Email Mountains Contain Gold, Says Attachment Inventor

Continued from page 1

What do you think the future holds for e-mail?
There are interesting things happening beneath the surface. There are two main areas I would identify right now. One of them is dealing with the power of spam and malware. It’s almost an insolvable problem, because it’s the one area where Moore’s law is on the side of the bad guys. I think it is going to fuel continuous innovation.

Value from the information bank

The other area where I think e-mail is going to see a lot of growth is in the back-end. That’s what my company Mimecast is doing. We provide cloud e-mail services for business, and at the core of it is e-mail archiving. Once you get all of a business’s e-mail into a single archive in one place, you have a resource that really never existed before.

Even in mighty IBM where I used to work, the archives would be scattered on servers all over the world. Once you have them all in one place, you can start to write applications that deeply mine the information that is hidden away in e-mail archives. A study suggests that up to 80 percent of a business’s information can be found in its e-mail. Now you can start to use this information to provide insight that will be helpful to business.

Imagine that as you are typing an e-mail, you have a sidebar. And the content of the sidebar is changing as you type. It is using what you type as a search key, and it is searching your company’s mail archive. And if you are sending an e-mail to a customer, the links come up as you type, maybe with the e-mails you sent to this customer three years ago, which you forgot about. A salesman might be sending an e-mail to a new prospect, but before he hits the “send” button, find out that three other salesmen have tried and failed with that prospect.

And then there’s the question of permissions. If the system would think that a certain employee needs to see a relevant e-mail, it could ask the author of the e-mail for a permission automatically. The author would then decide if he wants to share data.

These things are a solution to a major problem that organisational theorists have talked about for decades, “organisational memory”. The basic problem is: you need information and you know someone in your company has it, but you have no idea who. I think we have a real shot at making a dent in it. We can say “gee, this guy is working on “X”, he should see these seminal documents on “X” that somebody else owns”.

My CEO at Mimecast likes to talk about archives becoming information banks. An archive is basically just a safe deposit box. You put stuff in there, it sits there forever, and every now and then you look at it. But a bank doesn’t just store your data, it also pays you interest. In this case, the interest is paid in information.

Spotting communicators… and conspirators?

Another example of good use of e-mail has to do with organisational communication patterns. A lot of people have been trying to visualise how employees communicate within a company. It was noticed that with large organisations, you have clusters – geographic and functional, and a relatively small number of connections between them. Between those groups, there were a few people that were the key connectors.

Researchers at IBM went a little deeper, and found that those key connectors have something surprising in common: they all get bad performance reviews. They are spending so much time communicating, that they are producing fewer deliverables. And yet these people are very important for business. Until now the communication function wasn’t really measurable. We could analyse e-mail archives and send reports timed to the performance review, that show less obvious benefits of key communicators.

And then there are other things we can find by searching archives for patterns. I believe that by looking at email you could spot something like HIPAA (Health Insurance Portability and Accountability Act) violations in the US, as they are happening. A professor at Carnegie Mellon in Pittsburg was analysing the Enron corpus, and she believes that she found patterns of communication that are indicative of where conspiracies occur. If this is right, you could imagine that we could provide reports to security officers of possible conspiracies inside the company.

These are the kinds of things that we are empowered to do simply by getting all of organisation’s e-mail in one place.

How well do you know the language of the Internet?  Try our quiz