Facebook and Google know more about the average person than their closest friends do. And yet when you run ads on either platform, they will burn through your ad budget chasing people who are never going to buy your offer.
If they already know everything about us, how come your ads are not profitable?
Why isn't all that data enough?
Most advertisers blame the algorithm. The algo is dead. The targeting is off. The platform broke something.
Different version of the same conclusion, that whatever the platform does on its end has to be the part that failed.
But the truth is something else. The platforms still see almost everything they always did.
What they lost is the one thing they need to actually use that data. And that one thing comes down to something most advertisers never think about.
Knowing where someone lives is not the same as knowing whether they are about to buy your product this week. One is a profile. The other is an intent.
And the platform has only ever been good at the second one when a very specific signal is flowing in real time, no matter how detailed the profile gets.
The signal that used to be free
That signal is recency. Recent behavior. What people clicked, what they put in a cart, what they paid for, all in the last few hours and days.
The algorithm was eating it constantly, on every account, for free.
You probably remember what it looked like. A Shopify store with a copy-pasted pixel and a product that had no business making money was making money.
People look back and call the old algorithm magic. It was not. It was doing the same job it does today. The signal it was eating was just arriving clean.
The pixel was enough because the platform was already pulling fresh recent behavior off every browser on the open internet.
Clean recency is the one thing that lets an algorithm predict who is about to buy.
When the signal stopped coming clean
Then April 2021 happened. iOS 14.5 came out and started blocking the data the platforms used to pull off browsers automatically.
Ad blockers followed. Safari tightened its privacy. Other browsers did the same. The pipe that used to carry clean recency to the algorithm started to clog.
The recency signal did not stop existing. It just stopped arriving on its own.
Stale where it used to be fresh. Missing where it used to be complete. Days late where it used to arrive in seconds.
Same algorithm. Same appetite. Broken pipe. From that point on, somebody on your side of the screen has had to put the signal back.
Now they need the data from you
And that somebody is you - whether you ever signed up for the job or not. The platform's response to losing clean recency was not to get smarter on its own.
It was to start expecting you to send the missing signal back. Once you get that wrong, every dollar you spend gets handed to an algorithm guessing in the dark with your money.
The pattern is the same on all accounts. The first week or two looks great. ROAS holds. Then something shifts. Spend climbs but revenue does not.
Most advertisers look at the creative and call it tired. Some blame the audience. A few rebuild the offer. Almost nobody looks at the thing that quietly stopped feeding the algorithm in the background.
Here is what is happening underneath:
The platform can usually deliver initial sales out of its own existing buyer pool, even when the signal is weak. The first week is mostly the platform burning through that pool.
Once it runs out, the algorithm has to learn about your specific buyer for your specific product.
And the only way it learns is by being told who bought, which ad brought them, and when. On time and clean.
When that round-trip does not happen, the algorithm cannot learn.
But it still has to spend the budget. So it spends on guesses, and the guesses get worse as the signal decays. That is when the symptoms show up.
Results drop after a strong start. CPMs creep up. ROAS plateaus. A lot of advertisers have lived this exact sequence and reached for the wrong cause every time.
The platforms are sending you customers on the understanding that you will send the buyer data back in return.
Once that exchange stops happening, the algorithm has no choice but to spend your budget on guesses.
Tracking is the supply line, not the report card
Which is why tracking now has to mean something different than what most advertisers think it means. The old idea was that tracking watches your site and reports back.
The pixel sits there, sees a purchase, tells the platform. That mental model was accurate in 2019.
It is not accurate now. The job changed under the same word.
What tracking actually has to do today is bypass everything in the way, capture the conversion data anyway, and send it back to the platform on time.
The pixel firing green is one small piece of that, and the smallest piece by far.
What the algorithm is actually asking for is three things. Who bought. Which ad brought them. How recently it happened.
Pull any one out and the round-trip is broken, even when every dashboard on your screen looks healthy.
The platforms still see almost everyone. What they no longer see, on their own, is the moment somebody becomes a buyer for your specific offer.
Tracking is the pipe that carries that moment back. When the pipe runs clean, the algorithm has what it needs and delivers results. When the pipe is pinched, it does not.
Tracking is not a reporting tool sitting at the end of your funnel. It is the supply line feeding the algorithm the one signal it needs to keep finding your buyers.
The platforms have not gotten dumber since 2021. They have gotten hungrier. The signal they used to pull off the open internet for free now has to be handed to them.
The supply line that hands it over is on your side of the screen.
Once you see that, the question stops being why your ads are getting harder. It starts being whether the one pipe everything depends on is actually running clean.
And that is why tracking is more important now than ever.
