Creative is the new targeting: Solving signal loss with adtech

The collapse of targeting data forced the real performance lever to the surface: creative quality.

Over the past 5 years, 60-75% of the targeting signals advertisers relied on have vanished. Apple's App Tracking Transparency killed cookie consent for most iOS users, with only 25-35% opting in (a 70-80% collapse in mobile signal). Google Privacy Sandbox is replacing cookies with less precise alternatives, while GDPR and California's CCPA sanded down data collection across entire continents.

Roughly 50% of digital ad impressions are already served in cookieless environments. The signals you built your strategy on? They're gone. And they're not coming back.

But here's what the data shows: creative quality was always the real performance lever.

Creative accounts for most of your performance

Nielsen's research found that creative accounts for roughly 60% of campaign success variance. Targeting, media mix, reach, frequency: they split the remaining 40%.

Meta's internal testing backs this up. Their top 5% of creative variants perform 5-10x better than median creative, even when served to the same audience. Same people, same targeting, wildly different results. The creative was the variable.

A single ad contains 200+ measurable features: headline length, color palette, font choice, image composition, call-to-action placement, sentiment, product framing. Each one shifts performance. The combinations are practically infinite.

Dynamic creative optimization (DCO) proves this at scale. Platforms running creative testing see ROAS lifts of 15-40%. Some report 3x performance jumps, 40% faster approval cycles, and $1.6M+ in annual cost savings per client. These aren't theoretical numbers; they come from production data.

Signal loss made creative the only lever left

Before signal loss, targeting was cheap labor. You could throw a mediocre ad at a laser-targeted audience and still hit your number. Audience intelligence did the work. (Or at least, it felt like it did.)

That's finished.

The Association of National Advertisers estimates 25-40% of digital ad spend gets wasted. That comes to roughly $27B globally. The ISBA's programmatic supply chain research shows similar waste patterns in programmatic buying specifically.

Most of that waste isn't about targeting the wrong person. It's about showing the wrong creative.

Creative intelligence cuts through that problem. Studies show it can reduce wasted impression spend by 10-23%. For a $10M annual budget, that's $1-2.3M recovered. For a portfolio, that becomes a C-suite conversation fast.

Good creative travels. It performs across audiences, placements, and geographies. It doesn't depend on knowing someone's browsing history or email address. It converts strangers. Bad creative doesn't convert anyone, no matter how precisely you target it.

Where most advertisers still get this wrong

Most companies spend 80% of their strategy time on targeting and 20% on creative. If creative drives 50-75% of performance variance (which the data backs), that ratio is exactly backwards.

The old playbook: define your audience, buy the impression, serve the ad. Creative was the last thing you decided. You had a brief, a deadline, and 3 rounds of feedback from people who weren't creative experts.

That model is now actively expensive. Flipping the ratio toward creative testing, variant generation, and performance feedback loops is the real lever. Not a better pixel. Not a cleaner data layer.

The good news: the infrastructure to do this well already exists. Here's how to use it.

What works now

Measure creative performance separately from everything else. Think with Google and Warc's research on creative effectiveness both emphasize that creative attribution is nearly impossible if you're confounding it with audience, placement, and channel effects.

Test variants at volume. You can't find your top 5% of creative if you're only testing 3 versions. You need tens or hundreds. eMarketer's programmatic research shows that companies running 50+ creative variations per campaign see 2-3x better performance than those running 5-10.

Close the feedback loop fast. What performed? Why? Feed that back into your next round of creation. The compounding effect of 4-6 cycles of test-learn-create beats single-round optimization every time.

The IAB's research on programmatic supply chains also highlights that transparent creative performance data (knowing what worked and why) filters out waste faster than any audience signal ever did.

Signal loss was supposed to be a crisis. For companies that invested in creative intelligence, it's become a moat. They're deploying higher-quality creative faster, testing at scale, winning efficiency wars.

Your infrastructure should be built around creative quality, testing, and measurement. Because that's where the signal actually lives.