3,500+ Creatives Tested: How We Find What Actually Works
Creative is the biggest lever in paid social and most accounts barely pull it. Here is how we test at scale to find winners instead of guessing.
The Challenge
This one is not a single client. It is how we run paid social across every account, because the same problem shows up almost everywhere.
Most accounts rely on too few ads, test too little and never turn their results into clear creative learnings. They have plenty of data and almost no direction, so the same average creative keeps running and the account plateaus.
Too few ads per account, occasional or random testing and no system for turning what worked into what to do next.
How We Run It
We treat creative testing as the engine of the account, not an afterthought.
Test across every variable that moves the needle
Hooks, formats, offers, audiences and landing page angles. Not just swapping an image, but testing the things that actually change performance.
Test per platform, not one size fits all
Meta, TikTok and LinkedIn each reward different creative. We test for each rather than recycling the same ad everywhere and wondering why it flops on one of them.
Scale the winners, bank the learnings
Winning creative gets the budget, losing creative gets cut and every test feeds the next one so the account keeps getting smarter instead of starting from scratch.
In paid social, the account with the best testing engine wins. Disciplined volume beats clever guessing, every time.
The Results
Across managed accounts we have tested more than 3,500 creatives. That is the difference between an account that guesses and one that knows what works and why and can scale on purpose.
| Metric | Typical account | How we run it |
|---|---|---|
| Number of ads | Tested at scale | |
| Creative decisions | Data led, winner backed | |
| Testing | Systematic across hooks, formats, offers | |
| Platform approach | Tailored to Meta, TikTok, LinkedIn |
Key takeaways
Creative is the lever.
In paid social, creative moves performance more than almost anything else. Under testing it leaves money on the table.
Volume of tests beats guessing.
You do not need to be clever if you test enough to let the data tell you what works.
Each platform is different.
The same ad will not win on Meta, TikTok and LinkedIn. Test for each.
Bank the learnings.
A test is only worth running if the result feeds the next decision.
Notes Portfolio wide, not a single client. The 3,500 plus figure is cumulative across managed accounts and reflects the combined team track record, much of it before Elev8d (founded 2025). Delivered by the team now at Elev8d.
More Case Studies
View all →-
SEOLOCAL HEALTHCARE · GOLD COAST
7+ yrs
Holding the #1 Google ranking
How a Gold Coast Psychologist Held #1 on Google for 7+ Years
Needed long term organic visibility in a competitive local health market. Ranking (core term)Page 3→#1, held 7+ yrs View case study → -
GOOGLE ADSFintech · Australia Wide
90%
Lower cost per acquisition
How a Fintech Outranked the Banks on a Fraction of the Budget
Up against major banks outspending the fintech four to six times on the same keywords, with conversion tracking firing on the wrong actions. Cost per acquisitionHigh and rising→90% lower View case study → -
GOOGLE ADSB2C ECOMMERCE · AUSTRALIA WIDE
6x+
Return on ad spend
Profitable at Scale: 6x+ ROAS for a High Ticket Store
Spend was driving traffic but returns were inconsistent, with profit riding on one or two campaigns while the rest leaked budget. Return on ad spendInconsistent→6x+ across multiple campaigns View case study →
Running the same three ads and hoping?
Most accounts under test creative and wonder why performance plateaus. We've tested over 3,500 creatives across Meta, TikTok and LinkedIn because guessing is expensive. If your paid social feels stuck, let us look at what's actually running.