INTERVIEW |Google Australia's head of RD reveals three things you probably didn't know about Google downunder, including the fact that the Apps suite, which is so critical to Google's mission of beating Microsoft Office into submission, is partly developed here.
|Google del Mar: R&D chief Alan Noble at the beach
Three things you probably didn't know about Google:
- Google is developing Google Apps in Australia including several that are not yet announced
- Google Maps was invented in Australia, and key developers are still based here
- Google is creating new maps in parts of the world where quality map data doesn't exist
These tasty morsels -- and more -- are in a quick interview I did with Google Australia's head of Research and Development, Alan Noble at last week's Google Developer Day.
APC: Congratulations on the amazing turnout to the Google Developer Day today.
Alan Noble: Thanks, we're very excited that there are so many developers here.
APC: You had to change the venue to a bigger one at the last minute, so I'm guessing the turnout was a bit of a surprise?
Alan Noble: Yes, actually we were quite flattered almost by the amazing turnout. The original venue was going to be a hotel in the CBD, and we switched here to the Technology Park in Redfern to accommodate more people. It's really indicative just how popular some of our developer products are becoming.
APC: Does Google's phenomenal success ever surprise you guys internally?
Alan Noble: I'm pleasantly surprised. I don't think we ever take it for granted, but I've certainly been happily surprised by this turnout today.
APC: So, as the head of research and development at Google Australia, can you tell me about the kind of work you're doing here, and why you're not doing it at your Mountain View HQ?
Alan Noble: I'll answer the second question first about why we're doing things in Australia. We're doing things in Australia as part of a philosophy that we should do research and development around the world.
Really, the same reason that we're bringing these development conferences to 10 different cities around the world where we've decided, and I think this is the way to go obviously, that we should do our R&D around the world also.
In summary, we don't expect that every bright engineer is going to work on location in Mountain View, California, so Google Sydney is part of our global R&D.
We're here because we think there's great Australian engineering talent, and we've got big plans to grow the R&D centre right here in Sydney, and potentially even expand to other locations in the region in coming years. So that's why we're here in Australia.
As to what we're doing, well obviously we started out as Google Maps, that's our heritage I guess you could say. But as we grow and we have more and more engineers, we'll be taking on more projects.
We've now started to work on some of the online collaborative applications that make up part of the Google Apps Suite. Products that haven't been announced yet, but we're very excited about those. So we call that Content, Communication and Collaboration or CCC.
APC: Really, that's very cool.
Alan Noble: It's a bit of a mouthful, but really it sums up essentially the next generation of web applications, which is really all about content, communication and collaboration. And we think collaboration in particular is central to everything we do.
APC: So you guys in Australia are directly involved in the head-on challenge to Microsoft Office?
Alan Noble: You could say that. We don't tend to think of it in those terms necessarily, but certainly I expect Google Apps will be very, very popular. I think in Australia in particular, we've got a very large number of small to mid-sized businesses where frankly it's going to be very appealing.
The third area we're starting to work in is more on the infrastructure side, and particularly networking infrastructure. The reason we've selected that is a third focus area for our engineering is to tap into the engineering talent around here. There are a lot of great engineers here: good networking engineers particularly.
As cool and as exciting as web application development is, it's not for everyone, and we want to be able to offer great job opportunities for all sorts of developers, whatever their passion is.
APC: So what's the size of the development team in Australia?
Alan Noble: Well I can't really get into specifics, but I'll say we have well over 100 people in Australia right now, and about 1/3 of those are software developers.
We've also got a lot of other engineers with the network operations developers and systems operations and engineers. And that's growing very quickly. And ask me again this time next year; we'll be hopefully twice what it is today.
APC: I was wondering whether one of Google's imperatives in hosting these events is because it is quite hard to find developers up to Google's calibre. Is that the case?
Alan Noble: This event is about helping us get word out to developers, it's not about recruiting them obviously, but in general - we do face recruiting challenges, not just here in Australia, but everywhere. We do set a very high bar in terms of getting into Google. But I'm convinced there are lots of bright engineers here in Australia.
I think longer term it's going to be a challenge. We need to make sure that we're producing lots of IT and computer science graduates.
