ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

3 mistakes that prevent retargeting success

3 mistakes that prevent retargeting success Jarvis Mak

Retargeting -- it can produce incredible ROI. But you have to retarget the right way. And the "right way" involves a number of variables. How do you use your marketing budget to fill the funnel to ensure it delivers the optimal number of high-quality leads/purchases/desired outcomes? You know that an attribution solution would be ideal to solve this problem. Unfortunately, it's not easy to implement, and you may have internal stakeholders objecting to the black-box nature of an attribution vendor's algorithmic output.

Instead, you decide to segment your approach using the classic funnel divisions.

3 mistakes that prevent retargeting success

You have specific partners handle specific parts of the marketing funnel. You have search marketing and a dedicated retargeting partner at the bottom of your funnel to harvest demand that you have generated. And you work with separate display-advertising companies as prospecting partners, generating that demand at the top and middle of the funnel.

For agencies that split retargeting into its own bucket, avoid the following three critical mistakes when measuring the success of those efforts.

Having blinders on with a single retargeting partner

Why this is a mistake: Retargeting isn't all created equal. It's easy to retarget without technology, and it's especially easy to execute poorly. Visitors to your site can get bombed with far too many ads. They may be shown ads at a time when there is no way they will convert. Done poorly, vanilla retargeting can actually damage your brand's perception among your most valuable prospects. The problem occurs because once a marketer starts doing retargeting, the results seem so efficient that the marketer assumes it's not necessary to try another partner. Don't make that mistake.

What to do instead: First, conduct an A/B test between the incumbent and a challenger once every three or six months. It's important to have clean A/B tests. Second, be transparent -- let each one know how the other partner performs, but provide honest feedback as well. Unrealistic goals undermine trust between marketing partners.

How to do it right: Lay the groundwork for the A/B test via one of these ways:

  • Ideally, utilize tag-management software (e.g. BrightTag or TagMan) that splits pixel fires on your site into separate retargeting pools between your incumbent and challenger. If you're concerned about too much disruption to your current plan, execute a test that can carve out as little as 25 percent of the pixel fires for your challenger. Some solution providers can also set the rate of pixel files without the implementation of a full-blown tag-management solution.

  • If splitting pixels still isn't feasible, then conduct a geographic split for the A/B test. For example, splitting a long list of ZIP codes or DMAs that ends up being comparable from a lead-generation and revenue perspective.

Evaluating prospecting (non-retargeting) partners against a bottom-line CPA goal

Why this is a mistake: If prospecting partners are judged on a CPA basis but aren't allowed to retarget, then the only actions for which they get credit are first-time landers that purchase then and there. If the prospect doesn't convert the very first time on the site, the retargeting partner then takes over. How often do you go to a brand-new site and purchase the very first time you visit? Probably rarely. In fact, more formal research reveals that less than one-third of all site visitors transact on the first time. The consequence is that a CPA goal encourages prospecting partners to drive a different kind of traffic than what you actually need for your marketing funnel. What you want to reward is the driving of quality traffic at the top and middle of the funnel, not first-time landers that convert.

What to do instead: Judge prospecting partners on a cost-per-qualified-lander (CPQL) basis. The key term is that they are qualified. Otherwise, you waste your retargeting budget on people who won't ever convert.

How to do it right: Identify actions that mark an eventual customer (or lead) as being a qualified lander on your site. A simple example would be site visitors that reach a product details or a search results page. Once they've started browsing products or searching for deals, you've got them interested. Use web analytics to analyze paths to conversion -- identify key actions that precede eventual desired outcomes. The most important actions are highly correlated with those outcomes. However, you want actions that are present in a customer's initial visits to your site and may not be present during the same session as the conversion. Like product details or search results pages, other examples of predecessors of eventual conversions might be visiting the FAQ or clicking on a "learn more" button.

Establishing and sticking with one mid-funnel CPQL metric for prospecting partners

Why it is a mistake: Not all prospects are created equal. Prospecting partners can bring in varied customer segments. How much should you pay for your very best prospects? If one partner brings in prospects that are twice as valuable as another partner, it doesn't make sense to hold them both to the same standard. In addition, certain prospecting partners may be very efficient but can only handle low volume. As a marketer, you want the right balance between efficiency and scale without having to manage 20 prospecting partners.

What to do instead: Test and iterate the CPQL value for each partner. Establishing a mid-funnel metric like CPQL is just the first step. Lifetime value is the next key metric to consider.

How to do it right: Calculate the customer lifetime value originating for each partner and account for the multiple from the quality of traffic that each partner drives. (You can see how this can become significantly more complex once you factor in fractional attribution per partner for each customer. One step at a time, though! Start by ascribing just one partner per customer originally driven to your site as a "qualified lander.") Evaluate your whole marketing program based on total desired outcomes over an extended period. You may find that you are able to get many more prospects from partners with a more flexible CPQL goal, whereas you can tighten up the CPA goal for your retargeting partner. In the long run, that may be the right mix for your budget. It's like search marketing -- it's easy to harvest demand once it has been generated, but let's put the right amount of effort to generate that demand in the first place.

Short of deploying a formal attribution solution, these are the concrete steps to lay the groundwork of a basic attribution framework with your digital marketing partners, including specific prospecting and retargeting partners. Gather data, test, iterate, communicate with partners, and constantly re-evaluate. In doing so, you'll market smarter with your partners and be an attribution hero for your organization!

Jarvis Mak is senior vice president, customer success, at Rocket Fuel.

On Twitter? Follow iMedia Connection at @iMediaTweet.

Jarvis Mak runs the Customer Success group for Rocket Fuel, responsible for ensuring successful campaigns and delighting customers. He originally joined Rocket Fuel in 2009 as VP of Analytics. Previously, Mak was with Havas Digital as a senior vice...

View full biography


to leave comments.