We’ve been hearing that consistently since Google Shopping switched to a fully paid marketplace in October. This channel is suddenly akin to the early days of paid search where you could likely generate positive ROI just by turning it on.
Despite the good ROI seen today, as more competition flows into the channel we will certainly see CPCs rise and quality score become a stronger arbiter of positioning.
In order to compete and drive strong ROI, you will have to run your PLA campaigns much like you run your SEM campaigns today. If you take an active optimization role, there’s an opportunity to have PLAs be one of, if not the best, revenue driver site-wide.
Even though it might seem like a convenient strategy today, beware the feeds provider that tells you they are qualified to manage your PLA campaigns because they do it for Shopzilla and PriceGrabber.
The techniques below are most familiar to expert SEM practitioners and can drive significant improvement over baseline PLA performance.
Isolating Top SKUs
Most retailers have a core set of products that represent a significant percentage of their revenue, and help the brand stand out from the competition.
On their website, on their shelves, and in their advertising and promotions, these products are handled separately due to their importance to the business. This should be no different in PLAs and these products should have their own bid strategies, keyword targets, and promotional text so that profitability on these SKUs is maximized.
This is simple to execute on within the AdWords interface, just create Product Targets for each individual SKU that you’d like to merchandise separately from the others. Once you have that Product Target created, you can bid that product on its own as well as create negatives to ensure that it doesn’t get clicks on irrelevant queries.
Semantic Organization of Products
For a retailer with thousands or tens of thousands of products, it would be challenging to manage each of these SKUs individually, especially since many of them would not get enough volume to make an accurate bid.
However, the relationships between the products in a portfolio are incredibly important, in that they will behave similarly in a dynamic marketplace. Some examples of this would be products of the same brand, of the same attribute value (the color blue), or the same margin profile. Grouping these products together into product targets makes a lot of sense and allows for more effective control over the performance of the entire portfolio.
One example of this would be to group your products by category and brand. We know that JBrand Jeans and Levi’s Jean Shorts have wildly different price points, customers, and margin contributions. They should live in their own groups, not under the broad heading of Denim.
Aligning Your Feed Content to User Intent
PLA is a paid search channel and not a CSE (comparison shopping engine) channel. Products are displayed based on keyword level intent signals, however most retailers’ product feeds are designed for online/offline merchandising techniques, which don’t take searcher behavior into account. The content of those feeds are consistently misaligned with how users search for products on Google and therefore a lot of potential volume is left on the table.
There are ways to gently manipulate your product feed such that information light merchandising titles like Canon EOS Rebel T4i 18MP Digital SLR Camera with 18-135mm EF-S Lens – Black can be converted to Canon EOS Rebel 18MP SLR. This much better reflects how someone might search for this product on Google.
One truth in the paid search world, and similarly in PLA, is that the more control marketers have over the traffic they receive from AdWords, the better they will ultimately perform.
Google makes money through creating liquid marketplaces for each search, and they seek ways to intelligently increase coverage for their advertiser network. Many times this works to the advertiser’s benefit but often it does not, making it critical to include negatives in PLA at the product target level in order to ensure relevant traffic and to not compromise ROI.
If you have a Product Target for Red Sneakers, it would make sense to have negatives for the other primary colors as well as the other types of shoes that you are merchandising through your other product targets.
Image & Title Testing
Google doesn’t support ad testing for PLAs, but that doesn’t mean that marketers shouldn’t explore ad testing in order to drive performance. The most visible areas to test ads in PLA would be the product title and the product image.
There are a number of different ways to do this through testing different strategies either week over week for the same traffic, or at the same time for similar traffic.
For example, if you have two different styles of images on your site, one with the clothing on a model and one with it off the model, you could test all your products one week with the model shot and then rotate to the non-model shot for the next week. You should get a good sense of how clickable the different images are in Google’s marketplace from this test.
Keep Tabs on Your Margins and Competition
Much like in SEM, you don’t want to spend your precious ad dollars trying to compete for placement when you’re margin constrained or not price competitive. There are a lot of intelligent tools in the marketplace today that can help you understand your pricing relative to the competition.
If you aren’t competitive, it doesn’t make sense to bid high for position against a certain product or product type. On the flip side, if you are the price/volume leader on a specific product, you should bid very aggressively and continue to establish your leadership position.
As this channel continues to grow, it will become increasingly important to actively optimize your PLA campaigns. Once your PLAs are live in AdWords, you should start adopting these techniques to continue to stay ahead of the competition.