Web Moves Blog

Microsofts Search Engine Discussion
As a follow up to our look at what Microsoft’s new search tool could look like, Rob Sullivan, our Head of Organic Search, and I locked ourselves in an office and tried to tackle some big questions about what will happen when Microsoft enters the search industry. We suspect these questions have been on a few peoples minds.

Q: Given what Microsoft is working on for Search, what do you see Microsoft doing between now and the release of Longhorn?

Rob: Version 1.0 of search will out by the end of the year; Bill Gates has already stated this. It will look the same as everybody else, however; nothing too different or radical. They will be playing catch up for the time being, offering similar features to Google, Yahoo, etc. Trying to get MSN competitive with other portals.

(Note: as a follow up to this, you can see a beta MSN search portal at beta.search.msn.com. Hmmm..notice any similarities in the results to Google’s layout?)

Their deal with Overture lasts until 2005 so we won’t see too much change with that. Likely after their current contract expires they will go with yearly renewals only. Nothing too long term, and with the option to bail. This is for when they are ready to launch their own PPC.

By next year, I wouldn’t be surprised to see Microsoft buying some PPC outlet, like a LookSmart or Enhance.

Gord: I think Rob’s right. Microsoft has to start building a name for their own search product, so they’ll introduce it on the MSN portal. It will be introduced with a lot of hype, but little in the way of functional advantages. Ideally, Microsoft will be able to add one or two features that would give it a distinct user advantage over Google and Yahoo.

Another important thing to remember is that Microsoft is probably trying to recoup their investment in search as soon as possible. I don’t think they are prepared to sink hundreds of millions of dollars in a search black hole without a return for years. They’ve played this game in the past to capture market share and I don’t think they want to go there again. That’s why you’ll see them hang onto an agreement with Overture for the foreseeable future.

Q: Why Buy a PPC Outlet?

Rob: Buying a PPC provider is quicker than building one. The current PPC suppliers already have developed the back end systems, such as the database, reporting features, and so on. Also, in some cases (particularly Looksmart) the price is right. At least by buying a PPC supplier they can quickly monetize the purchase. After all, look at how much money Yahoo made this quarter alone, the first full quarter since the Overture purchase, because of the PPC supplier.

Gord: We have to remember that Microsoft will be throwing a lot of resources at developing their own search technologies. I agree with Rob. It makes more sense to buy an existing PPC provider and get access to the technology. The one caveat here is the Overture/Yahoo portfolio of patents, which currently has some PPC portals paying them license fees. Microsoft will be looking to steer around this. And this brings us back to Google. Google AdWords uses a system sufficiently different from Overture that it doesn’t infringe on their patent. Could this be one of the reasons Microsoft was originally looking at Google?

Bottom line: Microsoft will want their own paid search channel as soon as possible, but will be looking for short cuts to get there.

Q: Why won’t Microsoft hold off on unveiling their search until Longhorn?

Rob: The problem will be winning people over. Microsoft results are not the best right now. MSN search doesn’t have a very good reputation. It has traditionally been confusing for most people, what with sponsored listing, organic listings, web directory listings, featured sites and so on. They’re already changing their default search to something that has more reliable results but Microsoft will have to overcome this bad quality perception to convince people to use search. Also, as they roll out with new search features, they will continue to change their results pages’ displaying more or less ads, more or less free listings to find the right balance between monetizing the paid listings. And during the process, they have to build their market share.

One way they could attempt to win people over would be a side by side comparison of results with other sources. If they could prove to more advanced web users (the early adopters so to speak) that their results are at least comparable, but more than likely superior to other providers, they can start winning people over. They will have to do this specifically before Longhorn comes out. Microsoft has to convince people that MSN search provides quality. If they can’t win people over with MSN search, they won’t likely get them to use Longhorn’s search capabilities.

Gord: Microsoft has never built a search brand. Right now, Google owns that outright. Even if Longhorn ships with its default set to Microsoft search, people have to have sufficient reason not to change it. It remains to be seen what the Antitrust overloads will do to prevent Microsoft from monopolizing search (like they did with browsers) but we can be sure that the user will have to at least have the appearance of a choice.

Microsoft simply can’t allow Google to continue to own the search business outright for the next few years. They have to start establishing their own brand identity. For the next two years, the search fight will be fought on two fronts: through the portals (Google vs. MSN vs. Yahoo) and through toolbars and desktop search apps. Microsoft doesn’t have to win these outright, but they have to put up a good fight to build a position as a contender. If they can do that, they can eventually crush the competition through OS integration with Longhorn.

