Ask a platform how your advertising is going and it will tell you about impressions, reach and engagement.
All three go up when you spend more. That is the entire trick.
The chain nobody completes
Impressions, then clicks, then enquiries, then quotes, then jobs, then money.
Most businesses can see the first two, because the platform hands them over. Almost nobody can see the last three, because that requires connecting the platform to what happened afterwards in your own business.
Which means the decision about next month''s budget gets made on the two numbers that matter least.
What attribution actually requires
Every lead needs its source attached at the moment it is captured, and that source has to survive all the way to the invoice. That is what our data and reporting work is for.
Not "came from Google". Which campaign, which ad, which search term. The difference matters, because within one campaign some keywords produce work and some produce tyre kickers, and the average hides both.
What it looks like when it works
For an earthworks client over sixty days: search terms ranking went from thirteen to five hundred and eight. Pages earning attention went from nine to sixty one. Enquiries went from twenty five to forty two a month.
That third number is the only one that pays. The first two are useful because they explain the third.
The number worth arguing about
Cost per lead, and then cost per job.
On that same client, search cost per lead went from $577.98 to $148.10. Social cost per lead went from $221.14 to $92.34.
You cannot calculate either without knowing which enquiries came from where. Which is why attribution is not a reporting nicety. It is the thing that lets you decide where money goes.
The honest caveat
Attribution is never perfect. Someone sees a Facebook ad, does nothing, Googles you three weeks later and rings. The phone call gets the credit and the ad that started it gets nothing.
Anyone who tells you their tracking is complete is selling something. The goal is not perfection. It is being substantially less wrong than deciding on impressions.