On December 1st, 2010, Conduit, a provider of customizable toolbars and web apps with a claimed user base of 250 million, announced that it was ditching Google as the provider of its search results in favor of Microsoft's Bing. The transition was scheduled for January 1st and appears to have immediately begun paying dividends for Bing. Traffic measurement firms like Comscore and Hitwise reported big gains for Bing from December to January, but few called out the connection with the Conduit deal.
TechCrunch's Michael Arrington was one of those few noting a connection, citing figures showing that Conduit.com "lost" 165 million page views from December to January while Bing.com gained 138 million. Around that time, RKG noted that paid search traffic for the combined Bing/Yahoo Search Alliance appeared to be turning a corner, but reserved final judgement until further analysis was completed. We can now offer some additional insights on this perceived improvement and the impact of Microsoft's deal with Conduit.
Here's an updated weekly view of the Search Alliance's share of paid search clicks as we see it across our client base. This includes traffic from partner sites:
After steady declines dating back to 2009, the Alliance's combined PPC click share bottoms out the week of 1/2/2011 before reversing course. It should be noted that the Conduit transition to Bing was not instantaneous, as we continued to see Conduit search partner traffic through Google up until the end of January.
An interesting and important wrinkle here is that when it transitioned to Bing, Conduit began sending traffic directly to a Bing.com search result, rather than to its own domain as it did as a Google partner. We can't know the true motivations for the change, but one has to wonder if the public perception of Bing.com's traffic levels wasn't high on the list.
Here's our view of paid click traffic specifically to the Bing.com domain:
With Conduit sending traffic directly to Bing.com it has made it more difficult for us and others to precisely assess the impact of this deal, but we can make some informed estimations with the information we do have.
How Big is the Impact of the Conduit Deal?
First, we have historical traffic data for Conduit when it ran Google powered results on its own domain. Going back to the fourth quarter of 2010, we can compare Conduit's traffic as an AdSense partner to traffic on the core Bing.com domain:
Towards the end of 2010, Conduit's paid search click traffic was equivalent to 10-11% of Bing.com paid search traffic for RKG clients. That percentage falls off quickly in 2011 as the transition begins and Conduit traffic effectively becomes Bing.com traffic.
An advertiser running a publisher performance report in Microsoft's AdCenter will not see Conduit listed as they would have in a similar Google report; however, RKG is able to determine those clicks originating from a site using a Conduit search toolbar by looking at the full referer string of an ad click. That referer looks something like this:
We find that searches on Bing.com with the "pc=conduit" tag comprise about 5% of all Bing.com paid search ad clicks. That's a fairly large discrepancy compared to the 10-11% figure above, but much of that is likely explained by the tag not carrying over to subsequent searches once the user is on Bing.com. That 5% figure has been stable though and we have no reason to believe that Conduit suddenly saw its traffic cut in half when it moved to Bing, so the former figure likely remains a reasonable estimate of the boost Bing.com is receiving.
So, what would the click share graph above look like if we gave Conduit traffic back to Google? Using the Conduit/Bing.com click estimate of 10% and accounting for the gradual transition, something like this:
Combined Alliance share still trends up since the new year in our estimate, but it would average about 5% lower than we see it now. For the week of 5/8/11 that translates to a paid click share of 22.4% with the Conduit traffic and 21.1% had Google kept it. That doesn't necessarily sound like a huge difference, but Microsoft has spent hundreds of millions of dollars in order to achieve gains on that scale. No word on the terms of the agreement with Conduit, but recent speculation on a Conduit acquisition by either Google or Microsoft, put a deal of that scope at $1 Billion.
Here is the same estimate applied to Bing.com click share, showing that without the Conduit traffic Bing.com paid clicks have been essentially flat since the completion of the Alliance:
Is this sort of agreement fair to advertisers?
On first impression it seems a little deceptive for Conduit generated traffic to appear to advertisers as if it is coming straight from Bing.com, but it may not sound too unlike past deals Google has struck with Mozilla to make Google its default search engine for Firefox.
Perhaps what's disappointing is what this may portend for for the future of the search business. We've lauded Yahoo and later Bing for the controls they have given advertisers to manage search partner traffic and this is a step in the other direction. When it was a Google partner, Conduit provided traffic that was worth about 2/3 as much to advertisers as traffic on the core Google.com domain. That's a level at which some may choose to exclude it in order to drive more traffic at the same ROI from the core domain and the better performing partners.
That's a key difference between the Mozilla and the Conduit deals (along with the fact that a Firefox user can easily change their search bar engine, while a Conduit user cannot.) Someone using a toolbar customized for a specific site may just be less likely than the average searcher to be looking to make a purchase. While we have little control over our spend for Firefox users either, we've found their value to advertisers to be close to average. If advertisers are able to make these types of determinations, we would appreciate having the ability to act on them.
It will be interesting to see how these third party arrangements play out over the next year or so. Google's deal with Mozilla reportedly runs through the end of 2011 and you have to think Microsoft may just be willing to make a better offer to have Bing as the default search provider, even if it means working with an old foe. With Google's Chrome browser gaining share and directly competing with Firefox among other factors, a re-up of their deal is no guarantee.
There are also plenty of current Google search partners that Microsoft could try to pick off with more lucrative deals, particularly if they struggle to gain share through more organic user adoption. Conduit is a pretty big fish, but not nearly the biggest. Picking up half of Google's top 10 partners could net the Search Alliance on the order of 10-12 points of paid click share, but potentially at a very high cost to Microsoft and to traffic quality. Yahoo saw the reverse of this in 2009 when it lost eBay and up to 5-10% of some sites' paid clicks to Google.
While Google seems firmly implanted as the dominant player in search, the Conduit deal and speculation around big players like Aol, Apple and Mozilla shows that it's a continuous battle to retain that level of dominance and that a considerable portion of search share is under a tenuous hold.