Your best-performing ad might be working against you

There’s an ad in your account right now that everyone is proud of. Best return on ad spend, top of the report, the one that gets pulled up in the Monday meeting as proof that things are working. I want to make the case that it might also be the one quietly doing you the most harm. It does its job well. But the job we measured it on is only half the story, and the other half, the one where your brand lives, never lands in the number.

I’ve spent years watching people fall in love with their top-performing ads, and I understand the pull. In a world where most of marketing is hard to measure, here is one thing you can measure cleanly, so of course it gets the attention and the praise. But “bestperforming” is a label, and it’s worth asking who handed it out, and what exactly they were looking at when they did. Once you look closely at that, a few uncomfortable things fall out, and they’re worth walking through one at a time.

Your best ad won a contest that only measured one thing

Best-performing is a ranking, and every ranking depends on what you chose to measure. In almost every ad account, the thing being measured is the sale. Did this ad lead to a purchase, and how cheaply. That’s a fair question and a useful one, but it’s a narrow one, because it watches the moment of the checkout and nothing on either side of it.

Picture two ads. The first is a calm, confident piece that makes people feel something about you and gives them a reason to prefer you, with no discount anywhere in it, and it brings in a steady stream of people paying full price. The second shouts twenty percent off, ends tonight, and pulls in a flood of orders at a lower price. On the report, the second one wins, and it isn’t close, because it produces more sales for less money. But the two ads are doing very different things to your business. The first keeps earning for years and leaves you looking like a brand people are happy to pay full price for. The second earns once, and teaches a few thousand people that your prices are negotiable if they just wait. The metric cannot tell those two stories apart. It counts the orders and hands the trophy to the discount. And there’s the catch: the two ads are not even playing the same game. The first is working on how people feel about you, which is what a brand really is. The second is working on the price, which is just the product wearing a smaller number for the weekend. The metric only knows how to score the second one.

So the number can’t see whether an ad made you look like a brand worth paying full price for, or a brand nobody would pay full price for if they knew a code was around the corner. It can’t see whether it built the kind of preference that brings someone back next year, or spent that preference on a single sale today. It sees the transaction, records it, and moves on. “Best-performing” turns out to mean “best at producing a checkout this week,” which is a real thing and a good thing, but it isn’t the same as “best for the business,” and the whole gap between those two is where this piece lives.

Optimization has a taste, and the taste is cheap

Here’s what makes it worse. Once you let performance pick your winners, it doesn’t just measure the ads, it starts shaping them. The system keeps pushing budget toward whatever converts fastest, and it keeps asking you for more of that, and what converts fastest is almost always the same small set of moves: a discount, a deadline, a last chance that ends tonight, a lower price said louder.

You can see where that road ends by looking at the businesses that have driven furthest down it. The big travel booking sites are the clearest example, because they have optimized for the click harder and longer than almost anyone, and it shows in every line: only one room left at this price, eight other people are looking right now, booked eleven times in the last hour, deal ends in twelve minutes. Every one of those lines is there because it converts, and they convert beautifully, and not one of them makes you like or trust the brand any more. Nobody has ever looked at that wall of red urgency and thought, now there is a premium brand. It’s the highest-converting corner of the internet and the least premium one, and those two facts are really the same fact.

You can watch the same thing happen to your own account over a few months if you keep the old work. The ads from a year ago still had range in them, some built a feeling, some told people who you are, some just carried an offer. Then the reports came in, the offerled ones always looked strongest, so you made more like those and a little less of everything else. Do that four or five times in a row and your whole account has narrowed to one note, which is pay less and hurry up. It happens gradually, one reasonable optimization at a time, which is exactly why nobody notices it while it’s going on.

This isn’t only a marketing disease, and that’s part of why it’s so hard to fight. It’s the same trap a politician walks into with an election coming, where the win that shows up before polling day almost always beats the policy that pays off in fifteen years. It’s how you end up with climate and sustainability targets that get announced, half-funded, walked back, and announced again by the next person, because the slow and unglamorous work never fits inside a single term, and the incentive only ever rewards the thing you can point to right now. Your ad account runs on that exact incentive, turned into software and pointed at your brand. And short-termism like that has a habit of staying invisible right up until it isn’t. You don’t feel the cost of the skipped long-term work while you’re skipping it, you feel it years later and all at once, when the rivers are running low and the summers have turned hotter and the thing everyone could have paid for slowly has become the thing nobody can afford to fix.

And the real cost of the narrowing is hidden. The ads you slowly stopped making were the ones with something to say beyond the offer, and those were the ones shaping how people felt about you. They were never going to win on same-day return, because that was never their job. Trim them all away and you’re left with an account that is very good at closing a sale and has lost the ability to make anyone want you in the first place, which is a strange and expensive place to end up. The machine doing the steering isn’t out to get you, it simply has no opinion about your brand at all. You asked it what gets the most sales for the least spend, and it answers that question honestly and relentlessly, and it will trade away your brand for a better number without a second thought, because it can’t see the brand and doesn’t even know a trade is being made. And the marketing version of that bill lands the same way, slowly and then all at once. For a long time everything looks fine, because the offers keep converting and the report stays green. Then one day the sales start slipping, and they keep slipping no matter how sharp you make the offer, because the discounts and the urgency were never building anything, and a brand you spent years teaching people not to love at full price eventually stops pulling at all. By then the tactics that used to perform so well are firing into an empty room.

