Blog Post

Why You Won't See a Google-Akamai Deal Anytime Soon

Shares of Akamai (s akam) got a nice little bump earlier this week, partly due to rumors that Google (s goog) was interested in buying the company for its CDN business. But does such a deal actually make sense? Not really.

Yes, CDN (Content Delivery Network) technology could help Google deliver YouTube videos faster or serve up search results more quickly. But if that’s what the company is looking to buy, there are smaller, cheaper, safer acquisition targets out there. Here are the top three reasons why a deal probably won’t happen.

1. The price is too high. After the run-up in Akamai shares earlier this week, the company’s market cap stands at around $3.8 billion. That means Google would have to offer about $4.5 billion just to get Akamai interested in taking a phone call, analysts say. While Akamai is clearly the leader in the CDN market, it is by no means the only player. And if what Google wants to buy is CDN technology, it can get that for a lot cheaper than what it would cost to acquire Akamai.

“There are other companies out there that have proven you can build your own CDN for a much smaller price than that,” says Kaufman Bros. analyst Colby Synesael. “Look at AT&T (s t), who spent about $75 million building their CDN out. And Limelight (s LLNW) is trading at around $350 million, which is one-tenth of what Akamai would cost. By no means do you have to spend 10 times Limelight’s share price to achieve that.”

2. CDN customers are high maintenance. Akamai’s $800 million in annual revenue is nothing to sneeze at, but there is no reason to believe that Google would have any interest in running a CDN business and managing its 3,000 customers. “You have to figure out why [Google] would want to acquire a CDN, and from my perspective it would be more for the network than for its customers,” says Wedbush Morgan analyst Kerry Rice.

That makes sense. After all, the CDN business — and especially the one that Akamai operates at the high end of the media and entertainment and e-commerce markets — is a fairly high-maintenance proposition with regard to the amount of back-and-forth that goes on between salespeople and customers.  But Google has thrived in creating low-maintenance, automated and algorithmic business decisions, and labored in cases where it’s needed to step outside that comfort zone. It struggled when it needed to sign up premium content partners to contribute videos to YouTube, for instance.

3. Infrastructure issues. A Google acquisition of Akamai would raise serious issues surrounding Akamai’s network architecture. Its CDN works by distributing content from 50,000 servers located in last-mile ISP networks around the world. But after years of fighting with carriers like AT&T and Verizon (s vz), is there any reason to believe that those telcos would want Google servers located in their local network facilities?

“Part of Akamai’s value proposition is that they have free access to last-mile service provider networks to access their own servers,” Synesael says. “But who’s to say that will be maintained if Google were to purchase Akamai?” Chances are that access wouldn’t be maintained, in which case Google loses out on a big reason for making the purchase to begin with.

Tech M&A may be on the rise again, but the Google-Akamai rumor is just that — a rumor.

12 Responses to “Why You Won't See a Google-Akamai Deal Anytime Soon”

  1. I just want to add something to what Aaron said.

    Everyone should think of it this way. Almost every really really large Internet company (such as Amazon) eventually learns that by building up your own network of computers and paying for tons of bandwidth, etc, etc… eventually you reach a point where lots of server space and cpu usage and bandwidth goes wasted during times of lesser activity.

    So, it only makes sense to recoup costs by selling part of your network to the masses who couldn’t possibly ever think to reach the same scalability and reach and power and storage. This, in turn, becomes yet another lucrative source of income.

    Google makes LOTS of money (most) on advertising… but they could lose this… very easily… they weren’t on top before and they rose to power through hard work and clever tactics… someone else could eventually do this same thing to them. They know this. They don’t want to keep all of their eggs in one basket. So, while trying to make sure they remain strong with what they know, they are also trying to enter other fields. They already charge money for additional storage. They are charging money for hosting site contents (App Engine). I expect, in time, they will continue to find new ways to charge money… while still providing free services where charging money up-front makes little sense, but earning money later on ad revenue could be achievable.

    Given this… it’s only a matter of time before they get into the CDN business completely. When that time comes, do you think ISPs are going to let them right in the front doors? No way! However, if they buy Akamai… do you think all of those ISPs are going to happily give up all of that money? Not likely!

    So, an Akamai purchase would be the Trojan horse approach (though it was someone else’s horse). I think it would be a very clever tactic from several standpoints. Mainly:

    1) Instant reach through a presence in many last-mile locations.

    2) Instant customers. Maybe they can sell many of these customers on cloud computing solutions as well.

    3) Instant information. Suddenly, the amount of traffic going through their “pipes” increases dramatically, giving them more data and more insight.

    4) Instant growth in intelligent employees. Akamai wouldn’t be where it is today if they didn’t have some pretty good engineers. Google is always happy about acquiring more good quality engineers.

    5) Acerno. This shouldn’t be overlooked. Acerno is a growing force… and its techniques are different than anything Google is currently doing, yet is something Google surely will want to replicate soon. What better way to replicate than just purchase and integrate? Google has done this several times… Writely… Postini… etc…

    Would Google just keep Akamai’s systems exactly as-is? Initially, yes. In the long run, probably not. They’re probably more interested in the rack space and the bandwidth positioning. Will they keep all employees? Not likely. They’ll keep the good ones, though. Will they keep Akamai’s technology the same? Initially, yes… but they will certainly be looking at the good points… comparing to their own… and using this knowledge to improve their own design, then implementing this hybrid design across the board.

