You Tested Your Ads. The Numbers Say You Didn't.
You launched three ads. Maybe four. You ran them for a couple of weeks, watched the numbers, picked the one that performed best, and called it your "winner."
That wasn't a test. That was a coin flip disguised as strategy.
New data from Motion's 2026 Creative Benchmarks Report, analysing 578,750 creatives across $1.29 billion in ad spend, reveals something uncomfortable: only about 5% of ads ever become winners. The other 95% either underperform or get killed before they spend anything meaningful.
At a 5% hit rate, launching three ads gives you an 86% probability of finding zero winners. You didn't test a strategy. You played a lottery with three tickets and wondered why you didn't win.
The Power Law Nobody Talks About
Here's what happens inside almost every ad account, whether you're spending $2,000 a month or $200,000.
Common Thread Collective (CTC) tracks a metric they call "Ad Concentration." Across hundreds of accounts, they found that 3.5% of ads drive 66% of total ad spend. That's not a quirk of individual accounts. It's a pattern that shows up everywhere.Sam Tomlinson, one of the sharpest digital strategists writing today, puts it even more starkly in Issue #164 of The Digital Download: "1-10% of creatives account for 80-95% of spend and results." His definition of a "hit" is precise: an ad that produces 25+ conversions in 30 days, spends at least 25x your target CPA, and delivers at or below that target.
This is a power law distribution, the same pattern that governs venture capital returns, music chart hits, and book sales. A tiny number of entries produce almost all the value. Everything else is noise.
The problem? Most Australian SMEs don't produce anywhere near enough creative to find those winners. They launch a handful of ads, pick the least-worst performer, and convince themselves they've optimised.
Why Your Brain Ignores the Probability
Daniel Kahneman, the Nobel Prize-winning psychologist behind Thinking, Fast and Slow, identified a cognitive bias he called base rate neglect. It's the tendency to ignore statistical probabilities in favour of vivid, specific stories.
Here's how it works in practice. You see a competitor running an ad that seems to crush it. You think, "If I just nail the right creative, I'll get results like that." You're anchoring on a visible success story while completely ignoring the base rate: that 95% of ads fail.
The maths is unforgiving:
| Ads Launched | Expected Winners (at 5% hit rate) | Probability of Zero Winners |
|---|---|---|
| 3 | 0.15 | 86% |
| 5 | 0.25 | 77% |
| 10 | 0.5 | 60% |
| 20 | 1.0 | 36% |
| 50 | 2.5 | 8% |
Motion's data breaks this down by spend tier. Micro accounts spending under $10,000/month see a hit rate closer to 3.8%, which makes the probability even worse. Enterprise accounts spending over $1 million/month achieve around 8.2%, not because their creative is inherently better, but because they produce enough volume to find winners consistently.
The top quartile of advertisers within every spend tier doesn't outspend their peers. They launch 2-3x more creative concepts at the same budget level. They don't spend more. They bet more.
Your Winner Has a Shelf Life (And It's Shrinking)
Finding a winner is only half the problem. Keeping it alive is the other half.
In 2024, a strong ad concept could run for six weeks or more before fatigue set in. In 2026, that window has compressed to two to three weeks. For high-budget campaigns targeting narrow audiences, it can be as short as seven to ten days.
We've written before about why your best ad died two weeks ago. The data has only gotten worse since. Meta's own research shows a 45% drop in click-through rate after just four exposures to the same ad. Showing the same creative five times per week to the same audience slashes conversion rates by 30%.
The median ad set now hits its first fatigue signal at day 11 of continuous spend, with CTR already down 38% from launch.
This means creative production isn't a one-off task. It's an ongoing system. You're not just looking for winners. You're building a pipeline that replaces them faster than they decay.
Creative Is the New Targeting
Two years ago, you could get away with mediocre creative if your targeting was sharp enough. That era is over.
Meta's Andromeda algorithm now uses your creative as its primary signal for who to show your ads to. The platform reads the images, watches the video, analyses the copy, and decides which users are most likely to engage. Your creative doesn't just communicate your message. It tells the algorithm who your customer is.
The implications are significant:
- Roughly 60% of ad performance is now determined by creative, not audience selection or bid strategy
- Brands testing 20+ conceptually distinct ads per month see 65% higher ROAS than those running fewer than 10
- When Meta's similarity scoring detects that your ads are more than 60% alike, it collapses them into a single "entity" and makes them compete against each other
This connects directly to Byron Sharp's work at the Ehrenberg-Bass Institute. Sharp's research, spanning 130+ brands across 13+ product categories, shows that brands grow by being easy to think of across multiple buying situations, what his colleague Jenni Romaniuk calls "Category Entry Points."
