If this doesn’t convince you to always test new ad types, I don’t know what will.
One of our biggest ecommerce clients has historically had great overall performance and ROAS using traditional link ads in Facebook’s News Feed. In June, we wanted to see if we could extend that success even further by testing Facebook’s multi-product ads.
Over 10 days, we spent about $11k on each ad type and got the following results with MPAs:
–40% cheaper CPCs
–75% higher CTR
–26% cheaper CPA – from $22 to $16
–45% higher ROAS – from 142% to 207%
There are a few best practices we suggest based on our test:
Use the Facebook creative rotating tool.
In our case, we were optimizing around a purchase. Facebook’s auto-optimize feature rotates the multi-product ad so that the creative that is getting the most conversions will be the first in the ad’s series.
For some clients, this would be the wrong way to go if you’re using MPAs to tell a story with creative that requires a specific sequence.
For example, a subscription service might have Step 1: Register, Step 2: Select Your Subscription Type, Step 3: Get a New Box of Goodies Each Month, etc.
Shorten the funnel.
Send users directly to the product that you’re showing in your ad. Typically, we’ve had success sending people to a broader category page in order to let them shop around, but with this test, the MPAs really killed it when the destination was the exact product we showed in the ad’s image.
Mobile, mobile, mobile.
Desktop didn’t work well (for shopping, who’da thought!), so don’t forget to split your tests by device type – test mobile phones vs. tablets at the mobile level as well.
We’re also doing more tests like this for our clients in the EDU space with great results, so we look forward to sharing more stories like this in the future as well!