When assessing the success of your paid search campaigns, it’s easy to treat ad performance as a proxy for product performance. After all, if your t-shirt ads are performing well, that’s probably a good indication that t-shirts are in high demand, right? Maybe not as much as you think.
Using an internally developed click-to-cart report that examines product ad clicks and their attributed orders, Merkle puts a microscope over the order behavior of our retail customers. Google makes it easy to track item IDs associated with Shopping ads using ValueTrack parameters. Additionally, if text ad keywords are granularly categorized, one can simply compare the type of products purchased to the category of the keyword to assess differences.
Using this method, we ran the numbers on over 36,000 orders from some of our highest grossing paid search programs to find out how often shoppers were really buying what they clicked on.
What we found was surprising; one out of every five users that clicked on an ad of a given product category ended up ordering an item from a different category. Some advertisers find even larger shares of orders attributed to categories other than that of the product clicks, and the implications of this should affect how you’re spending your ad dollars.
Taking the right steps to account for these click-to-order discrepancies could help make the difference in taking your paid search accounts to the next level. So, what should we be doing if our customers aren’t buying what they’re clicking on?
Bid to the product, not the ad
Hypothetical: let’s say you’re an apparel retailer and want to push bathing suit products as you approach the dog days of summer. Sounds easy enough - just push bids for ads tied to bathing suit categories, right? But what if only 60% of bathing suit ads result in a bathing suit order? Should you push bathing suits at that point?
This is where Merkle’s advanced approach shows its value. We can selectively push categories based on the frequency they convert on the same category as the click, as well as identify categories of product ads/keywords that are likely to end in conversions for a different, targeted product category. Knowing that 32% of sandals orders, or 46% of intimates orders result in an order for bathing suits introduces other areas of focus and more opportunity to capture demand.
Group likes with likes
In basic terms, we like to group similar products together for bid calculation because they perform comparably. Obviously if an individual product/keyword has enough data to set a bid specific to just that product then we don’t need to group it with other similar products, but for products/keywords with less data we must aggregate performance in order to set effective bids based on meaningful data.
For example, dresses will generally rise in demand during the spring and summer, whereas outerwear gains traction in the fall and winter. So, it follows that we would group and bid the products and keywords with insufficient data tied to these segments separately.
That isn’t to say that all dresses perform the same; sweater dresses and sundresses perform differently depending on the time of year, so we opt to house those in separate product groups.
Click-to-cart analysis arms us with greater insight into true product performance, and allows us to think critically about how we are structuring campaigns and grouping products and keywords for bidding.
You might also like…
Retailers will often prompt users with suggestions for other products on their product pages. Being informed about what products customers are buying when they click on ads can help you recommend the right products, and drive additional revenue from the order.
Our hypothetical apparel client might not know that their sandals ads are often bringing users to the site to purchase bathing suits. Being equipped with that data opens the door to potential changes in product recommendations, site navigation structure, and product categorization.
Leveraging matches with promotions
If we’re seeing frequent click-to-order matches between categories, or certain products in particular, why not capitalize on the demand and promote them together? Whether it be on social media through lifestyle imaging or with bundle promotions, clients can leverage their matching data to push the best products to the right people.
Conclusion
Comparing the products clicked on to the products purchased enables retailers to look at product performance from a holistic view, and take necessary steps to react to consumer behavior. We advise clients run this analysis at least once per month to account for any seasonal changes in customer behavior. Doing so will help to ensure customers and data are the driving factors behind the Shopping decisions we make.
Authored by:
