Peter Bauer: My concern with that approach is it’s immature right now. It will be for a while. Since it’s unreliable, it’s also difficult to predict. You don’t know how information is going to be suggested to you. You don’t know what the underlying mechanism is. It can really be a distraction in the same way that something like Siri was hailed as someone super clever that can help you with all sorts of stuff. But I think a vast majority of people have left it alone and are waiting for some future iteration where Siri may be profoundly useful. It’s just not clever enough. There’s a trend towards that. I think it’s an exciting trend to figure that out. It certainly is not practical for business users.
The alternative way and the thing that we think about and are working on is this concept of facet-based browsing. It is very simple and it’s a part of the question “Suppose you were the most organized person in the world and filed all your information in a very structured way. Not necessarily just file it in one place at a time, you take each piece of content and then you’d file it in different categories, what might that look like?” You can imagine a traditional directory structure when you open up your C drive and you’d have these folders where you put things into. The approach is organize the information that way and then present it to the end user and give them search ability within that context. At the same time, allow them to browse based on those facets.
What I experienced with that so far is users are able to retrieve much more value much quicker than a simple search-and-scroll. The reason for that is in the way those facets are organized. It provides a very good memory recall aid for them. They remember context. They don’t have to think up what those things might be and enter them in a search box. They’re seeing arrangements of data and they say, “Let me look at that.” That’s a simple example. So let me look at a company, “Those are the people that I talked to in the company. Let me look at the data I’ve exchanged with those people. Here’s a file and it’s about X. Let’s explore it.” It’s different ways in helping the human memory connect with digital corporate memory and human-generated data.
Sramana Mitra: Not entirely, but more of the CRM contact management function, right? Today, you have to go outside of your email to get that kind of context off a conversation. Assuming that you are logging everything into your CRM, then that would be the place where you would find that view.
In fact, I’m not sure if you’ve seen it. Salesforce just acquired a company today. I just saw it on Techmeme. It’s Relate IQ. They’re a Big Data startup and Salesforce just acquired them for $390 million. What they say is that this company tries to draw those correlations and automate a bunch of the relationships and do things that you’re supposed to do in a sales process in a more automatic way by drawing those relationships. Relate IQ uses searches of unstructured data from emails, social networks, and calendars to automate large portions of the sales process. I have not looked at this company. I didn’t even know about this company actually.
Peter Bauer: That’s exactly the kind of thing that we’re interested in and see huge opportunities because of the vast amount of data that companies entrust us to keep. Providing those analytics on top of email when it’s in its native format inside an exchange server, you end up copying stuff out and hammering exchange service to get the insight.