Q: Of all the functionality we looked at in the last NetProfit, how much will ship with Longhorn?

Rob: The WinFS part is a given, that has to happen. They need to incorporate the XML features of the new file system with its ability to easily perform local searching. It will also give people the ability to see what MSN Search could become in the coming years; i.e. being able to search for a song based on genre, performer, songwriter, and so on. Once people learn of the power of the WinFS search feature they will likely come to rely on it for web search. This is key to Microsoft search. If people don’t buy into the power of the WinFS search, they won’t buy into Microsoft search.

They will have the Avalon desktop because the cool factor. Microsoft hasn’t changed the desktop since Windows 95. While they’ve added a ton of great features since then, they haven’t improved on the desktop. Having a desktop with the ability to render in 3D definitely contributes to the cool factor.

In addition, I see the subscription portion of MSN being more tightly integrated into the new OS. The ability to stream audio and video to subscribers will be a huge money maker for Microsoft. The subscription service will offer more features than the regular MSN portal.

Gord: In addition to what Rob has mentioned, I believe Microsoft will also introduce an iterative search interface, that will allow the user to tweak and filter their search and have the results update in real time. I believe this functionality is in complete alignment with how people actually search and will be a big factor in winning over users.

Microsofts Search Engine Discussion cont’d
Q: How about Implicit Query?

Rob: Not right off the bat. That will be likely wrapped into further enhancements. Also, the ability to anticipate queries, and aid in queries means that the computer has to know more about the user. Therefore it will have to monitor the user both locally and on the web to see how they search, what they search for, what results they picked and so on. Until the computer can understand the user Implicit Query won’t work.

Gord: I’m not so sure. Microsoft is testing Implicit Query pretty aggressively now, and I wouldn’t be at all surprised to see some version of it included in the first release of Longhorn. That, and the fact that the marketing possibilities of Implicit Query are staggering. I’m sure Microsoft is fully aware of that.

Q: How much of a selling feature will Search be in the Longhorn release?

Rob: It won’t be the major feature; Microsoft will be pushing the desktop features and functionality more than the web search feature, because of convenience. The ability to get information from formats such as sound and graphic files will be appealing to users. I think that initially many people may even change their default search preferences to a site like Google, especially if MSN search doesn’t perform. This is why they HAVE to get search right early on.

Gord: Right now, all Microsoft’s talk about the search functionality they’re developing is about its application on the desktop, not on the web. This leads me to think that they want to nail it as a localized search tool first, and then extend it to the web. For this reason, I think there will be a lot of marketing about ease of use and the ability to work more intuitively with your files without having to become a librarian. The web search implications will be rolled out later, as Microsoft moves their search focus online.

Q: What’s Google going to do?

Rob: Google has to get into the portal business. They need a built in traffic stream that’s looking for other features over and above search. That’s why they’re having an IPO – to raise money. Think about this for a second. You already have a hugely popular, highly profitable web property. Estimates are that it takes in a billion dollars a year in revenue with anywhere between $250 and $500 million in profit. Why would you IPO? Because you need to raise money. And with that money, they’re going to buy AOL or Netscape.

Why buy AOL or Netscape? Well either of these properties will give them a dedicated customer base, a portal, the ODP, a chat program (ICQ) Movie listings, mapping capabilities and so on. It puts them on a somewhat equal footing with MSN and Yahoo offerings. Not to mention that Time Warner has had nothing but headaches with the AOL branch since the two companies merged.

Gord: Google has to make a bold move forward in search functionality. They came out so far ahead of any search engine in 1998 that once the competition caught on, it took them 5 years to catch up. Now, that competition has caught up, and Google hasn’t been able to raise the bar that significantly since.

Google has to jump ahead of the pack again, and based on past experience, that jump on functionality will be squarely focused on providing web users with a better search experience. While I like Rob’s portal theory, I think such a move would split Google’s focus in so many areas that they’d lose sight of their core competency, which is search. In my mind, Google only has one slim chance to win the upcoming war with Microsoft, and that’s by continually focusing on providing the best search tool on the web.

What Impact will Yahoo dropping Google have?