I think short term, I don't see particularly insurmountable challenges. I think longer term, we need to make sure there's an ample supply of IT graduates coming out of Australian universities. But that's true of the entire IT industry in Australia. Which is why we're starting to get involved in some of the organisations like the Australian Information Industry Association.
APC: I wanted to ask you about how much of Google's success is down to having pockets deep enough to buy good content, not just build the technology to house it. For example, mapping: it must have been fantastically expensive to buy the intellectual property for maps and satellite photography worldwide. I imagine in some parts of the world you can't even buy it?
Alan Noble: You're spot on. There are definitely parts of the world where it's hard. And even Australia is relatively less well served than say the US or Europe. In fact I often like looking at the maps from Japan, because the quality of the other map tiles is just superb, because the rest of the world has changed.
There are definitely varying degrees of data quality as we go around the world, and it's not just maps, it's also business data for typical local business data, also known as Yellow Pages. So we see huge variation.
We are tackling these issues a couple of different ways. One thing Google likes to do is to use our technology and our algorithms and try to improve the quality of the data, so it's about teams working on for example how we can make local business search return more relevant, high quality results?
The raw data may not be that great, but with the appropriate algorithms, we can massage it and make it better.
And also we're looking for ways to work with our members and providers to improve the data quality. So it's a mix.
The third way is through users, and getting users to actually publish data. Wikis are a great example of user produced content.
So, we're looking at all three ways to address lack of data, but it is definitely something we have to deal with, especially in parts of the developing world obviously.
APC: I recently went to Beijing and I discovered that there were no Google Maps for Beijing, only satellite photography. Then I tried very hard to find them on other sites by searching for hours and I actually found it incredibly difficult to get any maps of Beijing at all. What were there were kind of blurry, scanned in tourist maps in giant PDF files.
Alan Noble: Some of those might be regulatory. Most constraints on us are to not reveal too much level of detail. I think there actually are Google maps in China. I don't know much about the quality of the maps.
I do know in India for example where there were no maps, we created our own.
Alan Noble: Basically a team of Indian engineers essentially created our own maps. So that's another option too.
APC: That would have been a huge undertaking.
Alan Noble: They just did all the major cities. Not necessarily every village in the subcontinent... [laughs]
APC: Sure. So you mentioned you're working in Australia on some Google apps. Obviously there's quite a sudden and big focus on making apps work offline as well so people don't need a live net connection every time they just want to use a word processor.
Alan Noble: Absolutely, and as we announced today, yes.
APC: Mozilla CEO Mitchell Baker was out in Australia recently talking about that from a Firefox perspective too.
What I wanted to ask is: there seem to be some key problems with online apps, for example synching with PDAs and, to an extent, providing adequate desktop alerts for calendar events and so on. I know Google has got a notifier, but it's still kind of basic compared to what Office can do.
I wondered whether Google is looking at solutions to build desktop apps that link to the online services, or is the philosophy simply to make everything work as well as possible within web browsers?
Alan Noble: It's really the latter. We feel quite strongly that the web browser and the web in general is the ideal platform. However, we want to make web applications run offline too using a locally cached database. So that's kind of our philosophy.
I think with Google Gears, some of these issues about synchronisation and notification will be addressed. They're going to fall into place quite naturally.
We're well aware of the fact that there are areas where - because we're relatively new in some of these products, the Google Apps Suite, we don't yet have all the full features of some other products out there.
The big difference of course is that we're putting this all into the open source. There'll be hundreds if not thousands of developers working on this.
APC: Gmail is one of those apps that is deeply impressive, but mostly because of the large storage capacity at Google's data centre. Do you think we'll ever see an offline version of Gmail?
Alan Noble: I think if you take a look at all the applications that we're using, email, calendar, Writely, which is our word processor and Google spreadsheet, I think it's entirely reasonable to expect that all of those will run offline as well as online. I couldn't give you a timeline, but that would be the normal extrapolation.
Obviously we're starting with Google Reader. We've announced Google Reader, but I would expect there'll be other applications for sure.
APC: So Gears in general is tackling the issue of finding web apps offline. Does it compete with Firefox's offline apps framework?
