Discovery MarketingBook a call

How we measure

Our goal is incremental sales.

Sales you wouldn’t have made without the ads: new customers, and existing customers who come back because of an ad. Here’s how we find them, and how we cut the ads that only take credit.

Two kinds of sales.

When you pay for ads, the number that matters isn’t how many sales the ads touched. It’s how many sales happened because of the ads. Marketers call that incrementality: revenue you have that you wouldn’t have had without the ad.

It sounds obvious. It’s also where a lot of ad money quietly disappears.

Our goal is new revenue and profit for you: new customers, and existing customers who buy again because of an ad. What we don’t want is to sit in the middle and take a cut of a sale you would have made anyway.

We won’t pretend every sale is new, though. We run retargeting, and some people who see a retargeting ad were going to buy no matter what. No honest agency can tell you all of its sales are incremental. Our job is to keep that share small, find the ads that are only taking credit, and turn them off.

  • Would have bought anyway. Ads that only take credit for these get cut.
  • Bought because of the ads. New customers, and ones who came back. That’s the job.

How some agencies take credit for sales you’d have made anyway.

Nothing in the ad dashboard flags this. Every one of those sales really did happen after someone saw an ad. The question the dashboard can’t answer is whether the ad made any difference.

Subscription retargeting

What some agencies do

Aim most of the ads at people who already subscribe. The report shows lots of sales every month, because those people buy every month. But they’d have bought anyway, and no new customers came in.

What we do

Keep retargeting to a small share of the budget, measure it with holdout tests, and judge it by new customers and extra purchases, not by how many sales it touched.

Example: who a subscription brand’s ads reached

  • Already subscribers: they buy every month anyway
  • New people: the only sales the ads could add
An example, not a real brand’s numbers.

Branded search

What some agencies do

Put a lot of money into Google ads on your own business name. The conversion rate looks amazing, as high as 40 to 50% in accounts we’ve seen, because those people were already searching for you.

What we do

Test turning it down. Keep it only where a competitor or a “best of” list is taking those clicks, and move the rest to what brings in new sales.

Free ad audit.

Send us a look at your current ads. We check how much of your money goes to retargeting (Facebook and Instagram ads to people who already visited your site) and how much to branded search (Google ads on your own name). Then we recommend a test you can run yourself, without us, to see whether you can lower your cost by cutting or reducing them.

Tests you could run yourself, for example:

  • Pause your own-name ads for two weeks and compare your revenue.
  • Cut retargeting to a small share of your budget and compare your new customers.
  • It’s free
  • We only look: read-only access, or screenshots
  • You don’t have to work with us to use the results

An email or a phone number is enough.

Which ad platforms do you use? Pick all that apply.
How would you like to share your ads with us?

Free. We only look. You don’t have to work with us to use the results. Privacy

When existing customers do count.

Existing customers aren’t the problem. A customer you’d lost, coming back because of an ad, is a real sale you wouldn’t have had.

Say you sell a daily health drink. Someone used it for a while, then stopped. A new ad comes up where an athlete they follow says they noticed the difference after three weeks, and they decide to start again. That person was already your customer, and the sale is still new. That’s exactly the kind of sale we want ads to bring in.

  1. BeforeBought every month
  2. ThenStopped buying
  3. An adA new reason to try again
  4. NowBack, and buying
An example. The sale after the ad is new: it wouldn’t have happened without it.

How we measure it.

You can’t see incrementality in one report. You have to test for it. These are the tests we use.

  1. Holdout tests

    We pause the ads for a set time and see what your sales do on their own. That’s your baseline, and anything above it while the ads run is what the ads are adding.

  2. Geographic holdouts

    Ads stay on in some areas and off in others that look alike. If sales rise only where the ads run, the ads are working.

  3. Facebook and Instagram lift tests

    Meta can split people into a group that sees your ads and a group that doesn’t, and compare what each group buys.

  4. Google branded-search tests

    People searching your business by name were already looking for you. We turn branded-search spend down and watch revenue. If it doesn’t drop, that money was taking credit for sales you’d have made anyway.

When a test shows an ad is only taking credit, we cut it and move the money to the ads that bring in new customers. We’d rather show you a smaller number that’s real than a big one that isn’t.

One honest caveat.

As spend goes up, ads reach people who are less interested, so the cost of each sale can rise. That doesn’t mean the ads stopped working. It means you’re reaching further. So we compare like with like, the same season, the same offer, a similar budget, before we call a change good or bad.

What this means for you.

You keep your own ad accounts, so you can see every number we see. We agree on a goal before we spend a dollar, we test whether the ads are adding sales, and we tell you when they aren’t. That’s how a percentage of ad spend stays fair: we only want to earn more when the ads are making you money you wouldn’t have had.

See how ads pay you back, and where we stop

Let’s find the sales you’re missing.

A free 30-minute call with Jerry. We’ll agree on your goal before you spend anything.

Ten days. $0 if you stop. Free 30-minute call with Jerry about the 10-Day Lead Program.

Book a call