Tying Offline Advertising to Online Results

No longer do people debate whether you can use an in-commercial response mechanism either to sell things (e.g., Ginsu knives, George Foreman grills) or build brand image, but not to do both at once. Brands such as Jenny Craig and Orbitz.com show the evolution of the use of “blended” media to simultaneously build call or click volume and craft a brand image.

Today, many prominent advertisers are doing the bulk of their ad spending offline in traditional media, but the purpose of those expenditures is to drive consumers to the phone or Web. The field of media measurement has struggled to keep up with this gush of activity — a brand has hundreds of commercials running at once, generating hundreds of thousands of Internet clicks — and to enable advertisers and agencies to marshal it into effective optimization across the realm of media choices.

In a particularly relevant example of blending, when Jenny Craig wanted to know which local networks were driving the most sales with commercials containing a nationwide 800 number (1-800-JENNY20), its agency, Carat, helped to compile a call database by specific creative execution. From that work emerged the ability to build empiric data on how specific creative executions performed by network. Taking things a step further, the agency-client team tagged each customer prospect with a unique ID and tracked the customer’s interaction in conjunction with the media analysis. Moving forward, every contact with every new prospect — regardless of how long it took the person to actually plunk down his or her money or whether that contact was made over the phone or Web — was tracked and the results were tied back to the initial media spend that drove the customer to Jenny Craig. As a result, calculating lifetime value for each customer was relatively easy.
Spurred by continued advances in technology, the availability of new data and more effective use of existing data, tactics for tying online activity back to the offline push continue to evolve. There are three prominent tactics: proximity, deduction and micro-tracking. Proximity is used when Web site response to a TV commercial happens so quickly that the correlation between offline advertising and online results is clear. Deduction is used either when response is not immediate, or when an advertiser is on air so frequently that tying Web hits to specific TV spots is impossible. Micro-tracking involves using dedicated URLs for specific media outlets.
When Gotham Direct launched GoDaddy using Super Bowl spots, the agency knew that people often responded faster to a TV commercial by using the Internet than they did by using an 800 number. The agency’s goal for GoDaddy was to push consumer signups within a mere 10 minutes of the first TV commercial airing. The client demanded immediate results, and that’s exactly what it got from the tight combination of provocative creative, crafty placement and a polished, ready-for-business Web presence. Without getting terribly technical about its approach to proximity, Gotham had examined traditional TV response patterns and applied that learning to data from its client’s Web site. This generated success metrics particular to each TV network. Drilling down further, when there were multiple ad occurrences, the agency decided to pro-rate responses based on program audience size. By constantly tracking each day’s cost per acquisition on an hourly basis, under-performing networks and programs were weeded out. Is proximity foolproof? Certainly not, but it provides strong, directional intelligence based on the quality and stability of the results it can generate.

Deduction is a way to ascribe a percentage of Web activity back to overall DRTV broadcast spending based on spending patterns and consumer signups. When the agency Direct Partners was helping EarthLink/People PC evolve its direct-response program, a big challenge was that the client was on air constantly, making it virtually impossible to precisely track commercials to Web site activity. A breakthrough was achieved by matching Web activity during the client’s occasional broadcast hiatus, or during times when its spending was up or down, to Web activity in the same time frame. When spending was boosted by 50 percent from one week to the next, there was a predictable increase in Web activity. This knowledge was used to create formulae for spending optimization. In this particular example of deduction, Web responses ended up being spread evenly across the entire TV buy, which resulted in a lower cost per lead.   

The most common approach to micro-targeting is with a suffix or sub-directory, such as www.thebrand.com/tv/123, to which consumer prospects will be driven and tracked back to specific commercials using special codes. This presents several problems, not least of which is having to create separate landing pages for every code used as a suffix. A second method is to use individual domains, like www.123thebrand.com, with no period between the 123 and the brand. But this requires buying and managing multiple domains based on how many media outlets are being used. What some advertisers have found most successful is using a prefix like www.123.thebrand.com, wherein the numeral is tied to a specific cable network or broadcast station. This model takes only one domain to yield multiple prefixes that can be created in seconds. The prefix 123 could be dedicated to spots running on Comedy Central, while 132 could refer to spots running on the History Channel. Pixel codes on a Web site transaction page capture all activity, from browsing to buying, and the transactions are tied back to the prefixes.

Carl Langrock is president and CEO of COREMedia Systems, Fairfield, N.J. He can be reached at clangrock@coremedia-systems.com.