The Value of Site Averages

Weve written repeatedly about matching the marketing objectives to the ad buy you make. First, in terms of measuring effectiveness, then later we wrote about matching the ad product and size to the campaign goals. Every regular reader of this column surely knows that Ryan and Whiteman believe in utilizing the correct tool for each job, and that the smart marketer needs a wide range of tools for the many different marketing goals that may be necessary.

For the ad sales professional, understanding what the various tools can be expected to accomplish is an essential prelude to effective client-centric selling. And part of that understanding has to include the ability to manage expectations around results.

What should a marketer expect to achieve with a given ad buy, and where can the responsible salesperson or site-side marketer serve their customer by clarifying expectations in advance of the buy?

We got an email this week from a reader wondering about average click-through rates for various ad types. Its a good question on the surface, but one that deserves a little more exploration.

It reminds us of the guy with his head in the freezer and his feet in the stove… on average, hes at a comfortable temperature.

Numerical averages offer only the most general guidelines and are rarely useful for any marketer with specific goals. We hear the 0.5 percent industry average click-through number as often as any other, and while it may be true, its not very helpful.

A true industry average would include click-through numbers for poorly planned, poorly placed, unattractive and unappealing ads for undesirable products, just as much as it would factor in the results of the truly compelling, well-conceived, skillfully targeted ads for a really hot property. Where your company, your product and your ad creative fall on that spectrum should tell you something about how happy you would feel about achieving an industry average result.

We know of incredibly targeted keyword buys that have achieved over 50 percent click rates, but generated only 15 clicks-the words were almost never typed. We also all know about the advertiser who has been running the same single banner creative on the same site for months without change, and the click rate has fallen to virtually nothing. When the range is that great, the average is meaningless.

To be useful, averages need to apply to the situation being compared. If we could know the average response rate to financial site teasers when targeted at 20-something women, that might help a like site know what to expect when focusing on that audience. It might even help an unrelated site gage general expectations for a similar campaign to the same target. But since advertisers are reluctant to share information that might help a competitor, we never get averages that are truly useful.

Wed say, forget the averages altogether.

Instead, think about what you want an ad to accomplish, and what budget would make sense to accomplish that. Think about what sort of acquisition cost per customer your product or category or current price points can support. Think about what a like goal would cost you to achieve in other media, to other targets, with other ad sizes or formats.

Track your own averages with a high degree of specificity about what was done, and how it worked, so you know which numbers will be valid to compare later campaigns. Make your marketing team a learning organization by focusing on learning more from every program you run (on and offline). Then, look at the reported average results with a high degree of skepticism, always asking yourself whether you really think your efforts should be compared to this industrys average.

For site reps, its good to know your own sites averages, but its better to know sub-set averages: whats the average pull by size of ad, by location on your site, by specificity of targeting. Know which creatives outpull others, so your advertisers can see clear examples of what has worked well with your audience. And know that marketers looking for general averages may actually be looking for help in clarifying what they really want to know to market more effectively.

Related reading