Alan Noble: It actually complements it really well. With Gears, we're trying to make it really browser agnostic, so we're trying very hard to make sure that it works with Internet Explorer, Firefox, Opera and Safari on whichever platforms those browsers run on. We'd actually like Gears to be a standard, and we're working with companies like Adobe and the Mozilla Foundation, so we can get some momentum around turning it into essentially the standard for offline web applications.
APC: Interesting that you should mention Internet Explorer. I noticed in the Google Gears press release that you had endorsement quotes from Mozilla, Opera and Adobe, but nothing from Microsoft. Microsoft hasn't really shown its hand yet in relation to offline web apps given their natural conflict with their desktop apps. Where will Google Gears will sit in terms of compatibility with IE and do you expect Microsoft will play nice with it?
Alan Noble: Well it certainly is compatible with IE. I mean IE is still used by the majority of web browsers today. I don't know the exact percent, but it's probably well over two thirds. And that's an important platform for us.
So we'll do everything within our powers to make sure it's fully compatible. And I'd love to see Microsoft sign up to the Gear standard. It would be a fantastic outcome.
APC: Coming back to Gmail, it's undeniably impressive in terms of how it has expanded what we expect from web mail, but there seems to be some curiously missing basic features, like the ability to search for messages by size, so you can show all messages above 10 megs and delete them or whatever.
I know that the Gmail team recently did a questionnaire on what people wanted from Gmail which I guess indicates that Google is actively developing it, but it's a bit hard to know what to expect from Google. At least with companies like Microsoft that make desktop apps, there is a somewhat certain roadmap where you can expect certain releases every X years, and so on. Does Google have a lifecycle plan that for the online apps?
Alan Noble: We do. I would say our lifecycle is much more aggressive. We're constantly trying to prioritise new features. To date, we've taken a lot of our feature prioritisation from end users. I think what you'll start to see is perhaps more enterprise requirements factoring into that prioritisation.
I think the dichotomy is really not as great as some people would think. Ultimately people in an enterprise are also the end users. And what works for an end user working in the office is probably also something they need to be working on at home or in a hotel or on the road from a hot spot.
So there certainly is a road map and it's certainly reasonable to expect in the future - there's also a tremendous amount of innovation happening in general. There may well be new email developments that go in ways that perhaps haven't necessarily been anticipated by the industry...
APC: You're teasing us here.
Alan Noble: Well, what I mean is that one thing Google does is - and this is definitely worth pointing out -- we have this fantastic thing called 20% time.
At any point in time, 20% of our brain trust is looking at new ideas. As we develop ... there are always new ideas. We're a very bottom-up company.
Gmail itself started out as a 20% type program. An engineer thought "gee, wouldn't it be nice if I could have a search interface with email?" So the concept of Gmail started out that way.
One of the great things about the way Google operates is that we encourage this kind of bottom-up creativity. Even as someone who's in a leadership position in the company, I don't always have visibility of what's happening, and that's exciting.
APC: Something that occurred to me recently is that one area where Google hasn't really played at all is in the area of web hosting. I mean obviously it's probably the world's biggest webhost for its own stuff, but you haven't done much in terms of enabling people to publish their own sites on Google's infrastructure.
Alan Noble: Well, there's Google Pages for publishing simple pages...
APC: It's rather basic though. I mean, I guess Google doesn't want to end up the entire internet, but at the same time, Google has provided innovative, free versions of just about everything else - why not some form of hosting that allows for simple CMSes and database-driven sites?
Alan Noble: So there's a couple of things we're doing in that area. We do offer Google Apps for your domain, where essentially we are hosting your Domain. It's not your website, but it's just your applications and it's your documents, and that's quite popular obviously.
The second thing we have is Google Base which is essentially a Google Database in Cloud for structured data. So we're asking all of our developers to basically consider using Google Base for online storage. If you combine that with Google Gears, you have the best of all possible worlds. You can have your large data store, leverage of Google's vast infrastructure, and then you can have your local Google Gears database cache.
In terms of just straight web hosting, I really can't speak to that. My view is I don't think that's part of our strategy.
