- Real return
- Everything bought by people whose first ever visit to the site came from this funnel's ads, divided by what the funnel spent. It counts their first purchase and every later one, with no time limit. Credit goes to the first visit, not to the purchase, so somebody who arrived through one funnel's ad and later bought a different product is counted in the funnel that first brought them in.
- Cohort
- A group of people defined by when they arrived, then followed forward. Here it is everyone who first landed between 17 July and 15 August, with everything they have bought since counted up to 3 September. The spend and the revenue describe the same people, which is what makes the return real.
- New visitors
- Every first-time visitor the ads brought to the site in the window, buyers and non-buyers together. Google Analytics identifies browsers rather than people, so one person on a phone and a laptop counts twice. Treat it as a measure of scale, not a headcount.
- Meta counts
- Meta only credits a purchase made within seven days of the click. Anything later it never sees. That is why its figure sits below the real return on funnels where people come back weeks later.
- Frequency
- How many times the same person saw the ads during the week. Under 2.0 is healthy, 2.0 to 2.5 is worth watching, above 2.5 is a problem. It counts unique people for the dates chosen, so weekly figures cannot be added together to make a month.
- Click rate
- The share of people who saw the ad and went to the site. It leaves out likes, shares and profile taps, which say nothing about whether the ad sent anybody anywhere. A falling click rate over three weeks, alongside frequency above 2.5, is what worn-out creative looks like.
The check reads the direction, not the level. There is no universal good click rate. It depends on the offer, the audience and where the ad appears, and across this account the campaigns run anywhere from 0.6% to 2.4%. So each campaign is measured against its own history: steady or rising is healthy, falling for two weeks or down more than 15% over the eight weeks is worth watching, and falling for three weeks with frequency above 2.5 is a problem.
- Buy rate
- The share of people who clicked through to the site and then bought. When this falls while the click rate holds, the ad is doing its job and the landing page, the offer or the checkout is losing people. New creative does not fix that.
- Duplicates
- Two or more campaigns selling the same product. They bid against each other for the same people, which makes both look worse than they are and shows up first as high frequency on a small audience.