Reach

Ad performance used to be reported as impressions alone. "Three hundred thousand impressions this month" sounds large, and if it beats last month it reads as a good campaign.

That number doesn't say what happened. Three hundred thousand impressions can be 60,000 people five times each or 250,000 people once. The first is repeatedly pushing at the same people; the second is meeting new ones. Doubling the budget and getting double the impressions with the same headcount is a common outcome — more money spent, no more people met. Reported as impressions alone, that state reads as growth for as long as nobody asks how many people were involved.

Count each person once no matter how often they saw it. Three things to settle.

1. Decide what identifies a person. Logged-in account or device changes the figure: one person who sees the ad on a phone and on a laptop counts as two under a device rule.
2. State the period. Weekly reach and whole-campaign reach are different numbers, and reach with no period attached can't be compared to anything, including its own previous value.
3. Put it beside impressions. Dividing impressions, the number of times the ad appeared, by reach gives frequency, the average number of times one person saw it. Three hundred thousand impressions against 60,000 people gives a frequency of 5.

Adding it up has limits. Reach of 100,000 on each of three platforms is not 300,000 total, because anyone who saw it on all three gets counted three times. With no way to deduplicate across platforms, the sum is always larger than the truth, so multi-platform campaigns report reach per platform and give the total as a range.

"Impressions doubled and reach didn't move. We're showing it twice to the same people."

The added budget didn't reach anyone new, so it bought repetition instead.

"The goal for this campaign is reach of 200,000."

The target is how many people get met rather than what any of them do afterwards.

"You can't just add reach across three platforms."

The sum hasn't been deduplicated, so it overstates how many people were involved.

The week Pagedeck raised its ad budget for an industry event, Ravi put the two numbers side by side. Impressions went from 180,000 to 340,000 while reach moved only from 52,000 to 58,000. Frequency had climbed from 3.4 to 5.9, and click-through rate fell from 1.1 percent to 0.6 over the same period. Rather than reverting the budget, Ravi widened the targeting so the extra spend had somewhere new to go.