When to Think Behavioral Ad Targeting
Assuming for a moment that a company that provides behaviorial targeting for web-based ads can actually do so - a separate discussion altogether - when might it be wise or necessary to consider this method? There's at least two answers to this. First, the obvious. It makes sense when your testing of this method results in an acceptable ROI. (I did say it was obvious!) And unless you're a somewhat small business truly lacking the resources to manage another campaign, there's little reason that you shouldn't try testing this method if you're playing in the search engine marketplace. The second is a bit less obvious. Behaviorial targeting makes sense when you have no other effective options or competing in the contextual marketplace is untenable for you.
There are several places where contextual online ads, (be they banners or text ads in search results), simply aren't going to be that effective for some specific products and possibly for whole categories of products. Yet collectively, these contextually weak areas represent increasingly huge traffic numbers. Consider just about any personal home pages or profiling service. (Myspace.com, Yahoo/Geocities and so on.) Though not always the case, the pages themselves lack any inherent theme or context. The same is true of most personal blogs. As a result, the ads themselves are unlikely to be as contextually relevant as in more structured content. Ad engines will crudely attempt a match to some keywords within the document, but will not be deeply contextual with today's technology. While those selling ads may say otherwise, it doesn't seem that ad delivery engines do well with extracting the actual semantics of a page, determine what general category/taxonomy the topic area is in, and then deliver appropriate ads. No, it's keyword based. Use of synonyms and such may be applied, but nontheless, this is the point at which contextual targeting may fall down for some page types. (As the web content space grows, this will be an increasing problem for search engine effectiveness as well; possibly most problematic for Google, which explicitly depends on a PageRank focused priority.)
Just surf some of these areas, (such as personal home page sites), and look at the ads. Sure, sometimes a page's content will be focused enough to allow for context. And keyword densities or whatever the keyword algorithm being used may work. However, keywords alone do not necessary abstract whole context well in pages with no clear overall topic focus. So when we're talking about millions of pages and billions of pageviews that are unclear meaning-wise, what do you do as an advertiser who wants to reach those web surfers?
Behavioral Targeting to the Rescue
When you don't have context for an ad placement via appropriate keyword matching, behavioral targeting may be the only way to bring relevance to a web page visitor. The whole premise of the success for text-based contextual advertising is related to the attention focus of a user. If a user is in a 'mind space' within a particular topic, they're more likely to notice anything with the same "information scent." People read based on word patterns, not just parsing individual letters and syntactically forming sentences. And it's this cognitive state of the user that's allowed such ads to command an attention grab from the main content.
Behavioral targeting is of course no different in the sense that it seeks to catch users' attention based on their current pre-disposition to certain messages. So if there's something about the sort of users that would be visiting these contextually poor areas that appeals to you, behaviorial targeting is something worth testing out. Your challenge may be finding a company that can sell you the segment you're looking for.
What About the Dollars?
Another solid reason for delving into behaviorial targeting might be if the nature of the keyword space your products are in is either hotly competitive or forces you to compete with ambiguous keywords. For example, you may be able to make an behavioral ad deal resulting a Cost Per Click (CPC), or other metric of your choice, that would rival the results from a more expensive set of keywords. (Of course, CPC alone should not be the only view you have on your campaign, but that is a separate discussion.) In any case, one way you could find yourself in an unacceptably expensive keyword space is if your keywords are ambiguous with other products.
For example, use a search engine to seek out "caribbean vacations." If you're a relatively small adventure travel business with SCUBA trips to the Caribbean, even if your trips have solid margins built in, you may not be able to cost competitively bid against large travel agencies, airlines and so on. Yes, there are likely other keyword phrases that can scope you closer to your prospective customers. But you're essentially locked out of the simple and most common phrases.
What to Do
In the end, behaviorial ad targeting is like any other web advertising product. It's worth testing and you'll need to have clear objectives on your required ROI. It will be important to remember that your creative is no longer going to be contextual to the page, only to the user. Ideally, your ad delivery partner will be able to offer you some advice in this area as you may be used to creating keyword oriented context and the required appropriate landing pages.
