It’s happened to all of us. You finally found a great deal on a gift for your significant other, and pulled the trigger. The package is on its way from Amazon, Wish, or some other retailer. You open another browser window to check the news, and are surprised to see advertisements for the exact same object you just purchased flooding the screen. As you’ve already purchased the gift, you rightfully ignore these advertisements and go on with your day, slightly miffed at the obliviousness of advertisers.
If you run an online marketplace, there’s a good chance this is happening to you regularly. Ad relevancy is a powerful tool when used properly, presenting users with options similar to items that they have already purchased. The intent is to encourage a click by advertising items the user has already expressed interest in, either through a click on an ad or a purchase on a website. However, the efficacy of this approach changes once the user has actually completed the purchase. In this post we’ll take a brief look at ad relevancy. We’ll explore some of the pitfalls of using relevancy naively, and discuss how we can improve relevancy – and click-through rates – with a slight shift of focus.
Stating the Problem
If you are running an online marketplace, you are very likely participating in a number of ad networks spread around the web. These ads, if properly implemented, should be driving clicks on your products that result in higher traffic on your marketplace – and accordingly, higher sales. The key to success of an ad campaign in this style is to present potentially relevant items to the end user, encouraging them to make a purchase based on items they’ve already expressed an interest in.
The most common approach here – what we’ll call the “naive” approach – is to present the user with products similar to those they have already purchased. If a user has purchased a red dress, for example, then it makes sense to conclude that they are interested in purchasing more dresses from online marketplaces. The user continues to browse the web and, in accordance with the above, they promptly see advertisements for red dresses in a similar style, blue dresses, green dresses, and other similar big-ticket items that they’ve already demonstrated a willingness to purchase. All is well in the world. Well, except for your click-through rates.
The Problem With the Naive Approach
The problem with the naive approach above rests in a faulty assumption. The base assumption is that if a user has purchased one item, they are likely to be interested in – and purchase – similar items. Someone who purchases a red dress may also be interested in another red dress, or may be interested in a blue dress with a similar neckline. While this approach seems logical at a glance, it ignores user behavior.
Put yourself in the shoes of a user shopping for a red dress. The user likely already has a closet full of other clothes, and is purchasing the dress to bolster their existing collection. While it is likely that they may purchase additional dresses in the future, odds are that they’re not interested in purchasing a big ticket item like a dress in bulk. This is a difference between the act of shopping, where you want to be presented with a range of similar options so you can make a decision, and the act of buying, which is driven by the underlying purpose that drove you to shopping online in the first place. The naive approach above optimizes for the action of shopping, but ignores that most of these ads are not being seen on your marketplace – in fact, once the user has clicked through to your marketplace, they no longer need to have the related products shown to them as ads. They’ve expressed an interest in purchasing this item from you, which means they’ve necessarily shifted their mindset from ‘purchase” to “buy”.
Improving Clickthrough Metrics
Once the user swaps their mindset from “purchase” to “buy”, the naive approach to advertising above immediately loses its efficacy. The user has just bought a red dress – there is not much of a chance they need two red dresses. The key here is to recognize that the switch from a “purchase” mindset to a “buy” mindset changes the desires of the shopper. They’ve completed their primary goal – buy a red dress – and have moved on to the original driving goal of building a wardrobe that includes a red dress.
To get around this misunderstanding, you’ll need to expand your definition of related products. A user who has bought a red dress likely doesn’t want another red dress. However, they probably are very interested in a red handbag, red shoes, or any other item that would be fashionable with their new garment. Bolstering this is the place where the user is actually seeing your advertisements. Once they’ve purchased an item from you, they’ve probably already left your website to continue their day. This means they’re seeing your “relevant” ads on other websites, either browsing the news, searching the web for additional products, or while checking in with friends and family. By changing the ads at this point to show items related to the red dress, you can once again capture their attention and drive them back to your marketplace.
Simply put, once a sale has been made advertising that shows similar items to the user loses all efficacy. The user has made their purchase, and moved on to other things. If you want to capture that user’s attention, you’ll want to present the user with items related to their purchase, but not similar. Presenting someone who has purchased a red dress with more red dresses will not be effective. However, presenting that same user with shoes, handbags, earrings, and other related items is far more likely to result in the additional click. By focusing on user behavior, and the factors driving their purchase decision, you can revolutionize your advertising and improve clickthrough rates, providing ads that are more relevant to the user and more beneficial to your bottom line.