APC: So with Google Base, can you deal with it in the same way you can with MySQL or something like that?
Alan Noble: Yes, there's an API. It uses this thing called GData API, so you can populate data to retrieve data search data. There are some great examples.
In the US we've put the entire country's real estate data in Google Base; obviously we'd love to do that in Australia too [grins knowingly]. Then you combine that with Google Maps, so you can see maps of properties up for sale or rent and all that stuff is stored in the cloud.
APC: The last question I had for you was around obviously getting involved with Google. There's been a bit of press lately about the fact that it's quite difficult to approach Google sometimes and you need to have a PhD to work there.
Alan Noble: Who's saying that?
APC: There was some blogger who had some great idea and he couldn't find a way to reach a real person at Google, one of those kinds of stories.
The question is A) do you have to have a PhD to work at Google? And B) other than Developer Day, how do people get in touch with Google to ask questions?
Alan Noble: Well you don't have to have a PhD, although about one third of our engineers do have PhDs, and one third have bachelor degrees.
Just getting in touch with us, this [Developer Day] is a forum where we try to encourage direct contact. There are at least 30 engineers out there and probably a similar number of non-engineers in support roles and we're here to talk to new developers.
So this is a great way to get in touch with people. Beyond that, I think when people blog about ideas, they've had. It may be the coolest idea in the world in their mind, but they should put it in context and realise that there are other people who have very similar ideas.
Sometimes the lack of interest is not a kind of snub, it's just that we've seen a lot of ideas and we only have so many hours in the day.
We are certainly working very hard to be seen as being more open and more receptive. For my part since joining Google which was just three and a half months ago, we're doing a lot more of these size of events, a lot more interviews, a lot more speaking events.
Yesterday I spoke at the AIIA's annual conference. At all of these opportunities, you've got an opportunity to reach out and find someone from Google and share your ideas.
I mean you can also try shooting us an email too, but what we really look for I guess is a bit of diligence on the part of the person contributing the idea. It's easy to say I've got a great idea -- we all have great ideas. But we like to see that you've said, okay I've got an idea, and I've taken it to the next step. I've built a mashup. I've used Google Base for my storage and created a gadget. It shows the person has what it takes to take the idea to the next level.
The point is it doesn't have to take lot of work, but if you've done some prototyping, that's when you'll get our attention.
APC: Can you just tell me a little bit about the gamut of your role?
Alan Noble: Well my role is roughly - I'd kind of divide it up as three responsibilities. I've got an internal kind of R&D management responsibility which is probably about one third of my time roughly. I've also got a role which is kind of relations, so external relations, media. And that also is internal relations too, because Google is a large distributed company. I spend a good chunk of my time back in the US making sure that Mountain View, California has good visibility in what we're doing.
The other third of my time, recruiting. A lot of recruiting. I typically interview two or three people every day of the week. Some days I'll interview five; today I'll interview no-one because I'm here. It's a big part of what I do, and it's a big part of what all Google Engineers do. When you're growing as fast as we are -- 100% growth a year in engineering - you have to do it.
I would probably also summarise, a kind of bigger part of our mission here is to be seen as really more a part of the Australian community. We're really trying to remedy the perception that we are not approachable. We want to be actively involved and supportive of the developer community.
We're doing a lot of work to support what we think are great projects, like the open-source development community, I mentioned in my key note. We sponsor at least a half dozen user groups - Python, Linux and so on.
And hopefully we're being a force for good in the Australian economy. It sounds a bit like a cliché, but we've got a lot of products, and I think that it will raise the game. I think it will raise the bar in ecommerce in Australia.
For example small businesses, this is a terrible statistic. I mean only eight per cent of small businesses in Australia have a web site, so how do they reach out to their customers, especially overseas ones, global markets, to generate precious export revenue?
I think Google can be part of the solution here. We can provide the tools to help businesses create innovative products. Focus on the innovative stuff. Focus on the stuff that gets users interested in products. Don't worry about the databases or the search and all that. That stuff, we can provide.
Then when it's time to make money, monetise it with Adsense for example, or sell stuff with Google Checkout, which is coming to Australia soon.
APC: Thanks for your time Alan, much appreciated.