How Microsoft can beat Google and win the Search Game
It sounds like everyone wants to beat Google these days. Great goal, but how do you do it? Everyone seems to think they have the magic bullet! Microsoft was about to buy Yahoo! for over 44 billions of dollars! They are investing billions in R&D. But do they have the right approach? Some people think you need more web pages indexed. Some believe you need a better interface. I tend to think the answer is right under everyone’s noses, they just need to use a little bit more passion, focus, to get the job done. But what should they be focusing on? What should the vision be?
What I propose below will take a huge team of brilliant people, a lot of money, and a lot of passion and insight into what is really going on with the Internet. Seeing as how I have limited resources, and no huge team to build this, I figured I would publish my thoughts on my blog, and see if any company out there would take on the challenge of building the next evolution in search. With the coming economic downturn, hundreds of thousands of web 2.0 companies are about to be wiped off the planet. Companies that rely solely on the Internet for their survival are going to have to bunker down. Even mighty Google has seen a fairly good beating in the last year, seeing their stoke price that was over $700 a share a year ago plummet to almost half at $386 a share. The free food is already disappearing.
Microsoft however, holds on strong because they are extremely diversified, and do one thing really well. Software! They have an incredible amount of money, and incredible army of brilliant people (nearly 100,000 employees). They also have some of the most incredible visionary leaders on the planet at the helm of the company. If anyone can create a better search engine, it’s Microsoft – Don’t argue with me you know it’s true! Will Google ever go away? No. But they could sure use some healthy competition to force everyone to innovate quickly!
Our Social Graph
The Social Graph that Facebook always talks about is extremely powerful. Facebook knows a lot about you. Other social networking sites and tools know a lot about you. They know who your friends are, where you live, and where you work. They have pictures, videos, and posted items that you find interesting. They even know what social groups you are a part of, and what events you plan on attending. Trust me when I say “Facebook” knows more about your friends likes and dislikes, than you do.
What’s Wrong With Search Today?
I find it absolutely absurd that when I do a search for something like “GAC”, I get the Geological Association of Canada. I am a software developer. Facebook knows it, Twitter knows it, and LinkedIn knows it. Google for some reason seems to be clueless. There is very little relevance detection built into the search. If my mom does a search for something, or if I do a search for something, Google treats us exactly the same. My mom loves crafts, and I love software development. We should get completely different results.
Why has nobody taken advantage of all the social graph data out there? Simply copying Google, re-branding it with a different logo certainly won’t work; people need a reason to switch. Adding context to search will give them that reason. Let me explain.
Let’s say a search engine were to partner with Facebook. What a minute, Microsoft did.
As Bill Gates has mentioned over and over in the last few years, search has a long way to go. All the demos I heard about a few months ago (new interface for searching with never ending scrolling, etc.) today quite frankly are not what Bill Gates was talking about. Scrolling endlessly so you don’t have to page? Not really an innovation.
What they need to do is a demo where they perform some searches with Google, and some with Live Search, and the Live Search results are better every time. And not just slightly better. Enormously better!
So What Can Microsoft Do?
It’s clear that Microsoft has the best software developers, architects, and engineers in our industry working for them. They also have a lot of money to invest into search, which was made clear when they tried to purchase Yahoo!. What they are missing is also very clear. While they know they want to win, they need that vision and roadmap to get there. They need the battle plan. Microsoft knows with Windows, Internet Explorer, and Office that it’s not about being first. It’s about creating an operating system that is better than the other operating systems out there. Creating an office suite that is better than everything else out there. It’s about creating a web browser that is easy to use and does what you want it to do. And lastly it’s about creating applications that all work together seamlessly.
So let’s put together a simple strategy that if executed correctly, could see Live Search gain incredible market share very quickly.
Live Search and Facebook Connect
Microsoft should be taking advantage of Facebook connect and connect Live Search with Facebook Connect. Using Facebook Connect, Facebook users could connect their profile data and authentication credentials with Microsoft Live Search. By connecting search, with the users profile data, you suddenly have created what I would call the ultimate search engine. And since Live Search is coming to Facebook already, having Live Search provide users with extremely relevant search results would be really powerful for users.
And it’s not just about context with Facebook. This is merely an example.
Microformats are the future of Search
Support Microformats better than any other search engine. Start with hCalander, hCard, and hReview. If you haven’t heard of Microformats yet, you should really read up on what they are, and why they are so important. Up until now, most of the data on the Internet has been completely unstructured. You perform a search for “Contact Miguel Carrasco”, you get my blog but that’s about it. You should really get a contact card link that you could import into Outlook, my phone number, and a picture of me. After all, that’s what you’re searching for, not my blog. Here is an example of an hCalander Microformat being used to create an event for our Winnipeg.net User Group event last week.
1: <div class="vevent" id="hcalendar-Winnipeg-.net-User-Group-September-Event">
2: <a class="url" href="http://dotnetwired.com/">
3: <abbr class="dtstart" title="2008-09-30T06:00-06:0000">September 30, 2008 6</abbr> –
4: <abbr class="dtend" title="2008-09-30T08:00-06:00">8am</abbr> :
5: <span class="summary">Winnipeg .net User Group September Event</span> at
6: <span class="location">17th Floor - One Lombard Place - Winnipeg, Manitoba, Canada</span></a>
7: <div class="description">What could possibly be better than enjoying some free pizza and pop with your peers while be entertained / educated by a presentation on a single .Net topic? Well, how about an open forum that includes some of the hottest topics in software development to date?! To keep the meeting energized, we will be limiting each topic to 20 minutes, and what's more, each topic will have a subject matter expert on hand to facilitate the session.
