why Google buys companies

via Google Blogoscoped

Watching Google from the outside – with the limited information that offers – it seems they buy companies mainly to get more:

• Data. Like articles, meta data, digital archives, photographs.
• Users. Or in more general terms, customers or market share.
• Technology. Mostly, that's software, like web or desktop applications.
• Developers. Or, in more general terms, call it employees.

Some of above items are interrelated; especially with technology and developers, there's not always a clear distinction.

Furthermore, Google sometimes invests in foreign partners out of legal or political necessity, like when they partner with Tianya or Ganji in China; or they may subsidize a company to skew the market in disfavor of competitors... e.g. when they pay Mozilla developers to progress Firefox, or pay Mozilla when users search Google using Firefox, to balance the market against Microsoft Internet Explorer. While there isn't a lot of evidence it may also be possible they sometimes buy companies just to silence a competitor, or to prevent a competitor from acquiring it and growing too strong. The philanthropic arm of Google, called Google.org, also invests in green energy and more to improve the world at large. But again, as we don't sit inside Google's strategic company meetings, much of this is just speculation.

The end goal of acquiring a company may be aligned with Google's overall mission. We can paraphrase it as 1) grab all the world's data, 2) make that data useful and accessible in order to direct user attention towards it, 3) profit from ads displayed with the data. Google's hardware business aside, the company indeed just sells attention. If you look at it from a bird's view, you could perhaps split up Google's goals into the two philosophies make money (more the manager or ad sales types) and build the ultimate AI (the developer types). Ideally, these two camps work hand in hand, as the ultimate AI would generate the ultimate user attention generating loads of money.
This whole mission is then accompanied by a moral framework that started with "do no evil" and changed to "do good," a more traditional but less powerful construct as it transcends from a restrictive consideration to a potentially restriction-free justification... a self-referential "do Google."

Looking at past acquisitions

Here is a limited selection of Google's many past acquisitions, checked mainly against the four parameters data, users, technology, and developers:

