an ad, you're guessing

Every few months a client tells us a campaign "just isn't converting." And every few months the actual problem turns out to be the same one: the sale happened, but nothing in the ad account ever heard about it.

The ad gets paused. The channel gets written off. The business keeps spending on whatever channel happens to report numbers it can see, whether or not that channel is actually the one doing the work.

Why "it felt like it worked" isn't good enough

Ask most business owners how they know an ad campaign is working, and the answer is usually a feeling — enquiries went up that month, or a friend mentioned seeing the ad. That might be true. It's also not something you can act on, because it doesn't tell you which specific ad, audience, or creative produced the result you're trying to repeat.

Without traceable data, every budget decision becomes a guess dressed up as a strategy. You can't scale what you can't identify, and you can't kill what's quietly failing if it looks identical, on the surface, to what's working.

A campaign you can't measure isn't a bad campaign or a good one. It's an unknown, and you're paying to keep it that way.

What broke tracking, and why it's not coming back

A few years ago, a single tracking pixel dropped on your website was mostly enough. Browser privacy changes, ad blockers, iOS updates, and cookie restrictions have each chipped away at that — a meaningful share of conversions now simply never reach the ad platform through the browser alone.

This isn't a temporary glitch to wait out. It's the direction every major browser and OS is moving in, and it's not reversing.

The result is under-reported conversions, which quietly skews everything downstream: cost-per-lead looks worse than it is, the algorithm optimises toward the wrong signals because it's not seeing your best conversions, and decisions about what to scale get made on incomplete data.

The three layers of an ad tracking setup that actually holds

An ad tracking setup that survives these changes doesn't rely on one signal. It stacks three, so that if one layer misses an event, another one catches it:

1. Browser-side pixel. Still useful, still worth having, but no longer trustworthy on its own for anything that matters financially.
2. Server-side event tracking (Conversions API). Sends the conversion event directly from your server or CRM to the ad platform, bypassing the browser entirely. This is the layer doing most of the real work now.
3. UTM-tagged links tied to your CRM. Lets you match a lead in your sales pipeline back to the exact ad, audience, and creative that brought it in — which matters even more once a lead takes days or weeks to close.

Getting UTMs right without a developer

UTM parameters are the small tags added to the end of a link (utm_source, utm_medium, utm_campaign) that tell your analytics and CRM exactly where a visitor came from. Most businesses either skip them entirely or use them inconsistently, which makes the data they do collect unreliable.

• Use a consistent naming convention across every campaign, so "meta_insta_reels" doesn't sit next to "Instagram-Reels-Q2" in the same report a month later.
• Tag every paid link, not just some. A single untagged campaign creates a blind spot in exactly the data you need most.
• Pass UTM data into your CRM as a lead field, not just into analytics, so sales conversations and closed-deal records carry the source with them.

Why the Conversions API matters more than the Pixel now

The Conversions API sends event data server-to-server, which means it isn't affected by ad blockers, cookie consent choices, or browser privacy settings the way pixel-only tracking is. Set up correctly, it recovers a meaningful share of the conversions that browser tracking alone now misses — and because the ad platform sees more of your real conversions, its optimisation gets better too, not just your reporting.

This isn't a developer-only task anymore. Most CRMs and ecommerce platforms now offer direct or near-direct integrations for server-side event tracking, which is worth setting up before spending meaningfully on any paid channel.

How to read numbers you can actually trust

Signal	What it actually tells you
Platform-reported conversions	A partial picture — useful for trend, not a source of financial truth on its own.
CRM-attributed closed deals	The real number: which campaigns produced revenue, not just leads.
Cost per lead vs. cost per closed sale	Two different metrics — cheap leads that never close aren't cheap.
Time-to-close by source	Some channels bring faster-closing leads than others; this changes how you should pace budget.

Fixing your ad tracking setup rarely changes what your ads are doing. It changes what you can finally see them doing — and that's usually the difference between a budget decision made on evidence and one made on a hunch that happened to be repeated often enough to sound true.

Common questions

Do I need a developer to set up server-side tracking?

Not always. Most major CRMs and ecommerce platforms now offer built-in or partner integrations for Conversions API setup. A developer helps for custom builds, but it's no longer a strict requirement for most standard ad tracking setups.

Will fixing tracking make my cost per lead go up or down on paper?

Often it goes up at first, because you're now seeing conversions that were previously invisible get properly counted against the right campaign. That's not a worse result — it's a more honest one.

Does this apply if I'm only running Google ads, not Meta?

Yes. Google has its own version of server-side tracking (Enhanced Conversions), and the same three-layer principle — pixel, server-side, UTM-to-CRM — applies regardless of platform.

How often should an ad tracking setup be checked once it's live?

A quick check every quarter is usually enough, plus a check any time you migrate CRMs, redesign your site, or change checkout/lead-form providers, since any of those can quietly break the connection.

What's the single biggest tracking mistake you see?

Relying entirely on the ad platform's own dashboard as the source of truth, with no CRM cross-check. The platform only knows what it can see — it has no way of confirming which leads actually became paying customers.