    Google isn’t as perfect as some assume. They’ve had some pretty high-profile network outages that were rookie mistakes… but mistakes you only make when you are on such a big scale as they are. I am sure they can learn a thing or two from Akamai’s design… and I’m sure Akamai’s design can be improved upon by Google.

    In my opinion… this would be a perfect match. Whether the price is worth it is open for debate.

  2. Aaron deMello

    Raj, Om didn’t write the article, Adam Kramer did.

    Also, Google has a variety of servers in their environment, so there is no “network confusion” by “having multiple server designs.”

    The quantity of Google servers is irrelevant since they are all in Google’s own data centers, which are geo-distributed based on availability of low-cost power and cooling.

    The entire point of a CDN is to be as close as possible to the subscribers accessing the content; Akamai’s servers are co-located at ISPs… like TWC, Verizon, etc.

    So its very complimentary, and will likely save enormous sums of money for rich media such as YouTube.

    Travis K and StareClips get it, not sure the original author does.

  3. If Google is serious about serving up Big Media Copyrighted material via pay per view or subscription then buying up a CDN sounds reasonable. Buying up a CDN would allow Google to acquire an infrastructure to boost bandwidth for said “copyrighted material channels” in cooperation with Googles newly acquired On2. On2 VP8 is supposed to make H264 look like an mpeg1.

  4. Om:

    I agree with you that Google won’t buy Akamai anytime soon.

    Google has many more servers than Akamai ( These may not be located as close to consumers as Akamai’s servers are, but for the vast majority of the US and European population you can safely assume that there are more Google servers in the vicinity than Akamai servers.

    Google has a specific way of building its servers and network ( Buying Akamai would cause network confusion due to multiple server designs.

    Akamai’s profits are generated by charging SMB customers who don’t own a widely distributed network of servers. Big companies pay lower margin per byte for using Akamai (and therefore are marginally profitable for Akamai). Google can serve these SMB customers directly, by creating its own CDN service (if it really wants to).

    The CDN market is worth a few billion dollars whereas the online advertising market is worth tens of billions. Why make millions when you can make billions….. Google could launch a free CDN service if it found a way to generate incremental advertising revenue from it.


    • Jonesing T

      What you don’t understand Raj is that Akamai’s customers hate Google. Take Viacom and News Corp for example. They want nothing to do with the monopoly of search. They have the premium content and want to monetize it. Google buying Akamai would put premium and long tail content together. Which is why I don’t think it’ll ever happen.

      Microsoft is in coopetition with itself here. They use Akamai for a ton, but they also aspire to be in control of their own destiny. A bidding war for Akamai would add up to too much valuation for either company to justify. Yet they can’t ignore what Akamai has — the largest collection of valuable content in the world. Not Google, Not Microsoft, Not Yahoo can compete with what Akamai has in aggregate.

  5. If i was google (i would have billions of profit yearly) and wanted to get gobble up a CDN company, why settle for a lower ranking player? If I wanted to get into that industry I would go for the industry leader, not second best.

  6. You really missed the obvious:

    1) Google is, to some degree, already building out their own CDN. If Google has figured out how to provide CDN services for CHEAPER than Akamai provides these same services, it would make sense for Google to acquire Akamai… gain those customers… switch those customers over to their cheaper infrastructure… then continue to collect the same money, reaping the huge benefits. Approximately 6% of ALL Internet traffic goes to Google’s servers, so they know a thing or two about handling massive load.

    2) Google can eventually push to convert CDN customers over to Google App Engine customers, giving them additional ways to monetize the existing client-base of Akamai.

    3) Google has become the expert at setting up data centers around the world. As they make their push into cloud computing, they could use more data centers. What easier way to pull this off than to purchase a company that is also known for its wide range of strategically positioned data centers?

    4) Google is on the verge of releasing the long awaited Google Drive. As they start providing gigabytes of cloud-based storage for anyone with a Google Account, it makes sense to be sure they’ve got lots of servers with lots of hard drive storage distributed in lots of places.

    5) As Google makes the push with Google App Engine and eventually Google Drive, they are getting closer and closer to competing directly with Akamai. It’s only a matter of time. It would make more sense to get rid of their future competitor before it is obvious that Google will eventually be a CDN provider.

    6) Akamai owns Acerno, a quickly growing advertising network which offers services that are vastly different than what Google currently offers. An Akamai purchase would be another way for Google to grow its advertising reach.

    • . Google just bought ON2 with Hantro. These are the cores that are behind the ARM, cores that power all the smarter phones. These encoding, codec, cores and VP8 video platforms power the •CCTV (official Chinese TV) and many of the flash video revolution seen buy most flash using sites. Dont think Goog dont understand iinfrastructure , China and world global growth. The US is only the slower growing country.

  7. Travis K

    You didn’t address the data angle.

    Akamai sees 20% of the world’s Internet traffic (figures from Akamai). That means they have a ton of data on users’ activities and behavior online. And the data Akamai has is exactly the data that Google does not have – big media and commerce companies’ user traffic. Akamai is already turning this asset into an advertising play. I’m sure this data gets Google M&A boys licking their chops.