As PPC marketers, we have a dizzying array of tools at our disposal. With the holiday season ramping up, we are often asked for our recommendation on how to maximise all opportunities and drive the best results possible during this key time of year. And for e-commerce advertisers going into the 2020 holiday period, one of our top recommendations is to make the most of Google Shopping campaigns and if you’re not already using it, test out Smart Shopping campaigns.
Dubious about Smart Shopping campaigns? We were too. Mourning the loss of control? We feel it too. Does it really work as well as Google says? From what we’ve seen, yes. Google has said, and we have found with our own accounts, that on average advertisers see a 30% (or more) increase in conversion value when using Smart Shopping campaigns. Smart Shopping has definitely come out of its infancy to become a mainstay campaign for this holiday season.
And so finally, the big question. Will you be missing an opportunity if you don’t use this campaign type? While success depends on a few factors, for the overwhelming majority of e-commerce advertisers our answer is yes.
So you’ve decided to give a Smart Shopping campaign a try this year. What next?
Decide whether you want to test a Smart Shopping campaign for a selection of products or for the full product catalog. If you’re testing on a selection of products, it’s advisable to choose product groups with high conversions (20+ conversions in a 15-day period) to give the algorithm as much data as possible.
Smart Shopping will take priority over any traditional Shopping campaigns, so whether you’re advertising on a selection of products or on the full product catalog, ensure you’re pausing those products in any existing traditional Shopping campaigns.
It is also a good rule of thumb for Smart Shopping campaigns to avoid unnecessary campaign segmentation. The machine learning algorithm generally benefits from having more data available for optimization so segment the campaigns only when it makes sense for a particular business outcome.
We recommend using a realistic target ROAS – something like the ROAS from the last 30 days for the selected products that you’ll be testing.