Targeted Display Advertising Tactics to Increase ROI

Last week I speculated about the imminent demise of print newspapers as advertisers flee to greener pastures on the Internet. Now I’d like to discuss some of the specific tactics that make internet advertising so much more attractive than its print counterpart.

Targeted internet advertising is a complicated, dynamic space with a ton of different players, but what it all really boils down to is feedback. When you see an ad in a print newspaper over your morning cup of coffee, there’s no good way for the advertiser to track or measure the outcome. The conversation is effectively dead before it starts. At best, they can sponsor focus groups to ask individuals how it affects them (which is extremely expensive), or compare sales in the ad’s region against other regions or time periods (which is incredibly imprecise). Ultimately, they’re not going to be able to come up with a satisfactory conclusion of how to best to focus their advertising dollars, or even to advertise at all.

The benefits of internet advertising begin with the ability to refine the calculation of return on investment (ROI). This allows advertisers to compare different campaigns, channels, and strategies with greater confidence, and ultimately to improve their advertising ROI.

Feedback is key to this process. At the most basic level, websites or advertising networks will track the number of viewers who click on a display ad. This simple, crude metric still provides a ton more data than traditional advertising, and it’s only the beginning. The following are examples of more sophisticated techniques ad networks and companies are using to better calculate ROI and improve the effectiveness of their advertising:

Measuring Latent Spending: Just because a viewer may not have bought the company’s product right away doesn’t mean the ad failed. By tracking IP addresses or other identifying features, an advertiser can tell if you came back days or even weeks later to make the purchase. This helps to more accurately measure ROI, especially for large purchases—such as cars—that often require a longer research period.

Aggregating Sales Channels: Similarly, the sale doesn’t always happen at the same channel where the ad was viewed. After seeing an ad, the viewer might instead choose to buy it at Amazon.com or pick it up on their next trip to Best Buy. Companies such as comScore allow you to aggregate offline or out-of-network data, such as Best Buy’s loyalty program, for their list of consenting consumers. By matching this data with results from your advertising campaign, you can gain a greater understanding of its effectiveness.

Behavioral, Geographic, and Demographic Targeting: Armed with a more complete measure of ROI thanks to the techniques above, a marketer can begin to systematically find ways to improve it. As consumers, our surfing habits contain a great deal of anonymous information about what we search for, what we click on, and what we buy. This information is tracked and disseminated to websites we visit. Combining this with very basic data about our sex, age, and location, advertisers and ad networks can hypothesize about what they think we might be interested in, target us with relevant ads, and collect feedback on whether their hypothesis was accurate.

Retargeting: A special subset of behavioral targeting focuses on getting viewers to return to a website they’ve already visited. Have you ever felt like you were being followed by ads for a product you looked at or a site you visited? You were probably being ‘retargeted,’ a tactic that has proved very effective in driving repeat visits and purchases.

Controlled Testing: Last week I briefly mentioned Monetate, one of OpenView’s portfolio companies, as an example of an innovative tool for companies to improve their marketing. Monetate is part of the emerging testing field, which centers around controlled experiments to separate promotional wheat from chaff. Even with the benefit of internet feedback and analysis, questions about the outcome of a promotion or presentation remain. Did it fail or succeed due to the timing, content, presentation, or audience?

Monetate’s software helps answer these questions by allowing you to experiment with different promotions, displays, or targeting criteria, and track how they perform against a control group. For instance, a retailer might find that by listing a certain product in underlined neon green, clicks on the item double and purchases increase by 22% versus the control group of users who saw the item in the standard black font. This may lead them to experiment with the color scheme for their entire site or just other similar products. As always, the ultimate goal is to drive sales and improve ROI.

Despite being around as long as the Internet itself, display advertising has, to this point, had much less success than its search counterpart. One major reason for this is privacy, as companies have been hesitant to put user information to work if it could potentially be considered invasive. Google and Facebook, on the other hand, have plunged headfirst into the droves of data at their disposal to see what they can come up with, and the result is advertising so targeted it sometimes borders on creepy.

The other thing it’s led to for Facebook and Google is tremendous commercial success. My bet is that with companies like these breaking the stigma, we’ll see much more targeted display advertising in coming years. Currently, however, the process is in its infancy, and the companies that can build on and refine the list I offered above stand to make a pretty nice ROI of their own.