Carbonite Eyes IPO, Aims to Be the Symantec of Online Backup

(Page 2 of 3)

Carbonite turned on its service in May 2006, but they all charged by the gigabyte—and they all forced users to use custom software to choose which files they wanted backed up.

“When Jeff [Flowers, Carbonite’s co-founder and chief technology officer] and I did our original market research, we realized that once you get beyond the techno-geek crowd, most people don’t have a clue about where their stuff is stored on their PCs,” Friend recounts. “And what customers were telling us was that they wanted something more like car insurance: ‘I’ll pay you $55 a year and I don’t want to spend any effort on this.’ That’s what [the flat-rate pricing] was for—everybody interpreted it at the time as a pricing thing, but it was really an ease-of-use thing.” That’s also why the company didn’t build a separate user interface for the product, but simply worked from the existing contextual menus in Windows—because Friend’s research told him people didn’t want to learn yet another piece of software.

The flat-rate strategy had its risks. For one thing, Carbonite knew that its plan wouldn’t initially bring in enough money to pay for the thousands of disk drives it needed to equip its data centers, given the relatively high cost of storage back in 2006. “There was no margin in it at the time,” Friend says. “We were praying that storage costs would come down, because there was no way we could make money at $50 a year.”

Luckily, the bet paid off. “Disks have been getting cheaper at a rate of about 40 percent per year. We were also praying that broadband would become pretty much ubiquitous, and that bandwidth costs would drop as well. It was a very good time to launch the product, because all of those things have really come to pass.”

But while memory and bandwidth trends moved in its favor, Carbonite had to scramble to make sure it could keep track of all of the files consumers wanted to back up. Typical commercial file systems just aren’t designed to store tens of billions of files, Friend notes.

“Every time you double the size of your installed base, pieces of software that used to work stop working,” he says. “You have to scramble and write around it. I remember when we outgrew Microsoft’s NTFS file system [the one used in Windows XP, Windows Server 2003, and later versions of Windows]. We’d call up Microsoft and they’d ask, ‘How many files do you have on the system?’ and we’d say ‘Half a million,’ and there would be dead silence. Eventually they would say, ‘We didn’t have that in mind.'”

So all of the networking software within Carbonite’s data centers—the rules that determine how incoming files get spread out over the company’s disk farms, how to compensate for dead drives, and the like—had to be written and rewritten in-house. The company will bring its fourth major generation of file management software online this spring.

“We have more data than Yahoo Mail,” Friend says. “There are probably 10 data centers in the whole world with this much data.” Though Carbonite won’t reveal how many customers it has, Friend says the company currently stores about 39 billion files, and that over time, it’s restored about 3.2 billion.

The scale of that effort creates an inherent barrier against the emergence of a second major competitor, Friend says. “If somebody said they were going to raise $50 million and try to compete with Carbonite, they couldn’t. Until you get to this size and this number of customers, there’s no way to test it. You have to grow incrementally.”

As Carbonite itself has grown, it has stuck largely to its simple message about helping customers survive computer disasters. But it’s also added a few innovations along the way. Last March it finally introduced a Mac version of its backup software. (Friend says that was a two-year project, because the first version of the Mac client “looked too much like Windows,” forcing the company to start over. The lesson: “You don’t take Windows developers and put them on a Mac product—the Mac people can smell it right away.”)

At the same time, Carbonite introduced a remote file access feature that let customers access individual backed-up files over the Web, from any computer. This was much more of a stretch than rolling out a Mac version, because for the first time, it meant the company was … Next Page »

Single Page Currently on Page: 1 2 3 previous page

Wade Roush is the producer and host of the podcast Soonish and a contributing editor at Xconomy. Follow @soonishpodcast

Trending on Xconomy

By posting a comment, you agree to our terms and conditions.

  • Shaw

    I use the service as a consumer and find it easy and comforting. A friend uses it, had to draw on the backup and said the help was great

  • Sara

    As a consultant in the industry, I can tell you that the money Carbonite spends in marketing may pay off for their brand, but its clear their spending less on their backend operations. Lost data, slow upgrades, and slow restores indicate that in terms of technology, Carbonite is going with the same marketing idea: “cheap”.

    In order for a business to rely on a vendor like Carbonite, price shouldn’t be the only differentiator. Software capabilities and reliability, as well as support are also important. Without sophisticated software and hardware, the brand promise will get much weaker over-time.

    Other companies to consider for businesses are: Backup My Info!, Iron Mountain, and Evault.

  • Thanks for the profile – lots of useful information for understanding their growth.

    I’m curious what kind of metrics (i.e. revenues, subscriber counts) do you think they will need to execute a successful IPO?

    Also, I’m confused by the statement (or perhaps, implication) that because they have grown revenues consistently, they do not need more money. In a business like Carbonite’s, the cost of customer acquisition is likely the single biggest cost. While it may make sense in the long run, they may be operating at a loss in the early stages of a customers life – in fact, their impressive growth rate and marketing spend almost certainly requires the sort of capital they have raised (and will continue to raise). I’m not suggesting there is anything wrong with the strategy, only that I think it is probably not the case that the won’t actually need the money they’ve raised to continue the marketing and infrastructure development required to achieve the scale necessary for a legitimate IPO. Thoughts on that notion?

    Disclaimer: I’m affiliated with another company in the online backup and storage space.

  • Interesting that they are so very “old school” interruption marketing driven. I hope that they do become a home run company here in Boston. Just seems ironic that here in the hotbed of Inbound Marketing, they aren’t up to speed. Surely that will have an effect on profitability in the long term. Also, they should look over their shoulder for Dropbox. I know they think it is for nerds, but that is so wrong, just ask my mom, even she can use it.

  • @tsummit: Hey Tom. Yeah, my sense is that Carbonite thinks of their target market as a group that probably isn’t even using the Web enough for “inbound marketing” to be effective. But we may also see some changes coming in their advertising strategy. I will check out Dropbox, I don’t know enough about it.

    @mike: You raise an interesting question. My understanding is that Carbonite has about $35 million in the bank. I suppose they could easily spend all of that on customer acquisition (counting advertising) between now and the time of their hypothetical IPO in 2011. Which would require some careful timing.

  • David

    Carbonite has seen a lot of success but I think it’s missing features that other companies are offering. Has anyone reviewed or used Egnyte? I did their demo and was very impressed.

  • I agree that apathy is the biggest competitor. But for the apathetic computer user online data backup services provided by the likes of carbonite is surely the easiest & most automated way to backup valuable information. Hopefully they will realise this before its too late