One Weird Trick to Improve Your Remarketing

Remarketing has come a long way. Just a few short years ago, marketers often needed to make a business case for the incrementality of remarketing (i.e., were you paying for users that would come back anyway or were you increasing sales?). Thankfully remarketing has won often enough that we don’t need to do this anymore.

We’ve gone from static banners stalking your every move across the internet (frequency caps are your friends) to a far more sophisticated offering across display, search, and social channels. The ability to segment users has greatly improved as well as the ability to message those segments differently. However, there is one particular segment of users that isn’t traditionally addressed well by standard remarketing efforts. Some simple changes may drastically improve the performance of your campaigns.

First, a quick primer on remarketing best practices

The general recommendation is to target by funnel depth (i.e., which step in the conversion process is the site visitor). For an ecommerce company, the funnel stages often look something like home page, category page, product page, cart abandoner, converter. Then exclude converters and progressively bid up for users based on how far into the funnel they’ve progressed. This makes intuitive sense; the further through the funnel, the closer to purchase, and therefore, the more likely to convert.

However, I’ve consistently seen bounced users, regardless of funnel depth, perform significantly worse than all other previous visitors. This too makes sense if you think a bit longer about how traffic acquisition works. No site has all of their traffic start at the home page. Equating funnel depth (which step a user is on) to funnel progression (how many steps a user has taken) is intuitively flawed. If your site has a robust SEM and SEO strategy, the majority of your traffic may start on a category page or product page. If that’s their starting point, and they take no further actions on your site, then you probably don’t have what they’re looking for and you shouldn’t pay for them to come back. In extreme cases, excluding bounced users can save over 20% of remarketing spend with a less than 5% decrease in remarketing-assisted transactions.

In extreme cases, excluding bounced users can save over 20% of remarketing spend

How to alter remarketing strategy

My weird trick, then, is to exclude bounced users across all remarketing lists, tout the ROI improvement , and call it a day. In some cases, it may also be appropriate to exclude bounced users across all marketing spend to prevent them from becoming a “new” user again. There are so many better ways to spend your time and marketing dollars than to fight for customers that have already signaled a lack of interest in your product by bouncing. However, this trick could lead to an unacceptable loss of traffic and require a more nuanced approach – particularly if your site has a high bounce rate (something to investigate regardless).

How to alter remarketing strategy, part 2

To start, make remarketing lists by funnel depth that include only bounced users. Then, avoid all dynamic remarketing options. I’ll explain this approach with a silly story. A long time ago, for reasons I can’t remember, I came across a product listing that had a ridiculous-looking tiger clock in the image, but the product listing was definitely not for a tiger clock. It was just a broken listing. Being a digital marketer, I shared it with colleagues—then that listing proceeded to stalk each of us all over the internet. While we got a good laugh and didn’t re-engage with it, something not too different could affect you and cost you money. Imagine you’re a clothing retailer and some person bounced on one of your dress pages. The reason they bounced was that you were out of their size. While you can’t know that explicitly, the fact that they bounced tells you something wasn’t right. Now, if that dress is in a Facebook carousel ad, they may still click on it, not realizing it’s an ad from the same company as before. And you still don’t have their size so you’re just burning money.

My general recommendation

Target and message a level up from the users’ previous max funnel depth. So product page visitors get sent to a category page where they may be more encouraged to explore and engage. Category page visitors get sent to the home page to see if they’ll engage in a new way. Home page bouncers are a unique case. It’s most likely that they arrived directly or via a branded marketing effort. Allowing these users to fall into standard marketing efforts is generally fine, as it’s unclear what value proposition you’d use in remarketing. If significant portions of home page traffic come from non-branded channels, that’s something that has to be addressed on a case-by-case basis. By changing up what these users see, you improve your chances of turning them into customers. Worst case, if this doesn’t improve performance sufficiently, you now have a much better argument to stop spending money trying to acquire them.

By changing up what these users see, you improve your chances of turning them into customers.

The biggest takeaway I want to convey isn’t just that not all previous visitors are created equal, it’s that some previous visitors could be worth significantly less than even new visitors. And adjusting your remarketing accordingly could yield immediate wins for the profitability of your marketing efforts. Plus, you’ll look like a genius if not playing by the book works out. And who doesn’t like that?