Mimecast's bread and butter may be email management, but its founders' vision of an information bank is alluring. By Paul Fisher.
We live in a time of exponential data creation, a mechanism that can only be switched off by the catastrophic failure of all the world's IT systems – an unlikely event. Welcome to the digital age.
Too much data doesn't sound much of a problem alongside global warming or the current financial crisis, but it is still a problem. Creating data is easy. Storing it is harder. Storing and using it efficiently and intelligently is harder still.
It's a concept that informs much of SC's conversation with the two founders of email management specialist, Mimecast: CEO Peter Bauer and CTO Neil Murray. These South African expats have set up shop in the rapidly gentrifying area around Kings Cross St Pancras in London.
The interview with Mimecast proves interesting not so much because of what they do now – manage their clients' email, hardly a unique proposition – but more for their conceptual thinking and for where the company may be headed.
They say they can already produce remarkable efficiencies with their grid approach to computing – just 500 machines doing the work of 7,400. When I comment that 500 seems a small number, they produce the maths to prove that this is possible.
“We're achieving a result that arguably is better and more robust. It's greater than the sum of the parts for applications we've converged,” says Bauer.
“When we started the company, we set out with a technological ideology to see if we could take a broad set of standalone email applications and build a unified software platform that could perform all of these functions. When you're virtualising for a very specific function and putting that on a grid format you get a great efficiency and you get great scale-out capability as well. So that's been the hard work that's gone on for the past six years,” Bauer adds.
The use of the word “ideology” is no mere throwaway. As the interview progresses, it's apparent that these two founders are intellectually smart, as well as technologically on the ball.
Murray steps in for his CEO and expands on the theory. “It's not trivial, because you have to have something that doesn't have boundaries, doesn't have limits, that can just grow. Google kind of showed us a few clues, but it has a different problem to solve. Our problem is profoundly different. It's in the application – and that's where you have to engineer it, from bottom up.”
Coming back to earth for a second, in credit-crunch Britain, IT managers could be forgiven for saying they can do most of the email management themselves, despite the efficiencies that may well be on offer in the future from Mimecast and its ilk. So what makes it special?
Bauer picks up the thread: “Yes, there are other areas companies need to think about – you know, web security and the providers that focus on that.
“It's something we've chosen not to go and get involved in, so that we can really be extremely effective in the email and data-at-rest, data-in-use piece, which lots of security companies don't need to think about – they're just thinking about the data-in-motion thing. And the web piece is largely a data-in-motion discipline.”
So, finally, we get to the heart and soul of the Mimecast mission. It seems to be run by two deep thinkers who want to be more than the guys who securely manage email for their clients.
Suddenly, a fresh concept is bandied about, around the table in one of the Mimecast meeting rooms. Bauer casually says: “Well, Neil would refer to us as an information bank.” That's exciting, I say. This is more than email management. This is, possibly, very new.
“Look,” Bauer says, “email is a fairly complicated part of data management, but it's not the only part and it certainly isn't the part we want to stop at. We buy literally hundreds of terabytes a month. We have customers signing deals with us where they give us their data for an extended period of time. We have to think like a bank in terms of security. We don't charge them to put money in or take money out, but we charge for things that we provide as an additional service. It sounds like money, but it's data.”
Murray elaborates: “If I said to you, ‘where do you bank your data?', you don't have at hand a list of companies that do that. But we are becoming one of those companies and our thought processes are being guided differently about what it is that we do.”
Where Mimecast now sees its opportunity is in adding services on top of the data it banks for its clients – services such as document management, knowledge management, file archiving and backup services.
“I'd say 85 per cent of the companies that come to us have problems with their email systems. They're either under attack, they add too much volume or they are spewing blowback spam, because they don't know what to turn on or off,” says Murray.
“But when it comes to ingestion, we get a company that says, ‘here's my 40 terabytes of data; please absorb that in some useful way'. Phew!”
I suggest that this analogy with money and banks is fine and sounds sexy, but only as far as it goes. After all, money flow by comparison is controlled and limited, especially now. How will the Mimecast bankers manage all that data?
“People aren't going to store less, but what they need to be able to do is find that information – particularly businesses and end-users. Businesses are going to have to retain metadata about stuff that describes this information, so that it can be retrieved and used in context,” says Murray.
But, I ask the duo, isn't data management all about search algorithms in the end? Google has proved that. How good are Mimecast's search tools? Bauer insists that search is only as good as the index of the content and the metadata that describe that content, both of which can then be referenced through a search engine. He says that how you aggregate is what enables a better user experience.
“But if you look at email,” Bauer continues, “anything that's moved through email, if you've tracked all the email behaviour and you've indexed all the content, you've got a lot more context for everything.
“Because you can take a document. And you know this document moved, in this email, it moved from this person to that person, at this time, and then on to that other person. So you can start to identify user communities of stuff.
“You can then relate back and say, okay, well, this person is in this active directory group, because we've got all of that information as well, we integrate with that. Say we know something, that this person happens to be in marketing, that's interesting. This went to this person out there, what's the domain? Okay, well, it's wpp.com, you know, investigate that and go and pull back some information off that website potentially.
“It's not foolproof, but it all helps to say, okay, that thing probably went out to somebody in the advertising industry from somebody in your marketing department. You're building a lot more context. You've got access to a lot more metadata, because you've seen the usage behaviours. It's not simply a file that sits on a hard drive and gets opened and closed,” Bauer says.
Murray adds that this has changed his whole approach to search. He now thinks about “the document he sent to Glenn”.
So how does this add up to being the world's first contextual information bank? The conversation turns to the theory of calculating the risk and value of data – concepts Bauer is very keen on.
“One does need to deal with it like a currency, but with even more care than one would money. People don't look at money and say ‘that's less risky than information'. They look at money and say ‘let's put it in a metal box or in a vault somewhere'. People are more casual about information, and that's difficult to fix, because people can't see information.
“Our role as a broker of unstructured data has the best perspective on security, the best perspective on that value, and the context of that information,” Bauer continues. “We are able to service information requirements of other application contexts, such as where the end-user wants to consume that information. That's very much the work that we're doing.”
End of conversation. Plenty of theory, but plenty of potential too. Whether you believe in information banks and intelligent storage or not, this duo's dynamism is such that you feel a fresh way of managing data overload may be emerging from the heart of Kings Cross.
World city: why Mimecast set up in London
Why did two South African entrepreneurs choose to start up in London – and has it proved a good decision? The answer is reassuring and pleasing.
“It's a fantastic place to have set this company up. The commercial opportunity, the ability to access capital, the talent pool that's available to us in this market is just mind-blowingly empowering for a company like ours,” says CEO Peter Bauer.
But it's also London's world city status that has helped, he says, its ability to draw in talent from around the world. Mimecast has a truly international staff – with, not surprisingly, more than a handful of South Africans.
“Having this international culture, it's been very good for us to build out our global operations from here, because we've got people who genuinely understand these parts of the world,” Bauer adds.
“We also knew a lot of really smart people who moved into the UK, so over time we've kind of cherry-picked them,” he says.