How many distinct ad concepts did you test last month? Not resizes. Concepts

If the honest answer is "we refreshed the hero ad last month", you've found a core leak. The failure smell is everywhere once you know it: CPA creeps up, the team reads it as channel decay or algorithm changes, and the response is to touch the budget, the bids, the audiences. Everything except the ads. The creative sits untouched because creative is "the agency's thing", or because there's one designer and a six-week turnaround, or because last quarter's hero ad was expensive and someone senior liked it.

Here is the uncomfortable arithmetic. Nielsen Catalina's analysis of roughly 500 campaigns found that creative accounts for 47% of advertising's sales lift. Targeting accounts for 9%. Most acquisition teams resource those two numbers the other way round: a media buyer optimising audiences full-time, and new creative arriving quarterly.

And the gap has widened since that study ran. Post-ATT, the platforms took most of the targeting away from you anyway. The auction finds your audience by watching who responds to your ads. Which means your creative is the targeting now. A team that ships two concepts a quarter has handed its distribution to an algorithm and given it almost nothing to work with.

Small companies get fixated on brand and polish. The anti-ego truth is that in a saturated marketing milieu, your brand and polish mean very little to a stranger mid-scroll. As long as an ad isn't antithetical or deleterious to your brand, push the creative envelope. CPA is pragmatic: it's driven by ads that work, with audiences that matter.

Creative excellence is the third result area we assess in acquisition, after audience clarity and channel selection, and the sequence matters. Creative is where your audience insight and your channel choice either pay off or get wasted.

What this result area actually is

Creative excellence does not mean beautiful ads. It means a working system for producing, testing and reading ad creative, so that winning patterns are found on purpose rather than by accident. The thinking follows Reforge's growth frameworks and structured creative testing practice: distinguish Pre-Flight tests (does this concept deserve budget at all), New vs BAU tests (challenger against incumbent), and Scaling tests (does the winner hold as spend rises). The test of maturity is simple to say and rare to pass: can you name the messaging angle, visual style and call to action that win for you, and is that answer written down anywhere other than in one media buyer's head?

The four levels of maturity

Every result area in the Revenue Engine framework is assessed on a four-tier ladder, from Foundation to Optimise. For creative excellence, the climb looks like this:

Two things worth noticing. First, T1 is a volume floor, not a quality bar. Two to three variations per campaign per channel sounds modest, and most teams still miss it, because the bottleneck is process, not talent. Second, the ladder climbs from output to patterns. T2 is not "we test a lot". It's "we know why the winners win", and that knowledge compounds across every future brief.

Best practice and worst practice, from the real world

The best-practice case everyone cites is the one most teams learn the wrong lesson from. Dollar Shave Club's launch video cost $4,500 and took a day to shoot. It brought in more than 12,000 orders inside 48 hours, and the company sold to Unilever four years later for $1 billion. The tempting read is "make a viral video". The accurate read is that the video was audience insight, articulated. Michael Dubin didn't make a funny ad about razors. He made a precise ad about the absurdity of paying $20 a month for brand-name blades, in exactly the register his audience used to complain about it. The production was cheap. The insight wasn't. That's creative doing the 47%.

The worst-practice case spent more on advertising in six months than most consumer apps spend in a decade. Quibi raised $1.75 billion, put at least $63 million into TV, web and print advertising, and bought a $5.6 million Super Bowl spot that ranked among the worst rated of the night. Six months after launch, it shut down. The reach was enormous. The targeting was professionally managed. And none of the creative ever answered the only question that mattered: why does this exist, and why should I care? You cannot buy your way past creative that doesn't land. Quibi bought the 9% and skipped the 47%.

The pattern holds at less famous scales, and I've lived the quieter version. At one company I worked in, confirmation bias ran supreme. Ads were created based on what the CEO liked, and they took three weeks plus to produce because the CEO micromanaged every one. Then they shipped, and were never fully measured, because ego was involved and the numbers might have had opinions.

So I built a test. Generic AI-driven ads, and they were ugly, at a much higher cadence, testing messaging first with the plan that winners would then be handed to design and beautified. The ugly AI ads demolished the polished, CEO-approved, "this will definitely work" ads. And a video ad the designer made on his own initiative performed particularly well. But it wasn't the CEO's, so it never got traction. The test worked. The politics didn't.

Where to start this week

One move: the concept count. Go through last month's active campaigns and count distinct creative concepts. Different messaging angle, different visual approach, or different CTA counts as a concept. A resize, a recolour or a new headline on the same idea does not. Most teams expecting a healthy number find low single digits.

Then install the T1 floor: 2-3 genuine variations per campaign per channel, every cycle, and a one-line log of what won and what you think it means. That log is the seed of T2. It costs nothing. It's also where the compounding starts.

Fix the leaks. Fix your world.

Creative excellence is one of six result areas in the Acquisition module, and Acquisition is one of eleven modules in the Revenue Engine diagnostic. If any of this resonates, our diagnostic maps your entire growth funnel and shows you where the real leverage sits. No deck. No pitch. Just clarity on where the engine is misfiring.

Book a discovery call

Jonathan Stanton-Humphreys is Founder and CEO of Revenue Engine. He spent a decade as a commercial executive in B2C and B2B tech, and built Revenue Engine after watching revenue leak through strategic and implementation dysfunction, and the joins that nobody owned.

FAQ

How many ad creative variations should I test at once? At foundation level, 2-3 genuine concept variations per campaign per channel. More than that before you can read results cleanly just adds noise. The discipline of always having variations live matters more than the number.

How do I know whether it's creative fatigue or channel decay? Ship fresh concepts before you conclude anything. If new creative restores performance at the same spend and audience, it was fatigue. If it doesn't, look at the channel, the offer or the audience. Teams that never test creative can't tell these apart, and usually pay for the wrong fix.

Do small teams need a creative testing framework? Smaller teams need it more, because they can't outspend their mistakes. A one-page log of concepts, hypotheses and outcomes is a framework. The point is that learning accumulates somewhere, instead of leaving with whoever ran last quarter's campaigns.