Ovum Says, ‘Microsoft Has a Plan’ for Search

May 24, 2008

Ovum, a British consultancy of high repute, asserts that Microsoft has its sights set on being “the king of search”. You can read its summary here. This article, penned by Mike Davis, is based upon a longer piece available to Ovum’s paying customers as part of the pundit shop’s Straight Talk service.

The Ovum conclusion, if I read Mr. Davis’ article correctly, is that Microsoft’s pay-for-traffic initiative is just one component of a far larger strategy to close the gap with Google. He writes:

The technology for the programme came from the acquisition of Jellyfish.com last year. The service is a different proposition to merchants than the usual ‘cost per click(s)’ such as used by Microsoft’s current nemesis Google. The payment model being used by Microsoft is called Cost Per Acquisition, and the advertiser only pays when the advertisement results in a purchase.

So, it’s not pay for traffic. It’s a rebate of three to 30 percent, requires a minimum balance of $5, and is designed to go after Amazon.com and eBay.com.

The point that jumped out at me is that Mr. Davis tosses the Fast Search & Transfer acquisition into the mix. Mr. Davis sees the pay-for-traffic plan announced by William Gates at the Advance 08 advertising conference and the $1.2 billion deal for Fast Search as signs of Microsoft’s determination to be “king of search”.

Let’s assume that Ovum’s research and Mr. Davis are right on target. This means that:

  • The Jellyfish technology underpinning the cash back for search play will generate traffic and hence ad revenue for Microsoft.
  • The Fast Search technology will allow Microsoft to break through the 50 million document barrier that some SharePoint users encounter with native SharePoint search
  • Consumers and advertisers leap on the cash back bandwagon and SharePoint licensees pay for the Fast ESP (enterprise search solution)

Each of these actions take place quickly and produce gains for Microsoft.

How much traffic and revenue does Microsoft need to become “king of search”?

The gap between Microsoft and Google is a reasonably large one. Recent data from an admittedly uneven resource suggests that Google has about 62 percent of the US search traffic. Google’s share of the global market is higher. In the April 2008 period, you can read Mashable’s quite good analysis here, Microsoft lost search market share. If the ComScore data are accurate, Microsoft accounts for 9.1 percent of the search traffic. The month before, Microsoft’s search traffic was 9.4 percent. Google’s share is growing, if the ComScore data are correct; Microsoft’s share of search traffic is degrading. Wow!

In order to close this gap, the pay-for-search scheme is going to have to reverse a declining trend, attract advertisers, and scale like the devil. I don’t think the pay-for-traffic scheme will work whether it is aimed at Amazon.com, eBay.com, Google.com, or Yahoo.com.

The Fast Search deal is going to have to show some sizzle. At the recent Enterprise Search Summit, I stopped by the Microsoft exhibit and asked about search. I was told SharePoint was quite good. I asked about Fast Search and I was told that Fast Search had a booth. I asked, “Please, show me the Fast ESP system running on a SharePoint system.” The nice Microsoft person said, “I don’t have that information.” So, no FAST logo in the Microsoft booth and no demo that I could see. Keep in mind that there were vendors such as BA-Insight, Coveo and ISYS Search Software, among others, showing potential buyers SharePoint search systems that worked, scaled, and delivered the nifty metatagging so much in demand.

I walked to the opposite side of the room where the Fast Search exhibit was. I asked to see the Fast ESP SharePoint demo. I was told, “Come back between sessions. We will have it up then.” I came back and was told, “We’ll walk you through the basic systems. SharePoint works the same way.” Iasked, “Where’s your Microsoft logo.” The really friendly person told me, “We don’t have that logo yet. Leave your card, and I will get that information for you.” I said, “No.” Your PR guy hassles me about not knowing anything about Fast Search despite my analysis of the system for the US federal government over a two year period.

Now putting the pay for traffic puzzle piece up against the Fast Search puzzle piece, Ovum sees a fit. I don’t. What I see is a very large orgaznization faced with market push back on three separate war fighting fronts. A three-front conflict is complex, not tidy. And what are the three fronts?

First, Microsoft controls the desktops of 90 percent of computer users and Internet Explorer is the default home page for the browser. Google’s market share means that people are consciouly navigating to Google to run queries even though the Micrsoft Web search box is the default. Most people don’t change their default home page, so extra clicks and typing are required. People like easy, but when it comes to search people go to Google. With a massive market share and the default browser’s search box, users go to Google. I find this pretty amazing. The longer Microsoft persists in losing market share, the more deeply ingrained the Google habit becomes. In the history of online, user habits–once set–become very hard to change.

Second, the pay for clicks approach is a double edged sword. Here’s why. There is a tremendous incentive for users to find ways to scam the system. Google has to work overtime to snuff out fraudulent clicks. Microsoft–lacking a high traffic site and the easy money of AdSense–will find that it must spend more to deal with tricky users. And if the pay for traffic play is really successful, Microsoft will have to scale its online system quickly. One edge is giving up some money by betting more traffic will yield cash. The other edge is that success means more scaling costs. The way I look at it, the pay for traffic play costs money and does not hold the promise of a way to lame the nimble Googzilla.

In fact, Google is adept scaling quickly and at lower costs due to its use of commodity hardware and its “smart” extensions to Linux. Microsoft has yet to prove that it can scale without taking extreme measures such as complex tiering, using super-fast, expensive, high-end branded server gear from Hewlett Packard and other vendors, and dealing with the time and bandwidth issues imposed by Micrsooft’s own 64-bit operating system and application overhead. Microsoft has to spend more to get the basic job done. My take? A huge success for Microsoft results in higher costs. In the short term, that’s not a problem. Over the longer term, higher costs can become a problem even for a deep-pockets giant like Microsoft. If performance lags or user trickery becomes evident, the gains may slip away leaving puddles of red ink.

Third, the Ovum analysis says that the pay for traffic play is based on the Jellyfish acquisition. The enterprise search initiative is based on the Fast Search acquisition. These two key components were not invented at Microsoft, and I have a hunch that integrating these acquired technologies into the Windows-based systems is a work in progress. Again, more costs and increased chance for technical and managerial friction. Microsoft’s ingrained project manager system and its silo-type structure make feudal squabbles between digital princes a feature of Redmond life. To be “king of search”, these destructive hot spots have to be remedied. Google’s certainly not perfect, but it seems able to innovate without clashes over interface and technology popping up when I use its system.

I wish Microsoft well in its quest to become “king of search”. I know Ovum’s management wants its analyses to be accurate and generate consulting business for the firm’s analysts. I hope SharePoint users find search happiness from Fast ESP. I hope Web searchers find that Microsoft’s Web search initiatives deliver the goods.

Microsoft has to find a way to leap frog ahead of Google. I’m not sure making acquisitions and paying for traffic fit together seamlessly. Furthermore I disagree that these two initiatives mesh, have been fully integrated, and represent a significant challenge to the GOOG. Agree? Disagree? Let me know.

Stephen Arnold, May 24, 2008


Comments are closed.

  • Archives

  • Recent Posts

  • Meta