8:
9: Come prepared with questions, project stories, and ideas to one of the most unique user group sessions we have ever had.
10:
11: Topics will include:
12:
13: What is BizTalk
14: A Real World Silverlight Application
15: What is NHibernate
16: Why Continuous Integration Is Critical
17: Open Forum Free-for-All Session</div><div class="tags">Tags:
18: <a rel="tag" href="http://eventful.com/events/tags/winnipeg">winnipeg</a><a rel="tag" href="http://eventful.com/events/tags/user%20group"> user group</a><a rel="tag" href="http://eventful.com/events/tags/.net"> .net</a><a rel="tag" href="http://eventful.com/events/tags/microsoft"> microsoft</a></div>
19:
20: </div>
A search engine that supported this Microformat could display this information right in the search results, and have a link to the web site to register for the event.
The Complete Search Interface
Something everyone seems to forget is that when Google came out, there was really only web searching going on. Blogs didn’t exist, video wasn’t really that big, and Facebook and other social networks hadn’t even been heard of. The future of search is context. People spend hundreds of hours a month on social networks, news sites, microblogging, and blogs. They are constantly feeding these systems with an incredible amount of data about themselves. What do they like? Who are their friends? What are they doing next week? What kind of mood are they in? Where are they traveling to in three weeks?
I use at least 20 different “social” services on the Internet in any given day. Wouldn’t it make sense that a search engine could not only search all this data for me, but also use the data to help give context to searches I conduct. For example, let’s say I’ve been tweeting for a few days that I am going to Miami in a week:
On Facebook, I’ve created a few events that I will attend while in Miami. On Digg, I have dugg a few stories about some new scuba gear that was just released. On Last.fm, I have created a few playlists of music titled “Miami Plane Ride”. On Facebook I get a few wall postings from friends that say I should check out a few key bars and dance clubs in Miami. Yet another friend has suggested a few beaches they like, and posted some pictures of the beach.
Now, current day, If I were to go to Google and do a search dance clubs, the top three searches have nothing to do with anything I am looking for. I get:
Like I said, none of these searches have anything to do with what I am looking for. Maybe I am being way too hard on Google. I should have typed in Miami right? Let’s do that to see what I get:
Now I at least get search results that show me Dance Clubs in Miami. But which one is the one my friends told me about? Why didn’t that one come up right away? Why didn’t their pictures show up, and maybe the video they posted. Where are their comments? Instead of being presented with information from a trusted friend, I get links for web sites that have a high page rank because they have a lot of links coming to their domain name.
With Live Complete Search, searching for “Dance Club”, would have instantly added context to my search, narrowing my results to Miami. It wouldn’t do this just because I was planning on going to Miami, but because it knew that I was just twittering my friends asking them what good Dance Club’s in Miami where. Therefore, performing the search, would have loaded up the following result:
You’ll notice the first result is Nikki Beach, the beach that one of my friends told me I should check out on Facebook. A picture is also added to my search, and the comment they left me is also visible right on the search page. The phone number is also provided to me using microformats, and I can also get additional details on the location using Twitter, Digg, or Facebook. Furthermore, Live Complete Search knows I created a Miami playlist on Last.FM, so it is placing a link to that playlist right in the search.
Of course this is merely an example, but I think you start getting the idea.
Building Live Social Profile
It’s fairly amusing to see all these companies trying to clone every other company’s ideas, when Google has already proven that you don’t really need to own all that data, you just need to be the one presenting the data and helping people find it. Microsoft has the number one IM client on the internet, MSN Messenger. Every single one of those users has a Live account. Some of them also have a Live Spaces account. I don’t know about you but Live Spaces are not really used that much. A quick click through some of my most active MSN users shows me that they in fact have never used their Live Spaces, or if they have, use them 1/100th of the time when compared to Facebook or other sites they do use.
My proposal entails turning Live Spaces to be just that, “Live” Spaces. Allow users to create a custom portal that meshes together their entire Internet social profile. doing so would create the ultimate social profile, and social “my page”.
The Proposed Interface
The proposed new Live Search interface is very simple. By default it offers the complete search capabilities of searching everything on the Internet and placing it right within the search results. For example, using complete search, you could get some results from web pages, a few blog posts, and a couple of YouTube Videos. You might even get a few Digg articles, and some twitter chatter on your search page. Want to only search your social graph? No problem, just click on Social, and the Search results instantly filter to only show social results.
Many filters could be added to the Live Search interface, including “Research” or “Children”. Schools could lock down their computers to only allow for Research searches, and parents could lock down their kids to only use the Children approved sites. By connecting all these disconnected systems through the Live Spaces and Live Profile, you create something infinitely more powerful than Google PageRank, and the pendulum starts swinging in the other direction.
Final Thoughts on Search
Hopefully you have come to realize merging social context, microformats, and a new interface that provides the entire web at your fingertips will be the next leap in search. Having search results that are 10 times more relevant based on social graph algorithms vs. PageRank and PageRank clone algorithms will dominate the web. I should note as well that this article only lists three out of the 15 items that I feel should be added to search to make it leap to the next level. It’s just way too much to write for one blog entry.
Microsoft you want me to switch from Google to Live Search? Provide better search results, and this is the way to do it.
If you liked this article, please share it on DZone, del.icio.us, StumbleUpon or Digg. I’d appreciate it.
Excellent post Miguel. You bring up a some good points on the shortcomings of generic search engines. While Web 2.0 has brought us so much personalization in many other areas of the Internet, the search engines still aren’t providing us with personalized results. You may be right; Microsoft may be the only one that could take it to the next level, but I would like to see someone else step up to the plate on this one. A smaller, but up and coming social network coupled with one of the smaller search engines could really blind-side the big boys.