Rob: Google will loose a substantial chunk of traffic, obviously, but it won’t have much of an impact on Yahoo regarding quality of search results. Yahoo will still have 1/3 of web users after they switch. They can replace with Inktomi, or a mix of AltaVista, Fast and Inktomi. How Yahoo will work is they will build features and test the new features on AltaVista. When they’re satisfied that the features do what they want, and are useful, they will implement the new features on Yahoo. The average Yahoo user won’t likely notice much of a difference from the day before they dropped Google to the day after.

As far as search functionality, they’re refining their semantic processing. They have had a year since they bought Inktomi, and Overture has had Alltheweb and Altavista for six months. It isn’t inconceivable that they have a huge amount of research and development going on to make a search product capable of replacing Google. (By the way, semantic search will also be a feature with MSN search therefore it’s safe to assume that Yahoo will be developing it as well)

Gord: The loss of Yahoo has been looming for ages, and I would hope Google has a Plan B, as well as a C, D and E. Really, the Yahoo arrangement has been a bonus for Google from the beginning. The fact is, Google still owns a huge chunk of search traffic outright, and they have to concentrate on this before anything else. If we’ve learned one thing in the past few years, it’s that you can’t depend on partnership arrangements for your success.

Google will be finding ways to build search traffic directly to Google. The launch of the Google Toolbar was the first move in this direction, and a brilliant one, in my opinion. I think toolbars and other desktop search apps will be the next hot battleground of the search industry.

Rob: The down point to that is people have to agree to download from Google. With Microsoft, it will all be built in. Again, a huge competitive advantage to Microsoft.

The thing is, if Google does do something revolutionary, and offers it as a free download, all Microsoft has to do is build it and release it as a patch or service pack to get it implemented.

Gord: And that’s why I think Microsoft can’t be beat in the long run.

Rob: I think Google can get back on top, and the application of the semantics db is just the first step. They have to get that working first. I think they are close and getting closer all the time, but they still have some tweaking. Once they do get it fixed they have to get in front and then stay in front, so the competition is always aiming for a moving target. Even then I don’t think superior search is enough to keep them in front. They have to offer more.

Google will keep trying to introduce ways to make search more intuitive and useful. For example, the slide show search at Google labs is kind of cool, if they find a way to make it more applicable to people.

Q: What about Yahoo?

Rob: Yahoo is in the most unique place right now. They have nothing to prove and a solid customer base. Anything they do is an improvement, so they have no where to go but up.

I think they’ll have to go into semantic search, like Google and MSN. The first roll out of a pure Yahoo search will be vanilla organic search, but they’ll be changing that as time goes on. By this fall they’re going to want have something out before MSN. MSN is the key in Yahoo’s formula. Yahoo will want to be ahead of MSN, and Google wants to be ahead of everybody else.

Gord: I wouldn’t want to be Yahoo right now. I can’t see how they’ll win in the long run. The one area that’s interesting is the Yahoo shopping search engine. Perhaps Yahoo will eventually have to concede the general search contest, and become a niche player providing specialized search functionality. But they’re not going to go quietly. It’s going to be a huge battle in the coming few years.

Microsofts Search Engine Discussion – More Q&As
Q: Does Yahoo offer anything unique or superior?

Rob: They’re just relying on their brand. They really don’t have any features that set them apart. That’s not to say that they won’t develop these products, but I think with Overture making so much money for them that, at least in the short term, they don’t need to innovate to stay in the game. Once they realize that they are getting left behind (or at least are simply status quo) they will invest more into organic search R&D. If it will be too little too late is anyone’s guess at that point.

Gord: Overture may provide another key to survival for Yahoo. In addition to the revenue Rob mentioned, Overture has always been ahead of Google in providing better management and reporting tools to PPC marketers. I think it would be a smart move on Yahoo’s part to build this advantage aggressively over the next two years. We know search marketing is a hot market, and if Overture can build some loyalty, it will give them some high ground to defend their position from.

Q: Does Yahoo have a chance?

Rob: Microsoft’s marketing power is huge – as is their budget. Yahoo is also a big company, but they can’t compete head to head with Microsoft. MSN is going to crush Google and then aim for Yahoo. They’ll be competing portal to portal. MSN has to beat Google at search first. Once that does happen (and I do believe it will ‘ provided they can get that crucial version 1 search right), MSN will set its sites squarely on Yahoo.

Gord: As I mentioned before, I don’t think Yahoo can win the search battle head to head with Microsoft or Google. They’re going to have to rely on their strengths in other areas to survive.

Q: Does Google have a chance?

