How We Cut Meta Ads CPA to One-Tenth: The Full Story of Testing 120 Creatives
“CPA for our event recruitment stays stubbornly high.” “We want to improve it, but we don’t know where to start.” Anyone in charge of ad operations has felt this at least once.
StandArrow joined a staffing-related service company that had no dedicated ad operator in-house. By running about 120 creative tests and multi-angle PDCA across six axes, we improved acquisition CPA to about one-tenth. This article presents that case.
Case Summary
- Challenge: No one with specialist knowledge to run ads, so no testing framework was in place, and CPA had risen sharply
- Approach: Rigorous testing across six axes: “media, user flow, time of day, day of week, creative and targeting”. Produced and ran more than 120 creatives in total
- Result: Cut Meta Ads CPA by 90%. Cost per event participant also came in well below target
- Key point: Not mere ad operation outsourcing, but consulting that went into user flows such as LINE, apps and landing pages (LPs)
Challenge: The Cause of the Rising Cost Could Not Be Identified
The biggest bottleneck was that no one in the company had specialist knowledge of ad operations. They could not analyze “why we acquired customers” or “why costs rose”, so performance testing was a black box. Operations became intuition-driven, and they had fallen into a vicious cycle in which CPA kept rising.
What We Did: 6 Axes x 120 Creative Tests
1. Searching for Winning Patterns Across “Six Testing Axes”
We broke down the factors that affect ad performance into the following six, and ran A/B tests on each one.
- Media: Selecting platforms with a strong affinity for the target audience
- User flow: Comparing which landing point, among an LP, an official LINE account, a dedicated app and a web form, causes the fewest drop-offs
- Time of day and day of week: Identifying the moments when people are most responsive
- Creative (CR): The messaging of banners and videos
- Targeting: Refining segments
2. PDCA on More Than 120 Creatives
We did not just “make it and be done”. We produced and ran every pattern, changing the angle of the message and the tone and manner of the design while running PDCA at high speed, so that we never ran out of winning creatives.
3. User Flow Optimization Consulting
We also analyzed in depth what happens after the ad click. By comparing and testing the LINE, app and form flows, we identified the route with the highest participation rate.
Result: A 90% Reduction in CPA
As a result of thorough data-driven testing, CPA fell to one-tenth or less of its level before operation began. Notably, this is not a one-off result. More than three years of ongoing support is proof of a solid acquisition setup that holds steady even as market trends change.
Why Did It Fall This Far?
Ordinary ad operation outsourcing tends to stop at “bid adjustments” and “swapping banners”. In this case, we went into the very core of the user experience, the user flow (“LINE or a web form?”), and that led to the dramatic improvement. Most results are determined by the breadth and volume of testing.
If you are “running ads without really knowing why, and not getting results”, the cause is very often a lack of testing. We can take on the testing that is hard to handle in-house, such as producing a wide variety of creatives and designing complex user flows, so please get in touch about building an operating setup grounded in evidence.
FAQ
Q. What was the biggest factor in reducing CPA to one-tenth?
A. Rigorous testing across six axes (media, user flow, time of day, day of week, creative and targeting) and a PDCA cycle of more than 120 creatives. What sets it apart from ordinary ad operation outsourcing is that we compared and tested even the post-click flows such as LINE, an app and a web form.
Q. Is the effect continuing, or was it only a temporary improvement?
A. The effect has lasted through more than three years of ongoing support, and we have built an acquisition setup that holds steady even as market trends change.