Labor Day weekend is over, and the advertising industry is back from its annual August hiatus. As clients and online ad buyers review budgets, plans and schedules for Q4 and Q1 2005, behavioral targeting will, no doubt, be one element they’ll evaluate.
When full-year budgets were built in late 2003, behavioral targeting was but an afterthought. Now, with dozens of case studies and research reports about its effectiveness out in the field, a number of buyers are scrambling to include it in Q4 programs. This is great news for publishers with behavioral targeting capabilities, and will be a great test for development of the sector and new products. It will bring to the forefront some of the key issues in behavioral targeting, like segment “granularity,” or “segment sizing” (for those who like to avoid words like “granularity” in their lives.)
When you discuss behavioral targeting with market participants, whether ad sellers or ad buyers, the same question inevitably arises: how big (or how small) should audience segments be? For example, if you sell or buy “Auto Buyers,” is a segment of 100,000 monthly audience members enough? Or is 250,000 a better number? Is a segment of 10,000 too few? Of course, everyone asking that question wants a simple, straightforward, black-and-white answer. They don’t want “it depends.”
There is a simple, straightforward answer to the question. The right size is where buyers and sellers meet. The right size is the volume of targeted audience a seller is willing to sell at a certain price when it coincides with a buyer’s willingness to buy that audience at that package and price. The market dictates it.
In the end, this is no different than how decisions are made when selling contextual advertising or sponsorships. Sellers make judgment calls about how to size ad packages, recognizing that there’s a certain “friction” cost in every transaction. They can’t make the packages too small or they’ll never be profitable. The same holds true in behavioral targeting. Publishers must: 1) assess the volume of targetable audience segments they have available; 2) assess advertiser interest and appetite for each segment; 3) determine the overlaps between the segments to determine potential inventory conflicts; and 4) package and price their segments proactively and according to their own business goals, whether they be market share, share of wallet, profitability, or inventory yield management.
Of course, at the tactical level, the discussion will always be driven to buyers who want to buy very small, extremely granular segments, such as suburban women with children and household incomes over $150,000 in the market for a car who have searched for minivans or SUV’s in the past 3 days (the members of such segments may only number in the hundreds or thousands). Sellers will want to sell large, more generalized segments, such as anyone who has read automotive content over the past 45 days.
Successfully negotiating those conflicting agendas is what drives the media business and makes some publishers profitable, others not. Keeping a sales staff from promising any segment a buyer could ever want at almost any price is what sales management is for. Sellers are supposed to sell. Managers are supposed to maintain control over what they sell.
I suspect general identification of the issues surrounding the sizing of audience segments for behavioral targeting leaves those on the front lines a bit wanting, so the next column will dig into specifics on this issue. I’ll strive to address “segment sizing” in ways that make sense to help sellers, buyers, traffickers, and operations folks keep our industry moving forward. Your questions, comments, or suggestions for issues to address are very welcome.
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more
Google Analytics comes with lots of standard reports and settings, but with a little customisation you can extract much more value. One way is ... read more