When Rob May founded Backupify, it had a completely different name: Lifestream Backup. His original intent for the backup service was to help ordinary Internet users save backup copies of all the content they were putting onto sites such as Flickr, Photobucket, Friendfeed, and Twitter.
But the more traditional backup business turned out to be more lucrative, and thus the company rebranded as Backupify. In this phone interview, May discusses the importance of backing up personal online data, especially at a time when even quite major technology products and services continue to go under, as was the case for Google Reader, and which is impending for task-management app Astrid at the end of August.
May also imagines a future in which we have immediate and seamless access to our entire personal data collectionthe ability to not just recall conversations, but actually replay them or display data we've capture during them.
Jill Duffy: Backupify started as a "life-streaming" backup service, but since then it's grown to be a more traditional backup service. How did that happened? I thought it would be more natural to go the other way, to start is a core backup service and grow into things like social media and personal photos.
Rob May: It's a great question. The gem of the idea came from a friend of mine who want to back up his Flickr account. This was back in late 2008. I talked to some friends, and everybody had a set of things in the cloud that they wanted to backup.
We positioned the company as a service that backed up your life stream. And so the name of the company was Lifestream Backup. The reason we decided to change happened when we looked at our early paying customers. For us, it made a little bit more sense to move to the SMB model.
JD: It sounds like it was more of a business decision than following the trend.
RM: Somebody will surely build a successful version of what we were originally trying to build: a service that backs up all of your social services.
And it will probably go the same way other backup companies on the PC side have gone. Once it backs them up, it will try to do other things with [your data], such as make that data accessible and in more formats.
We built a product with Symantec called Norton Ditto [which backs up some life-stream content]. It's a beta product out of Norton Labs, and it has some really nice formats. We can make a PDF of your Facebook profile, your Twitter stream and things like that. It's a pretty cool. I think that's an example of where that market will go.
JD: You mentioned that Backupify first grew out of wanting to back up Flickr photos, and now you've mentioned Facebook streams and Twitter feeds. What other kinds of data do you see people wanting to collect?
RM: Email is the biggest because it's important to so much of our lives right now. Photos are probably next. Beyond that, it depends.
One of the things that's surprising is note-taking applications, apps like Evernote, Remember the Milk, those kinds of things. People are putting a lot of their lives into those kinds of applications.
And then beyond that is something we haven't done, and that's text messages. A lot of people are sharing so much of their lives via text now that it's out there and there's no easy way to archive, and categorize, and go find them. I expect to see somebody pop up and fill in that gap.
JD: What about services such as Evernote? One of my assumptions as a user is that because it's a cloud-based service, all that data are by their very nature backed up. But I know you're concerned with that using a cloud service does not necessarily mean that you have a backup. Is that true?
RM: It's a confusing topic. It's an area that people don't always think clearly about if they don't come from the IT world.
There is a difference between disaster recovery and backup. Most of these cloud services are prepared for disaster recovery, which means if a server fails or a hard drive fails, they have another one that can take its place. They have the data duplicated in another area. If the data center goes away because of a hurricane, there's another data center that you can go to.
Google's recovery services are great for Google's disasters, not your company's disasters. So while, yes, if Google's server wipes out, that's Google's problem and they have server backups to deal with that. But if Mary in accounting deletes a customer list, or an invoice, or an email that's important, I don't know how you'd call Google and get that back.
JD: That brings up an interesting point. Finding something after it's been deleted or after it's been lost, whether it's in a backup or it's on your primary memory device, is very difficult. It sounds like what you're saying is services like Backupify take that into account.
RM: Yes. Yes, that's very true. It's interesting. The problem of really trying to find what you have A lot of people use Backupify because, depending on the way you like to search, we may have better search than the application that you have backed up. So for example, we let you search by date range. I know email lets you do that too, but other applications may not. That's an advantage.
For a lot of people, we're also an aggregation point. We don't support Dropbox or Box today, but we will within the next year, and we do support Google Drive. So, say you have a file, and you can't find it, and you can't even remember if it's in Dropbox or on Google Drive, or who shared it with you. You could go into Backupify as a single point of contact and be able to find it.
One of the interesting things about backup across cloud services is it's actually much more useful than just backup. Cloud services are siloed. Being able to aggregate that data across those silos in various ways has some value.
JD: Earlier this year, Google Reader shut down. And the task-management app Astrid did, too. All the people who used those services were left in a lurch because they had to find replacement services, and figure out how to export their data and which services support the importing of that data. How is backing up a strategy that protects your from that kind of situation? Or what happens if you stop paying for a service and the company yanks the plug and deletes your data?
RM: As this evolves you'll see a couple of things. You'll see more normalization around data types and structures in the cloud, and that will help ease the movement between applications.
One thing we talk about a lot at Backupify is "What is 'beyond backup?'"
I think the ability to migrate data between applications is a common use case. One of the big use cases when we were very consumer-focused in the early days was this idea that, "Hey, these services shut down. They go away if they can't monetize." When Facebook bought FriendFeed, we added a FriendFeed downloader just for that specific use case. We had several thousand requests for that.
There's no technical challenge to moving it around. There's a business model challenge. Even though your average person wants that data and wants to keep it, your average person is reluctant to pay to put that data into a new service. Businesses, on the other hand, pay a lot for migration. It's a huge business on the on-premise world, and I have no doubt it will be a huge business in the cloud world some day.
JD: What do you think people will be doing with these huge caches of data that they're storing now in the future? I have these daydreams of people who are akin to the genealogy experts of today, building together pictures of people's lives and their histories based on their data. What do you speculate on that end?
RM: Two things. Number one: I think we're going to see artificial intelligence advanced to the point where we're going to be able to augment our brains so that we don't have to remember so much.
I'll give you an example. Say you're talking to somebody and you think, "Oh, let me tell you about" and it might be a restaurant or an idea or a bookwhatever it is. Or, "my friend told me about what was the name of it? Crap. I can't even remember who told me." But you'll remember some context, like, "We were sitting at that coffee shop," or "it was somebody from work."
You'll have a partial recollection. The ability to go find that informationyou can't search for thatbut to have a service that would let you ask, "What was that book that Fred told me about over lunch a few weeks ago?" Imagine if we have tools that can record your whole life, and you could go back and recall a conversation or recall the tweet, or the note, or the email, or whatever, and find it. Right now, if you don't know the name of something, it's hard to find.
So I think one thing you'll see is automated agents that access our data and do things to it that make it relevant to you at certain points in your life in time when you need it.
The second thing I think we'll see is a closer augmentation directly with humans. I'm a big fan of Ray Kurzweil's The Singularity. I think it's going to happen. If you look out a decade or maybe two decades, the brain-machine interfaces will bring new advancements. I think we're going to be automated with chips that are going to be able to call on our own big data libraries. Everything that's ever happened to us we won't keep in conscious memory, but we'll easily be able to access via these interfaces if we want to.
It will change a lot of social relationships. I'm a believer that many of our social relationships are built on little white lies, or that we don't remember things clearly, or stories that grow bigger over time, or legendary stories. And all of that will stop, because everybody will remember how something happened.
JD: Or they'll have all the notes from their perspective, at least.
RM: It could lead to the entire breakdown of society [laughs]!