This is the subhead for the blog post
Recently our team met with a client that had a pretty daunting request: they wanted us to completely change all of their high-performing, evergreen ad copy to promote a gift card with purchase offer. In addition, they wanted to do this twice, to create a sense of urgency. So, the first set of ad copy had “Offer Ends 5/27!”; the next set had “Offer Ends 6/10!”
Historically, I had been against changing all the copy, preferring to go the safer route of rotating in one piece of promotional ad copy. But, I didn’t have any actual proof that it would be problematic and it’s a request that we get a lot from ecommerce clients. One scenario I thought likely is that there would be a brief decline is Quality Score, which would then climb back to the account average as the new ads got good CTRs. So, we agreed that we would run it as a test – after all, maybe the increase in revenue would be worth any negative impact to QS.
Unfortunately, we found that completely changing ad copy resulted in dramatically lower QSs that didn’t recover quickly (within our two-week time frame). Additionally, the promotional offers drove higher CTR, but not a higher CVR, meaning that we spent more for the revenue we got.
I did find it interesting that launching the second round of completely new ad copy did not have a second round of equally negative impacts to QS, which dipped low the first time and simply stayed low with the second set of new copy. (The account appeared to have reached equilibrium around 5.8.) I was also surprised by how quickly QSs went up again once we set the evergreen copy live. Within a week, QS rose to 6.4, much closer to the historical average for the account.
Lesson learned: replacing brand copy with promotional copy across the board is not a good idea. The client agrees after this test that we will now rotate in promotional copy, but keep our high-performing evergreen copy live to mitigate the negative impact.