• In 2001, Google acquired Deja's usenet archive. This seems to have been done simply to get more data... the archive contained over 500 million discussion group postings, Google stated back then, saying that "Usenet and its thriving community is one of the most active and valuable information sources on the Internet." Today, Google displays advertisements next to these messages. They also utilize the archive as a bit of a lure to get people to sign-up with Google Groups, a discussion group tool merging usenet and Google's own groups software. A good groups search was also beneficial for Google's search engine soul, though they removed the link from the main homepage links after some time (currently putting it in the "more" menu on the homepage).
• In 2003, Google acquired Pyra Labs/ Blogger. Having a working blogging technology might have given Google a chance to be ahead of competition, but I would find it hard to imagine that was the actual reason they acquired this company. Moving the software, which worked sluggish for years, to their own system might in fact have caused more trouble than building an in-house application would have. Rather, it's more likely Google simply acquired blogger to get the existing users (all those who signed up with Blogger and produced posts) as well as the data (every word written on a Blogspot blog resides on Google's servers, so they may more easily index and analyze it).
• Also in 2003, Google bought Kaltix. Larry Page at the time stated, "Kaltix is working on a number of compelling search technologies, and Google is the ideal vehicle for the continued development of these advancements." This seems to be an example of a technology as well as developer acquisition.
• In 2004, Google acquired a stake in Baidu. The shares are sold by now, but finding a local partner in China might well have been out of political reasons, indirectly triggered by legal restrictions of not being able to simply offer their own technology due to China censorship.
• In 2004, Google acquired the Picasa desktop software. The application is for photo management. I find it hard to tell what this acquisition was good for; the technology of photo management seems trivial in relation to some of Google's other undertakings. Perhaps this was a bulk developer acquisition, in terms of quickly adding the right team for Google's photo management strategy. In a press release from mid-2004, Google argued, "Picasa is an innovator in the field of digital photography, and we're excited that the Picasa team is joining Google." Out of Picasa's offline suite at Google grew Picasa Web Albums, a product not unlike Flickr.
• In 2004, Google also acquired Keyhole Inc. In a page copyrighted to 2004, Keyhole advertises its product that was later turned into Google Earth: "Keyhole 2 LT is a software application that you download and install. It's only 9MB, but with an annual Keyhole subscription, you can fly through 12+ Terabytes of Earth imagery and data – spinning, rotating, tilting, and zooming. Think magic carpet ride." More than a data acquisition – Google still needs to license satellite data from e.g. Tele Atlas, DMapas, Navteq, Geographic Data Technology Inc, MapData Sciences, Georoute IGN France – this was possibly a developer and technology acquisition.
• In 2005, Google snapped up Dodgeball. Like the addition of many other mobile companies – Reqwireless, Android, AllPay, Zingku – this seems to have been Google trying to get a foothold in a market area where they might not feel positioned strong enough. In the case of Dodgeball it seems hard to imagine they did it for the Dodgeball user base; perhaps it was just to get the Dodgeball developers on board to speed up Google's own mobile projects (existing Dodgeball technology may have been a reason as well). The Dodgeball founders weren't too happy with what happened to them at Google, though. When they called it quits in 2007, they gave the company a thumbs-down and said, "It's no real secret that Google wasn't supporting dodgeball the way we expected. The whole experience was incredibly frustrating for us – especially as we couldn't convince them that dodgeball was worth engineering resources, leaving us to watch as other startups got to innovate in the mobile + social space."
• In 2006, Google acquired YouTube. According to an ex-Google employee, many people inside Google were surprised at the acquisition of the video site at the time, saying "But they have no technology!?" Google already had the technology in-house, named Google Video (superior, too, just considering technology and not the social side). However, YouTube was also ridiculously popular with many people, and a whole lot of YouTube videos were embedded in blogs, directing a whole lot of attention towards them (with attention being Google's main product, in a way, as mentioned above). So this was both a video data and video user acquisition. And considering the YouTube team continues to work partly separated, perhaps it was also a developer/ employee acquisition – Google buying a company that "gets" communities. Today, YouTube continues to thrive and YouTube video results are more visibly integrated into Google web search results (receiving special rating stars in the seemingly neutral web search results). The Google Video program director Jennifer Feikin, on the other hand, left the company last year.
• In 2006, Google also announced the acquisition of Neven Vision. The company focuses on image object recognition. As such, it is most likely to have been added to Google's repertoire due to technology and Neven Vision developers. The Picasa product manager in 2006 said Neven Vision comes to Google with "deep technology and expertise".
• Google also acquired Writely in 2006. This product was turned into the Google Docs documents editor, so it was likely a developer and technology acquisition. For similar reasons, Google acquired Tonic Systems to help develop their Powerpoint web app clone Google Presentations. With acquisitions like these, you get to wonder what would be the more successful strategy for someone to develop a product for Google: to apply for a job with them and then launch something in-house, seeing it potentially repeatedly shredded apart by the many management layers of this 16,000+ people company... or to just start-up your own company outside, and then get acquired by Google. If the latter would turn out to be easier, it's probably not the best motivator for Google employees.
• In 2007, Google started the acquisition of DoubleClick. This seems to be a good example of acquiring a user base, though here the word customer base – all the advertisers using DoubleClick – may be more appropriate. Also, with DoubleClick Google snapped up a lot of employees with good connections in the market, growing their web ad presence closer to the size of a monopoly.
• In 2007, Google also acquired Gapminder's Trendalyzer and the team. It seems the motivations to get more user and data can be ruled out almost completely for this acquisition. And while the technology of the software might have been a reason, I would guess Google was mostly interested in grabbing these Scandinavian developers. When Google's Marissa Mayer welcomed the Gapminder team to Google, she wrote that "[b]uilding flexibility into search, email, and other Google products is critically important".

Trends in Google's acquisition strategy

As the chess master once replied when asked about his favorite piece on the board: "My favorite piece is whichever wins the game." Google seems to be getting more and more pragmatic about acquisitions in terms of buying whatever advances their strategic interest. Formerly, the focus was slightly more on technology and data acquisitions, while nowadays it's also often about mere land-grab of user base. But even one of the earliest acquisitions – Blogger – was a lot about users, albeit in a more technical and geeky space (blogging) than e.g. the acquisition start of DoubleClick (hit-the-monkey ads). If Google starts focusing too much on land grab acquisitions though, they may get into legal troubles due to accusations of being a monopoly – just buying market share does not progress technology, which in the end hurts users.

When looking into the future to understand which companies Google might buy next, it still helps to check against all four main parameters. Some companies sole reason of existence seems to be trying to create "Googlebait," wanting to be snapped up by Google. This hardly seems to work because the goals of a company which has time to think about being acquired are apparently not sufficiently directed towards data, users and technology, perhaps due to a lack of great developers on-board.

Some companies also try to make it seemingly really easy for Google to acquire them, by already working a lot within Google frameworks: RememberTheMilk's todo suite (which works on top of Google Calendar) or the Zoho office suite (using Google Gears) come to mind. Indeed, Google may look at these companies in terms of adding new good developers to their own teams in that space. However, in many acquisitions Google may also want precisely the opposite, namely to increase their developer knowledge delta. Just buying a team that understands Google may not help Google tackle new problems.