Google recently announced an update to its in-store visits tracking which will allow advertisers to view the estimated number of store visits tied to paid search clicks at the ad group and keyword levels. Previously this data was only available at the campaign level.
This is a huge upgrade for the product, and should help advertisers to better take advantage of these estimates in calculating bids, though the estimates are still limited by traffic significance considerations.
More Granular In-Store Visits Data Means Better Bids for Advertisers
AdWords bids are set at the ad group or keyword levels, and the recent update allows advertisers to bake the expected in-store impact into bids at these levels as opposed to having to apply adjustments based on broad campaign-level impacts.
To understand why this is important, imagine a campaign that contains five different keywords with bids set at the keyword level, and that the in-store impact is different for each keyword.
Prior to the update, if this advertiser were looking to adjust bids based on the in-store visit estimates, they would have had to adjust their keyword level bids for all five keywords by the same amount based on the campaign level impact.
Now with the data available at the keyword level, advertisers will be able to adjust keyword bids based on the measured in-store impact of each keyword, making for smarter bids.
However, data significance will limit how many keywords and ad groups these in-store estimates populate for, as is the case with campaign level estimates.
In-Store Visits Only Populate if Google Deems the Data Significant
Per Google’s documentation, only advertisers with ‘thousands of ad clicks and many store visits’ are eligible to track store visits.
Even for those advertisers that do qualify, however, specific parts of their accounts may never register any in-store visits with this tracking. While in some cases this is because those parts of the account aren’t driving users to stores, in many cases it’s because Google doesn’t have enough data to provide statistically valid results, and thus does not populate store visit estimates.
For example, in Q1 of this year one advertiser saw a 28% lift in conversions driven from PLA clicks when adding in-store visits to single device online conversions, but no lift at all for non-brand text ad campaigns. While it could be that text ads just weren’t driving store visits, it’s more likely that the data was insignificant for the text ad campaigns as traffic is divided up between many more campaigns than is the case for PLAs.
Similarly for keywords and ad groups, this data is only going to populate for those that have enough data for Google to feel confident that they are reporting on real in-store traffic.
Thus, keywords and ad groups that have zero in-store visits attributed to them may still be driving store visits, they just aren’t driving enough tracked visits that Google will populate the data in AdWords.
As such, it may make sense to look at the overall impact of paid search clicks in order to inform how much an advertiser might want to adjust bids for those keywords and ad groups for which there are no in-store visits attributed. In this case, one key bit of segmentation that would be worthwhile to consider is what the in-store impact is for brand vs non-brand, as brand ads tend to drive more users in-store than non-brand.
This is a great update to Google’s in-store visit tracking that will certainly help advertisers assign offline visit impact at a more granular level than was formerly possible.
While data significance impacts how well this data populates, perfect in-store attribution isn’t a possibility for any tracking tool at the moment, and Google will likely only get better over time at ensuring the data is as accurate and complete as possible.