S.G. Cowen has search's slice of the pie reaching nearly 60 percent by 2010.
Per the Interactive Advertising Bureau, in 2005 search constituted 41 percent of all online ads, with the rest of the pie breaking down as follows: display, sponsorships, slotting fees, and rich media--34 percent; classifieds--17 percent; referrals/lead gen.--6 percent; and e-mail --2 percent.
All this pie talk is making me hungry. But I think we're cooking with the wrong ingredients here.
When considering what the current online media landscape looks like--and will look like in 2010 and beyond--I think we should be looking at a pie in which search's share is zero.
No, I'm not saying search is disappearing. Nor will increasing bid prices or click fraud cause its share to dwindle. I am simply proposing that we refine how search is evaluated in the context of measuring online advertising spend, so that we can truly gauge its penetration.
A few years ago, it was relatively easy to separate online search and display ads from a publisher, marketer, agency and consumer perspective. To publishers, search ads were those that appeared on the results pages of user-initiated keyword queries on their sites (typically via syndication). To marketers and agencies, search ads were those that could be purchased through search engines like Google and Yahoo and appear across their distribution networks. To consumers, search ads were those listings at the top or on the right of search results pages (and, even though many consumers still don't know the difference between paid and organic listings, they sure never confuse them with banner ads).
Then Yahoo Content Match and Google Ad Sense changed everything. Suddenly publishers (large and small) could display "search" listings on their sites without user-initiated query functionality. Marketers and agencies could capture more traffic from their "keyword" buys. And consumers began to see subtle, relevant text ads on content pages of the sites they visited every day in place of "Punch the Monkey" banners.
So how should these keyword-contextual ads be classified? Should budget spent on those placements fall under the search slice of the pie? Sure, they're purchased through search providers on an auction basis, but they are not based on user-initiated queries and, hence, are often no more relevant than a behaviorally or demographically placed ad. And they appear on publisher sites in standard banner, leaderboard, and skyscraper ad units--sometimes even in graphical format with CPM pricing.
And if keyword-contextual ads didn't blur the line enough, consider what MSN AdCenter brings to the table with the ability to target display ads to consumers based on their past search keyword queries. When a person who searches for "home improvement" on MSN and two days later is found on MSN Real Estate and served an ad for Lowe's--is that a search ad? Or, if it's in a banner format, does that make it display?
So far, lots of questions and no answers. Since there's no one right answer, allow me to propose a cook-off of sorts.
First, I think we need to bake a few different flavor pies.
Let's start with one for online ad formats, breaking down spend against text listings, in-page graphical units (i.e. banners, leaderboards, skyscrapers, large rectangles, etc.), floating ads, pop-ups, e-mail, etc.
Then, let's break out creative format. Here, again, we have text, along with gif/jpg, Flash, roll-over, video, etc.
Now, let's look at targeting. This could consist of behavioral, demographic, geographic, content, keyword-contextual, and user-initiated keyword query, among others. Although, given the MSN example of overlaying user-initiated query targeting with content targeting, there could be multiple targeting tools in play for any given campaign--so the sum of all percentages here would likely be above 100 percent. If this were a pie, the cherries would be spilling out over the crust.
As my fellow Search Insider Bill Wise pointed out, another key feature of search is the auction environment in which ads are purchased (although I'll save my argument about whether that's the real reason search ads perform so well for another time). And auction-based pricing is quickly expanding far beyond search. A pricing pie would help us assess just how rapidly this is proliferating. This would consist of fixed and auction-based pricing. And another pie could break out CPM, CPC, CPA, and Revenue-Share.
These are just some of the many ways we can slice and dice the online advertising pie. My hope is that, rather than considering a pie made up primarily of search and display, we can isolate the ingredients that really provide insight into where the market's at, where it's going, and the true impact of search along the way.
Following this recipe, all the various cooks in the digital media kitchen will be able to create campaigns that are fully baked. And, further, we'll be able to prepare marketing meals that will stand up to the distinct palates of Internet, TV, radio, print and out of home consumers alike. After all, as I pondered in a recent column, it won't be long before all of these mediums are fully digital and searchable.
So, shifting gears, how will the overall advertising pie--in which Jupiter projects online ad spending at 9 percent--look in 2011? I'll save that for my next column-- but, for now, let's just say that I think search is the key ingredient.