Your ROI target should come from your margins, not an industry chart.

Ask five business owners what a good return on ad spend looks like and you'll get five different numbers, usually repeated from somewhere they read once and never checked against their own books. Return on ad spend, or ROAS, measures how many dollars come back for every dollar spent. A nail salon owner hears "3x ROAS is the bar" and either panics because they're below it or relaxes because they're above it. Neither reaction means anything until the number is run through that salon's own math.

Why the industry number can't tell you what you need to know.

Your business is not the average. It has its own costs and its own customers, and the only ROI target worth trusting is one built from those two things.

Start with the plainest case: a coffee shop spends $200 on local Facebook ads and gets 40 new customers through the door, each spending $6 on their first visit. That's $240 in first-visit revenue against $200 in ad spend, a 1.2x return, which sounds mediocre next to any benchmark chart.

But coffee shops don't make money on first visits. They make money on the customer who comes back three times a week for the next two years. If even a third of those 40 new customers become regulars, the $200 wasn't a weak campaign. It was one of the better $200 that shop ever spent. A benchmark number, built from averages across businesses that don't share this shop's repeat-visit pattern, would have told the owner to pull the plug.

This is the problem that Neil Thomas Bendle and Charan Bagga describe in their review of marketing metrics: the metrics themselves aren't broken, but practitioners routinely misunderstand what each one actually measures and where its limits are. Return on ad spend measures a transaction. It was never built to measure a relationship, and a lot of small businesses are built on relationships, not transactions. No single ratio can capture whether an ad is building the kind of trust a relationship depends on, which is why Knowing learns your business first and judges the ad second.

The number that actually matters is what you can afford to pay.

The useful question isn't "is my ROI good?" It's "what can I afford to pay to win one customer, and still come out ahead?" That number comes from two things you already know: your margin on a typical sale, and what a typical customer is worth to you over time, not just on day one.

Economists and marketers call that second number customer lifetime value, usually shortened to CLV: an estimate of the total profit a business can expect from one customer across the whole time that customer keeps buying, not just their first purchase. Mohammad Safari's research on the subject ties CLV directly to profitability management, arguing that the rise of transaction data and a shift toward customer-centered marketing are what pushed businesses to estimate this number rigorously instead of guessing at it.

The industry average was never your business. It was an average of businesses that are not your business.

Once you know roughly what a customer is worth over their relationship with you, not just their first receipt, you can work backward to a number that actually means something: the most you can spend to acquire one customer and still turn a profit. That's your real target. It will look nothing like the number in the benchmark report, because it's built from your margins, not an average of everyone else's.

There's no one number that settles the question.

Lynette Ryals makes a related point in her chapter on assessing marketing performance: businesses shouldn't settle for a single "silver metric" to judge whether advertising is working. Part of what that chapter addresses is how to account for the indirect value of a customer, the value beyond the receipt in front of you, which is exactly the gap a pure ROAS number leaves open.

A referral is worth something. A customer who posts about your shop is worth something. A customer who never complains and never churns is worth something. None of that shows up in a transaction-level ROI calculation, and all of it should show up somewhere in how you judge whether an ad campaign earned its keep.

This is also why marketers have struggled for decades to prove advertising's value to the people who control the budget. Jonathan Knowles, Neil Thomas Bendle and Moeen Naseer Butt argue that marketers have spent too much energy trying to get external financial reporting rules changed to capture marketing's value, when the real opportunity is in a business's own internal, decision-oriented numbers instead. Their framing: marketers have pursued the wrong audience, chasing external financial reporting standards instead of the internal, decision-oriented numbers that actually run the business. The same logic applies to a shop owner chasing an industry benchmark instead of their own ledger. The right audience for your advertising math isn't a trade publication. It's you, looking at your own margins.

Picking a way to calculate what a customer is worth.

Customer lifetime value isn't one formula. Shirley Yvonne Coleman, Daniel Walker and Mohammed Rahman-Chowdhury reviewed several different approaches to calculating it — some built from how often customers return, some from how much they spend per visit, some from how long they typically stay a customer — and found real tradeoffs between them, recommending that managers choose the approach that fits their specific situation rather than defaulting to whichever formula is easiest to find online.

For a small business, the practical version of that advice is simpler than it sounds. A salon with appointment records can look at actual repeat-visit data: how often clients come back, how much they spend each visit, how long they typically stay a client before they stop. A newer business without that history has to estimate, using a rough guess at repeat rate and average order size, and should treat that estimate as provisional until real numbers come in.

What this changes about judging an ad before it runs.

None of this means ROI or ROAS are useless numbers. It means they're incomplete on their own, and dangerous when borrowed from somewhere else. The fix isn't a better benchmark. It's doing the arithmetic once, using your own margin and your own repeat-customer pattern, and keeping that number as the bar for every campaign after.

That math also changes what you should be asking before you spend the money, not just after. If a campaign's job is to bring in first-time customers who tend to stick around, judging the ad on first-sale revenue alone is the wrong test from the start. The question worth asking before launch isn't whether the ad will hit some borrowed number. It's whether the ad is built to bring in the kind of customer your own math says is worth having.