Different creative concepts reach people in different mental states. An ad that resonates with someone researching a solution won't land with someone who doesn't know they have a problem yet. We've explored this idea in depth with the six buying triggers most marketing misses. Diverse creative isn't just a platform requirement. It's how you build mental availability across the full range of situations where someone might need you.
The Portfolio Approach: What Top Performers Actually Do
Sam Tomlinson has written extensively about applying portfolio theory from investing to ad creative. The logic is straightforward: no fund manager puts their entire portfolio into three stocks and hopes for the best. They diversify across enough positions that the winners more than compensate for the losers.
Your ad account works the same way. Motion's data shows the typical breakdown:
| Category | Percentage of Ads | What Happens |
|---|---|---|
| Winners | ~5% | Scale and drive majority of results |
| Mid-range | 38-46% | Spend modestly, break even or slightly positive |
| Losers | 50-53% | Get killed before day 28 |
The hit rate varies by format, too. Text-only ads win at 11.6%. Product images with text overlays hit at 8.75%. Lifestyle imagery lands at 7.59%. UGC-style creative converts at 7.56%. High-production video comes in at 6.87%.
That last number should make you pause. The most expensive creative format has the lowest hit rate. The cheapest has the highest.
This doesn't mean you should only run text ads. It means the return on creative investment isn't where most people think it is. You're better off producing 15 simpler concepts than pouring your budget into 3 polished ones. The data from System1 and Effie's Creative Dividend research, presented at Cannes 2025, found a 12x profit multiplier between the worst and best-performing ads within the same brand. The gap between good and bad creative is enormous. But you can't find your best-performing ad without testing enough to discover it.
Mark Ritson's qualification here is important: when emotion, brand fluency, and time-in-market are all present, campaigns return up to 21x more profit. But that's for brand campaigns that run for months. For performance creative on Meta and Google, where your typical SME is spending $2,000-$5,000 a month, the probability game comes first. Find the winners. Then invest in making them better.
How This Actually Works for a Real Business
Take a typical Australian service business spending $3,000 a month on Meta Ads. That's around the median for SME digital marketing spend in Australia, where digital ad investment hit $18.4 billion in 2025.
The standard approach: create 3-4 polished ads, spend $750-$1,000 on each, run them for a month. At a 5% hit rate, the odds are stacked against you. You'll likely end up with no clear winner, just varying degrees of mediocrity.
The portfolio approach: create 15-20 simpler concepts. Allocate $150-$200 to each for the first 3-5 days. Kill anything that doesn't hit minimum benchmarks (cost per result above 2x your target). Shift budget to whatever emerges as a winner. Replenish with new concepts every 2-3 weeks as fatigue sets in.
Same budget. Radically different probability of success.
| Approach | Ads Created | Expected Winners | Probability of Finding at Least One |
|---|---|---|---|
| "Perfect ad" method | 3-4 | 0.15-0.2 | 14-19% |
| Portfolio method | 15-20 | 0.75-1.0 | 54-64% |
You're not spending more. You're spending smarter. The constraint isn't budget. It's production volume.
This is why your marketing budget works better as a portfolio than a savings account. Concentration feels safe. Diversification actually is.
What This Means for Your Business
The shift from "find the perfect ad" to "run enough experiments to find the winners statistically" changes three things about how you operate.
1. Production speed matters more than production quality (for initial tests). Your first version doesn't need to be polished. It needs to exist. A rough concept that goes live today beats a perfect concept that launches next week. Test the idea first. Polish the winners. 2. You need a system, not a campaign. Campaigns have start and end dates. A creative system produces, tests, learns, and replaces continuously. Build a process that generates 4-5 new concepts every week, even if they're simple static images or short-form video. 3. Track the right number. Most businesses track ROAS or cost per lead for individual ads. Start tracking your hit rate: what percentage of ads you launch achieve your performance threshold? If your hit rate is 5%, you know you need 20 concepts per month to expect one winner. If it's 10%, you need 10. The hit rate tells you how much creative you need to produce to keep your account healthy.The businesses that win on paid media in 2026 aren't the ones with the biggest budgets or the flashiest creative. They're the ones that understood the probability, built a system around it, and stopped gambling on three ads.
Further Reading
- Motion 2026 Creative Benchmarks Report - Comprehensive analysis of 578,750 creatives and what separates winners from the rest
- Sam Tomlinson: "Creative Is Math Wearing a Costume" - The portfolio theory approach to ad creative management
- CTC: Ad Concentration and Creative Demand Scores - How 3.5% of ads drive 66% of total spend
- System1 + Effie: The Creative Dividend (Cannes 2025) - 12x profit multiplier between worst and best creative within the same brand
- IAB Australia: Digital Ad Spend 2025 - $18.4 billion in Australian digital ad investment
Dream Outcome is an Australian digital marketing agency helping SMEs grow through Google Ads, Facebook Ads, and Email Marketing.