Rob: Not really, and that’s why Google needs a portal. They can’t compete on search alone. If it’s not a portal than Google has to offer other unique features.

Gord: The only way you can compete on search is to be best of breed. And Google isn’t the clear leader in search any more. To be honest, Google impressed the hell out of me in the beginning, but I don’t think they’ve done anything noteworthy in a long time. I think they’re showing the stress factures inevitable in a young company that suddenly has to compete in the big leagues. I don’t think they’re up to taking on Microsoft head to head. I can’t think of many companies that could.

Rob: But if Google is building in semantics and a constantly improving relevancy feedback mechanism, they should continue to improve. After all, they are already collecting data on the number of abandoned searches, the number of returned results, the average clicked on position and so on. It shouldn’t be too difficult to integrate this data into the ranking algorithms to make them better (if they haven’t done so already). Remember that if this is Applied Semantics technology being applied by Google, then the software is supposed to ‘learn’ based on what results were picked or not picked. It is supposed to be able to refine results for the next time.

Q: Does anyone else have a chance?

Rob: They’ll become niche players..fighting over the scraps that are left. And there will always be an element of the population who are anti-big business. Many people adopt Linux because they are anti Microsoft; I think you will see similar sentiments towards the search engines if they become too commercial. Already we are seeing signs of open source search engines, and other smaller engines trying to compete.

Gord: It may seem ironic, but the biggest potential losers in this will be the ones that go head to head with Microsoft, namely Yahoo and Google. The smaller players will probably benefit. The profile of search will grow, along with the profitability. The small players will be able to move quicker to identify niches and capitalize on them. And they’ll probably be able to strike new partnership, providing specialized services to the big 3 in the upcoming battle. As niche players, I think the future looks good for some of these services.

Q: When Microsoft enters search’will the industry grow?

Rob: Yes, search volumes will continue to grow, so we should see search continue to grow. After all, markets outside of North American and Europe are growing faster than anywhere else. In addition, broadband usage is growing, hardware prices continue to come down, and more people are getting hooked up to the ‘Net. There will be a point where internet growth tapers off, and search plateaus, but I think that is many years away.

Gord: I think we’ll see incremental growth in search as a whole, with a possible jump coming with OS integration. But I see exponential growth in the commercialization of search, and therefore the revenue generated. Implicit Query and other ‘push’ search channels will change marketing all over again. Search, in its eventual evolved and converged form, will be one of the biggest marketing channels, bar none, in 5 to 6 years.

Q: What will Microsoft do on the paid search side with the release of Longhorn?

Rob: I think this is where implicit query kicks in, and the sponsored results will be shown first. Consider this: How much would one ad across the tool bar in an application be worth to an advertiser? That advertiser essentially has a captive audience. We’ve talked about this before ‘ the application ‘watching’ what you are doing and offering help ‘ by way of sponsored (or other search) listings ‘ appearing conveniently in the application you are using. Another resource for listings could be the desktop sidebar (another of Longhorn’s new features). It is also built on XML so it should be flexible enough to display ‘best picks’ listings ‘ whether paid or organic. Combine this with Longhorn’s ability to learn from you and refine its ability to provide what you want and you have a powerful new advertising medium.

Gord: It’s a different pricing model’a different delivery technique. It would very easy for Microsoft to serve up as much advertising as they want, they have to know where they start irritating people. It will be a whole new paradigm, and it remains to be seen how people respond to it.

That said, Implicit Query changes all the rules for advertising. It introduces huge considerations for privacy and online intrusiveness. We’re getting to the online equivalent of the personalized ad message that we saw in the movie Minority Report.

Rob: But Implicit Query and result returned don’t have to be that obvious. Microsoft has already experimented with inconspicuous helpers. Remember Smart Tags? They are still used by Microsoft applications. They appear as a little box with an ‘I’ in it. There could be a time where Longhorn recognizes a phrase and associates it with a smart tag which is linked to a search result which provides more information via organic results, or paid search listing. This type of system opens the door to many different types of web marketing we haven’t considered before.

Conclusion
Microsoft, Yahoo and Google will take search in totally new directions in the next few years. That means search marketing will also change dramatically. The medium will become more powerful than ever, probably prompting new investigations and concerns by consumer groups, and new legislation by government. I hope the questions we posed and tried to answer help clear up a very fuzzy future just a little bit. One thing is for sure. The way you search today will be significantly different from the way you search in 2006.

Author: Gord Hotchkiss