TR: In the US Gmail is third by market share. What's the objective with Gmail now – topple Hotmail?
Jackson: We tend not to focus too much on competitors, we focus on users, and the needs that they have. Growth is one of the things that we care about. We hope that more and more people will use Gmail – and not just that more people will use it but that people will use it more often.
This is something that Larry and Sergey, the founders, tell us – focus on usage not users, because the people who are using your product the most, the most active demanding users are the ones who are going to be helping inform you about what the future is.
The leading edge users are where everyone else is going to be two to three years later. So they are the ones that can help push your product to where it needs to be for the future, so we're really focussed on innovation.
TR: But if you are looking to grow, stuff like a lack of folders can be a deal-breaker for new users…
Jackson: We worked on this recently with our project where we improved labels. We wanted them to be accessible for people who are familiar with folders. We believe that the label model is a good model because it allows something to be in multiple labels. If I receive an email that's from my family but it's about a vacation I want it to be both in the 'family' and the 'vacation' label.
At the same time we realised that most people just didn't get labels. All our research showed this. And so we did this big project to improve them – while still keeping them as labels we wanted to make them familiar to people who use folders so you can drag and drop, you can 'move to' – which is essentially moving something out of one label and into another label so we made them sort of, still implemented as labels under the covers but look and feel and behave like folders so that users who are familiar with folders could use it.
And after we launched this we saw the usage of labels go way up and we saw the number of people doing the traditional label commands go down a little and the amount of people doing the 'move to' and the folder-like commands go way up.
TR: Why isn't the search in Gmail as good as Google's web search – you need to be much more accurate with your search queries in Gmail.
Jackson: Gmail works on pretty much exact string matching. There's an incredible amount of knowledge that's baked into Google web search and we're trying to incorporate more and more of that into Gmail. So you will see improvement in that area in the future. It's one of the bigger back-end projects that we're working on right now.
TR: So we'll see search suggestions such as 'did you mean?'
Jackson: Similar. You have to make sense in a mail context. We refer to this as 'stemming' – certain terms that are the stem of a longer term. It's a basic property of web search. And things like synonyms and bigrams and anagrams – all that stuff we want to work well in Gmail. It doesn't work yet but it's something we are working on.
TR: So when will that be implemented?
Jackson: We try not to be too forward looking with what we announce. The reason is that oftentimes because of the fits and starts that projects have… imagine if four years ago we told you that we were working on Buzz.
Projects happen organically at Google and they are very engineering driven and sometimes they start and stop and resume later. And we also don't like to tease users – we want to announce it to users at the moment that it's ready for them to use it.