A guide to Google’s new Shopping campaigns
Within the next couple of months, Google will be transitioning all existing product listing ad (PLA) campaigns over to their updated format, called Google Shopping campaigns. With Shopping campaigns, online advertisers can expect a host of new features, as well as a simplified, retail-centric user interface for managing and optimizing ads connected to their product catalog. On the back end, nothing will change – the product feed you used for your PLA campaigns will still work for Shopping campaigns, and algorithmically, Google has changed nothing about the product target bidding engine and ad unit. They do, however, recommend that campaign managers be proactive and manually convert their product ads to a Shopping campaign as soon as possible, since an automatic upgrade tool will not be available for quite some time. Here is a quick startup guide on moving to Shopping campaigns, so you can take advantage of the added functionality ahead of schedule, and avoid any major bumps along the way.
Creating a Shopping Campaign
Because there is currently no turnkey solution for upgrading PLA campaigns with the click of a button, you will need to go ahead and create a Shopping campaign, which is now a new campaign type of its own:
After managing some basic settings, like connecting the appropriate data feed to the campaign, you’ll notice your first ad group defaults to targeting ‘All products’, with the option to bid, or expand out your selection with a plus button:
One of the main benefits of Shopping campaigns is that you can – and in many cases should – manage all or your product targets from within just one ad group. By filtering product categories out from the top down, rather than bidding on discrete combinations, it’s easy to understand, analyze, and ultimately bid on any number of targets from this one screen. I’ll show you what I mean.
Click on the gray plus icon to bring up your first opportunity to get more granular with your targeting than ‘All products’. Refer to your current PLA campaign structure for an idea of how you want to begin sorting your products at a high level. For example, for one of my hockey equipment clients, I find it more effective to prioritize my bidding at the brand level rather than product category, since certain brands provide greater margins than others for their business, which informs variable ROAS targets across brands. So, my first subdivision of ‘All products’ will be by Brand:
Notice that Google has flipped the onus of targeting and bidding on product targets; with PLA’s, you had to (accurately) type in category labels in order to target a particular segment of your feed, whereas now Google’s interface pulls that information in for you from your feed and presents it in a “click-and-drag” interface. This promises to make implementing new product targets much, much more intuitive for advertisers.
I recommend starting relatively broad at this point, and only selecting a handful of your brands to start. Using this sporting goods client as an example, I’ll add every brand with more than 600 products (SKUs). Obviously, if some of your brands with fewer SKUs are of relatively high importance to your business, then you should break those out as well for maximum control.
Your ad group should now look something like this:
Along with the subdivisions you chose, you’ll see a row labeled “Everything else in ‘All products’” at the bottom. This is the catch-all for every product not represented by the brands above it, so make sure your bid reflects this fact. For now, just be sure you are bidding competitively on your “Everything else” SKUs so that they can gain impressions. You’ll be able to easily go in and pluck top performers out into their own rows once the data begins to roll in – another win for Shopping campaigns over PLAs.
From here, you have up to five more subdivisions you can manipulate (you can think of your ad group as a bulleted list with a maximum of six indentations below ‘All products’). You can parse out your campaign as elaborately as you feel comfortable dealing with right now, but it bears repeating that the best practice is to limit yourself to this one ad group. If the structure of your legacy PLA campaign has worked well for you, it’s fairly easy to emulate it as a Shopping campaign once you understand how the subdivision hierarchies relate to product targets. However, I recommend utilizing your broad “Everything else” categories, and not going too granular, for when you first turn on your Shopping campaign and turn off your PLA’s (don’t worry, your product ads will keep serving!). As you become more comfortable with the new structure and start to have sufficient data in the Shopping interface, building out more complex targets and bids will naturally follow.
Improvements over PLA campaigns
Although I have yet to see just how much the new campaign type will affect my product ads’ performance, there are quite a few things that Shopping campaigns improve upon PLAs right off the bat. One that strikes me as very helpful in particular is that you’ll find that it’s nearly impossible to have conflicting bids on overlapping product targets, since rather than bidding on a group of SKUs, and also bidding on subsets within that group, the bids are set within the group – only at the level you’d expect to be bidding on – and unspecified products are bid on separately. I know that sounds confusing, so here is what I mean visually:
In this instance, the bid for all Bauer products is, in effect, the catch-all group, and its bid is set at 50 cents. Unless careful attention is paid to ensure that the highest-level category has the lowest bid, which can be difficult in larger accounts, then it is possible to run into the situation where a more precise product target is actually being outbid by the less granular bucket – thus giving the impression to the more general group and negating the usefulness of breaking hockey pads out at all.
Here, you can see that rather than competing on concentric targets, Shopping campaigns pull that long-tail group of products out and away from the subcategories you specifically want to bid on, leaving no overlap within the whole of Bauer products, and no conflicting bids. In this way, Shopping campaigns allow you to easily bid lower on a brand’s subcategory than on the remaining unspecified subcategories within that brand grouping, without worrying about interrupting the performance of other product targets of that brand.
Another new and powerful tool boasted by Shopping campaigns is the ability to see benchmark statistics at every tier of your campaign structure, alongside the usual performance data. Google now provides a benchmark click-thru rate, cost-per-click, as well as your current impression share in-line with each product segmentation. These new columns are included to inform campaign managers on how “other advertisers’ Product Listing Ads for similar products are performing,” specifically the relative effectiveness of their product offer or price point, how their bids compare to the market at large, and how much room there is for improvement, respectively:
This peek under the hood will almost certainly affect the way advertisers as a whole approach the product ads auction, and I’m sure we can expect some shifts in performance over time due to the newly available information. Ultimately, however, more transparency is rarely a bad thing when it comes to online marketing, and these features look to become a huge improvement over the basic insights provided by PLA campaigns.
Another great way Shopping campaigns allow you access to the information you care about is in the Dimensions tab – specifically, there are now ‘Shopping’ dimensions that will break down the true performance of your campaign, and show you where in your product catalogue you are actually generating conversions (or wasting money). For example, here is every possible “product_category” identifier in my feed, and the performance of each, viewed from outside the campaign management paradigm:
This same report can be pulled at every level of product identifier, even down to the individual ‘Product ID’ level. It is easy to imagine how informative these reports will be once your Shopping campaign begins to accrue impressions, clicks, and conversions, and I envision marketers being much more confident in their bid decisions after analyzing their product ads performance at this level of granularity.
One feature that I would really like to see brought to Shopping campaigns is the ability to apply flexible bid schedules with ROAS targeting. Currently, AdWords allows you to create these optimization rules around ROAS (based on an X% desired return on ad spend), but there is no way to then use it in a Shopping campaign. Product ads are inherently related to e-commerce, so advertisers would be remiss not to consider their conversion values by product relative to its advertising costs, and having Google’s powerful conversion software working with you towards optimizing that crucial KPI would be an enormously helpful asset here.
If you are still having trouble transitioning your legacy PLA campaigns over to Shopping campaigns, there are plenty of resources in the AdWords Help section on the subject. I have provided some links below with further information, and feel free to ask any specific questions in the comments.