Today's Search Insider column is based on coverage of Gian Fulgoni's keynote at the Search Insider Summit, still in progress here in Captiva Island, FL. Follow me on Twitter for more updates from the event, check out who else is tweeting the show, and read more at MediaPost's RAW blog. Also check the Search Insider Summit photo pool on Flickr.
Dreaming of a Super Model
Does search need a new business model?
ComScore Chairman Gian Fulgoni suggested as much in his opening day keynote at MediaPost's Search Insider Summit this week. He stressed that marketers are realizing so much more value from search than they're paying for, and that means there's money being left on the table. Here's some of his analysis as to why (read more stats in my blog's coverage of the session):
- One-third of ad dollars are focused on brand building, which is the reverse of traditional media, so we need to figure out how to use search and display to increase branding value.
- There are three components of how search drives buying: direct online effects (16%), latent online effects (21%), and latent offline effects (63%), so 84% of the value isn't being monetized by search engines, and marketers aren't generally measuring it.
- Enquiro did a study that showed a 16% brand lift when a brand was advertised in the top sponsored and organic results, so even without a click, there was value.
- ComScore's data shows that only about 5% of Google's paid links result in a click; the other 95% of ads are really "unpaid links," yet they deliver value to advertisers.
All of this led Fulgoni to muse that the cost-per-click model may not be the most effective. He noted that perhaps a cost per impression (CPM) model or another tracking view-throughs could better account for branding value and, in the process, bring over more branding dollars.
Here's the problem with that: the model and the measurement aren't the same. We don't need to rely on the pricing model to determine what marketers measure.
Fulgoni seems to be looking at the industry from the perspective of the search engines, which seem to be at fault for not monetizing searches as well as they can. It's ironic in light of last week's analysis of Jakob Nielsen's missive, where Nielsen suggested search engines were actually making too much money and should give some of it away. One says they're too poor, another they're too rich, and I'm like baby bear saying that in this instance, it's just right.
The search engines will probably do fine. Fulgoni said as much later, noting that while marketing is the first thing to get cut in an economic downturn and advertising is the first part of marketing to get hit, the accountability of search marketing can make up for it.
So, what would a CPM model add to search? In this economy or any economy, not much. Yes, it would show marketers that there is more they should value. At the same time, that would make search marketing much less appealing. There are competing forces: the engines aren't cashing in as well as they could be, but marketers have enough pricing concerns with search even when they're getting a great return on investment, let alone when they're struggling.
There are two major challenges preventing any sort of CPM model from working with search. The first is that the engines would consistently need to demonstrate the impact on branding, latent conversions, and offline conversions for every single ad that runs. Last I checked, comScore doesn't give that away. Still, the engines can continually strive to make their reports more robust, especially in ways that will appeal to brand marketers.
Secondly, search engines have shifted entirely to a performance
model where they penalize ads that only deliver branding value. Some sort of
immediate online action needs to occur often enough for the ad to rank well, if
at all, or it will get bumped. I touched on this over a year and a half ago
with the column "
Google to PPC Branding: Drop Dead," and that was before Yahoo launched
Maybe the business model, which works well enough for most parties most of the time, isn't what's broken. It's also interesting that CPM and view-through models would be seen as an improvement, as they still don't account for brand lift, latency, and offline impact without intensive additional measurement.
So what will bring more branding dollars to search? The key is
developing integrated digital marketing strategies combining search with
display, video, and social media, all backed up by behavioral targeting.
Improving the measurement and reporting will help further still. None of this
provides the panacea that could be accomplished as easily as changing search
engines' pricing models, and it's all easier said than done, but it should help
bridge the disparity illustrated at the Summit