Every month we turn down paid media work for the same reason. The tracking is broken, the client wants to launch anyway, and we say no.
We will not manage paid media against tracking we cannot trust, and we do not think you should either.
That is a strange position for an agency to hold out loud. It costs us business. It delays our own first invoice. We hold it because we have seen what happens to companies who spend two years buying clicks against numbers nobody ever checked, and because the fix is usually cheap compared to the waste.
What tracking you cannot trust looks like
This is not a hypothetical problem we invented to sell an audit. These are the patterns we find over and over when we open a new account.
- A pixel that stopped firing after a site redesign nobody told the agency about. Months of spend, no data.
- A form that was never connected to a conversion goal. The platform believes the campaign produced nothing, so it quietly pulls budget away from it.
- A form plugin firing its events into a custom data layer that Google Analytics never reads. Everything looks wired. Every report says zero.
- A contact form with a broken CAPTCHA setting that silently rejects every submission. The form looks fine from the visitor’s side. Nobody notices for months because nobody expects a working form to eat leads.
- Duplicate conversions. The platform counts one, a plugin counts another, a manual import counts a third. Now your cost per lead looks three times better than it is.
- Call tracking that exists but attributes every call to direct traffic, which is the same as having no call tracking at all.
- Platform-reported conversions that never reconcile with anything in the CRM.
That last one deserves a number. Here is an illustrative example, not a client account. Google reports 400 conversions for the quarter. The CRM shows 240 closed deals traceable back to those campaigns. The gap is not the problem by itself. Some of those 400 were duplicates, some were junk, some were real people who never bought. The problem is that every bidding decision for the whole quarter was built on the 400.
Smart Bidding is very good at its job. It optimizes toward whatever you tell it a conversion is. Feed it bad signals and it becomes extremely efficient at producing the wrong thing. It will find you more of the leads that never close, because those are the ones your tracking told it to value.
Running ads on bad tracking is not a neutral choice
There is a common assumption that broken tracking means you fly blind for a while. Annoying, but harmless. You still get the clicks you paid for.
That is not what happens. Budget actively moves toward whatever the platform believes converted. Wrong signal, wrong allocation, every day, compounding. The campaign that actually produces revenue gets starved because its conversions never registered. The campaign producing nothing gets scaled because a duplicate tag made it look like a winner. Three months in, the account has been optimized away from the thing that works.
An agency that launches into that is charging you to make the problem worse. Then it reports on its own performance using numbers it already knows are wrong. We are not interested in grading our own homework with a broken ruler.
Why most agencies launch anyway
The incentive is not subtle. Retainers usually start when spend starts. Measurement work is unglamorous, hard to sell, and pushes the first invoice out by weeks. Nobody signs a contract excited about server-side tagging.
So the work gets bundled into the media retainer, where it competes with campaign management for attention and loses every time. Somebody adds it to a launch checklist, ticks the box, and moves on. Three months later a redesign breaks half of it and nobody is watching.
We separate the two deliberately. Measurement is its own engagement, built and paid for before the budget conversation, not smuggled into a retainer where nobody has time to check it. Our ad management fees are fixed monthly, and we never mark up your ad spend, so we have no reason to want your budget bigger than it should be.
What we tell the client who wants to launch Monday
This conversation happens constantly, and we understand the pressure behind it. Somebody has a number to hit. The competitor is already running. Waiting feels like losing.
Here is what we actually say. If we launch Monday, we will be spending your money to learn nothing. In six weeks you will have a report full of numbers neither of us believes, and we will still have to do the measurement work, except now you will have paid for two months of misallocated budget on top of it.
Two weeks of fixing tracking usually saves more than two weeks of spend would have earned. That is not a sales line, it is arithmetic. And if your tracking is genuinely fine, the audit says so and we launch. That outcome happens. We are glad when it does.
I have told a client no on a Monday launch and lost the first invoice over it. What they did not lose was two months of budget optimized toward form fills that never became customers. I will make that trade every time, and I have.
Sean McKay, founder and CEO of Site Hub
What we mean when we say we trust the tracking
Trust here is concrete and testable. It means the systems agree with each other.
GA4, the ad platforms, call tracking and the CRM tell the same story about the same lead. A closed deal traces back to its first touch. Real conversion values flow back to the platforms making bidding decisions, so the algorithm is optimizing toward revenue instead of form fills. Calls are wired to source and campaign, not dumped into direct. And platform-reported conversions sit next to CRM-verified ones in the same view, so the gap between them is visible instead of theoretical.
That is what a CRM-integrated build produces. When the systems reconcile, arguments about which channel works stop being opinions.
Tracking breaks quietly, and it keeps breaking
The hardest part to explain is that this is not a one-time fix. A tag manager change, a form redesign, a renamed CRM field, a plugin update, a developer removing a script that looked unused. None of these announce themselves. Tracking almost never fails loudly. It just starts reporting a slightly different number, and the number looks plausible, so nobody questions it.
That is why we send a monthly signal memo in plain English rather than a dashboard nobody opens. Somebody has to be responsible for noticing. If it is not written into the engagement, it is nobody’s job.
When we will spend against imperfect tracking
We are not absolutists about this, and it would be dishonest to pretend otherwise. There are two situations where we will run paid media before everything is clean.
The first is a genuinely short test with a hard budget cap and a written manual reconciliation plan. We agree in advance what we are testing, what we will count by hand, and when we stop. The client knows the data will be rough and knows why we are accepting that.
The second is a business whose only meaningful conversion is a phone call, where call tracking is already wired to source and campaign. The rest of the setup can be messy. If the phone data is clean and the phone is the whole business, we can work with that.
The line is this. We will spend against incomplete tracking. We will not spend against tracking that is actively wrong, and we will not spend against tracking we have not looked at. Those are different things.
Five questions to ask any agency before they spend your first dollar
- Show me one closed deal in my CRM and trace it back to the campaign, ad group and keyword that produced it. If they cannot, attribution is a guess.
- What is the gap between platform-reported conversions and CRM-verified ones in my account right now, and what explains it?
- Which of my conversion actions are currently feeding the bidding algorithm, and what value is assigned to each one?
- How will you know if a tag breaks next month, and who is responsible for checking?
- Do you mark up my ad spend, and does your fee go up when my budget does?
The answers matter less than whether the person can answer at all. An agency that has looked will tell you exactly where the holes are. An agency that has not will talk about strategy instead.
Frequently asked questions
Because budget moves toward whatever the platform believes converted. If the conversion data is wrong, the misallocation starts on day one and compounds every month. Fixing tracking later does not recover the spend you already wasted, and it does not give you clean history to learn from.
It reviews your GA4 setup, ad platform pixels and conversion actions, call tracking, and CRM, then checks whether what each system reports reconciles with the others. The fee is $4,500 fixed and it takes two weeks. Half the fee credits toward any build signed within 60 days.
Then the audit says so and we move to the paid media conversation. That is a real outcome and we are happy when it happens. You get a documented baseline either way, which is useful the next time something breaks.
Rarely, and only under conditions we write down first. A short test with a hard budget cap and a manual reconciliation plan, or a business whose only conversion is a phone call where call tracking is already wired to source and campaign. Anything longer or more complex waits for the build.
If you want to know whether your tracking is telling you the truth, start with an Attribution Audit. Two weeks, fixed fee, and a straight answer about what your numbers are actually worth. Call us at (585) 300-4640 or stop by the office at 350 East Ave Suite 1 in Rochester.