Every discount is a lesson, and customers are quick learners

Before the lesson, a word about why the discount is there at all, because it’s rarely the innocent little tool it looks like. A discount is a confession. When sales come in soft, the quickest thing to reach for is a price cut, and it always works a little, so it feels like the answer. But you cut the price because people didn’t want you enough to pay full price, and wanting is an emotional thing. A discount is a rational argument, pay less, and you cannot fix an emotional gap with a rational lever. A lower price has never once made someone love a brand they felt lukewarm about, it just gives the lukewarm ones a cheaper reason to buy this one time. So it calms the symptom for a week and feeds the thing that caused it, and if a competitor is discounting too, you both end up cutting deeper and deeper just to stay level, each handing away more margin to look slightly cheaper than
the other.

And a price-led ad doesn’t only make a sale, it teaches the person who sees it something about you. It teaches them that your real price is lower than the one on the label, that if they wait a week something better usually comes along, and that the smart move is never to buy from you at full price.

People learn those lessons fast, and they don’t unlearn them. You can watch it on a national scale every year now. Plenty of retailers see their sales go quiet in the weeks before Black Friday, not because demand disappeared, but because we have trained an entire market to wait. Everyone knows the big discount is coming, so everyone holds off, and the full-price weeks that used to pay the bills slowly hollow out. That is the same trap running at the scale of a whole economy, and it started with a lot of individual businesses each deciding a discount was the easy answer.

And watch how the market answers that. Nobody responds to shrinking full-price weeks by discounting less. They respond by discounting more, and earlier, and deeper. Black Friday isn’t even a day anymore, it’s a week, then two, then most of November, and the offer has to get bigger every year because last year’s stopped feeling special. And it is spreading. Singles Day, Valentine’s, back to school, Mother’s Day, Father’s Day, one ordinary date after another is slowly turning into a global discount event, because the moment one brand turns a day into a sale, everyone else has to match it or look expensive standing next to them. That’s a race to the bottom playing out in real time, and the finish line is a market that only ever buys on offer.

At the company level it tightens on itself the same way. Once you’ve trained people to wait for the offer, the weeks without one start to look weak, sales dip whenever there’s nothing running, and the obvious fix in the moment is to run another offer. So you do, and you teach the lesson again, a little harder, until the whole business leans on the promo calendar just to stand up.

And here’s the part the report never shows you: this costs you three things at once, not one. There’s the margin you give away today, on every sale you discounted that didn’t need it, usually to your best customers, the ones who would happily have paid full price if you’d let them. There’s the pricing power you lose over the following years, because once people have decided what you’re “really” worth, charging full price starts to feel to them like you’re overcharging. And then there’s the one that does the most damage and never shows up in any number, your brand power, the plain fact that people will pay more for you than for the next option. That drains away because a brand that is always on sale slowly stops reading as premium or desirable and starts reading as the thing you only buy when it’s cheap. You can rebuild a margin in a quarter. You cannot rebuild the feeling that you’re worth it anywhere near that fast.

A lot of those conversions were coming anyway

There’s one more thing hiding inside that top-performing ad, and it’s the most uncomfortable one. The ads that win on return on ad spend are usually the ones sitting closest to the purchase: retargeting someone who already has the product in their basket, catching someone who just typed your brand name into search, handing a discount to a person who was already on their way to buying.

Think about what branded search really is. Someone already knows you, already wants you, and goes looking for you by name. If your ad is the thing that greets them, it gets to record that sale, even though that person was walking through your door either way. Retargeting is the same shape. It works on people who already came, already looked, already showed real intent. These ads convert brilliantly, which is exactly why they sit at the top of the report, and a big share of what they “delivered” was going to happen with or without them.

All you really did there was bring the moment forward a little, and if the nudge you used was a discount, you didn’t just fail to create that sale, you earned less on it than the customer was already willing to pay. You put a lower price in front of someone who had their wallet out at full price, gave away the margin, claimed the win, and taught them to wait for the discount next time. Three separate things happened there, and only the good-looking one showed up in the report, which calls it a win and puts it at the top. A lot of the time it’s really just an expensive way of taking credit for something that was already yours. This is why measuring whether a channel truly created a sale, rather than stood near one, matters so much, because it’s the difference between an ad that grew the business and an ad that took a cut of demand you already had, and charged you a discount for the privilege.

Judge an ad by what it leaves behind, not just what it closes

None of this is a reason to walk away from performance. Performance marketing is very good at what it does, you’d be mad to give it up, and plenty of your top ads are earning their place honestly. It’s a reason to stop letting a one-eyed metric crown your winners and shape your brand for you.

A few things help. Look hard at what your top ads are really saying, all together, and ask what they would teach a customer who saw them ten times in a row, because plenty of people will. Watch margin and full-price sell-through by channel, not just return on ad spend, so the cost of all that discounting shows up somewhere you can see it instead of hiding inside a healthy-looking number. Where you can, measure whether a channel is really creating sales or just standing next to them, because that changes which winners you want more of. And when sales come in soft, before you reach for a price cut, ask whether the problem is really the price or whether it’s that people don’t want you enough yet, because those two problems have completely different fixes and only one of them is cheap.

There’s one more, and it’s the simplest. Decide up front how much of your marketing is allowed to be pure offer, and protect the rest for work that has something to say. If every dollar has to prove itself by tomorrow morning, you’ve guaranteed that nothing ever gets to build anything, because building takes longer than a day. Give some of the budget a job that isn’t judged on the checkout at all, and hold the line on it when the weekly numbers get tempting, because they always do.

Do all of that, and the best-performing ad and the best ad for your business can absolutely be the same ad. They just won’t be the same one by accident, and they definitely won’t be the same one if the only judge in the room is a number that can only ever see the checkout.

Hugo van den Biggelaar spent 8.5 years at Nike across finance, insights and brand. He now runs BITSing in the US, a growth methodology built over thirty years and used in more than a thousand organizations. He also runs Think Biggie, a Brooklyn-based creative studio, and is a founding contributor to the Fospha Academy.