If You Can’t Be A Good Example, Be A Horrific Warning: ECs + CO
Here, I think, I will be a bit of both.
Part I: A Horrific Warning
So the story begins…with an enhanced campaign migration. In our legacy campaigns, we were running a display campaign on Display Conversion Optimizer (DCO) using CPA targets on desktop only. After migrating to Enhanced Campaigns on Google, we were forced to run on both desktop and tablet, and started to see something concerning within the first couple days post-EC migration. While our desktop CPL had, for the most part, remained the same, the tablet CPL was quickly approaching a $1,000 CPL value. Our first reaction: WHAT IS GOING ON!??!
After speaking to a Google Enhanced Campaign specialist, his first thought is that it could possibly be a bug because the difference in CPL was so high between desktop and tablet, and the CPA targets were being vastly exceeded. Particularly because the DCO system should be shifting traffic away from tablet traffic since its performance was well below sub-par.
After looking into this a bit further and deep diving with more EC specialists and Google reps, we found a couple things that may be occurring:
- The CPA value for the display campaign was set specifically for desktop computers. Once we migrated to ECs, tablets were introduced, which the campaign had no history for, so DCO was still adapting to the presence of tablet traffic within the campaign. This was causing budget to be spent on poor conversions.
- The algorithm is still attempting to drive to the desired CPA target and is currently treating tablets as a new space and exploring available placements. Once the CPA targets start to take effect with the addition of tablets, traffic will become limited to tablets since their CPL values are so much higher and less efficient than the target, and the traffic will once again shift towards desktop computers.
- However, while DCO will be adjusting for tablets and desktops, it will probably take about 4 weeks.
So the question remains – what if you don’t want to pay for the system to figure out how to become effective again?
Part II: A Good Example
Here are a few things we’ve learned to help keep your costs down and to have a little more control over your DCO campaign.
1) First and foremost, if you haven’t implemented CO yet on your campaign and want to play it safe, wait at least 4 weeks until you enable it. On the flip side, if you have it enabled on a legacy campaign, you could disable CO before you migrate to ECs and wait 4 weeks before you re-enable it. This way, the campaign can gain some historical performance with the addition of tablets and help Conversion Optimizer make better, more informed decisions.
2) Secondly, if tablet conversions are a viable option for your particular client, then ensure that conversion tracking is set up properly on all possible conversion paths. If conversions on tablets are not a possibility, then disable conversion tracking for tablet users so that the higher expense will trigger DCO to reduce traffic further. This should make the algorithm stop trying to force a conversion based on tablet traffic and eventually lead to making tablet traffic virtually non-existent.
3) If you’ve migrated a legacy campaign to Enhanced and already had CO implemented, you are doomed. Just kidding! Try limiting your campaign budget until the tablet CPA value is low enough that you can begin to increase your budgets in increments. Also try optimizing and lowering your CPA targets to move traffic away from the higher CPL devices (tablets) a little more aggressively. You can do this in addition to disabling conversion tracking for tablet users as described above (2).
Part III: The Moral
And there you have it. Not everyone will experience this same issue when migrating a legacy CO campaign to EC, and for those that do, this is not panacea; but it is a good start on being a more strategic and having a little better idea of issues you may want to be prepared for during migration. And hopefully you’ve learned a valuable lesson from this tale of woe.