My flight left Shanghai at 1pm, got into SFO at 9am, and I headed directly to the Palace Hotel, the location of the conference as fast as I could. By 11am I was in my seat and listening with as clear a head as I could muster.
The first talk I caught was You Are Here: How Google Maps Keeps Innovating, led by Bernhard Seefeld and Elizabeth Windram. They talked about their experiences pushing GMaps along. The biggest idea I got from their talk was that they focus on and design for power users — not in the traditional sense of super-techy people, but rather the users who want or need to use the service often and therefore have both the highest level of patience and the highest expectations. I like this attitude — assuming you can expect continued growth, you can and should design for a better/more saturated day.
Another key line from the preso was: “Google Maps is Google on a map.”
Interestingly, the GMaps team seemed to go with the genius design + churn approach — throw something out there as accurate and useful as you can, and then churn out an update as soon as the masses shout loud enough for it. More about “GD+C” in a minute.
There was a useful portion about “two search boxes vs. one” — in other words, (text box one) “italian food” NEAR (text box 2) “san francisco”, compared to (single text box) “italian food san francisco”. The latter is much more intuitive and much more DMMT (don’t make me think), and it’s the kind of thing that, when it works, makes us regular users go “Wow, they’ve got some black magic over there.”
But it led me to stand up and ask the presenters a question. What I was trying to explain was — Google Maps in San Fransisco is like a mind-reading genius because there’s so much data and it’s properly attributed and Google did a lot of legwork in this area to fill in the blanks and so on. But go to, say, my hometown of Shanghai, and you’re shown the same friendly interface, but you get an epic fail in terms of results. One, the map doesn’t give good results for fuzzy input, and two, the map doesn’t connect well to the data that I already know is in Google.cn, because I find it when I do a regular Chinese-language search.
So my question was, why does that happen? Is it a character set issue? A semantic issue? An IA issue? A quality of information issue? Or just a matter of not getting around to it? I got some apprehensive reactions from Bernhard and Elizabeth — I guess China + Google is a touchy area — but in the end I felt that, as much as Google pushes the long tail on areas where it’s a newcomer, such as advertising, it’s quite content with the status quo in its core competencies/regions. In essence, it’s more valuable to Google to make the functionality twice as good in the Valley than it is to make it decent in Shanghai. Well, I have only one thing to say about that, and it’s: Baidu.com 76%, Google.cn 20%, and that’s after Google’s big push here.
Finally, since we always want to know what’s in Google’s secret sauce, here’s the content from their “innovation” slide, which I liked a lot:
- Be self-critical
- Focus on core competencies
- Be inspired by people who push the edges
- Use power to drive simplicity
- Establish a framework to evaluate new features
You can follow me on Twitter at @n_kruse and track the conversation at UXWeek via #uxweek. If you’re if you’re in San Francisco and want to talk about UX or about how Reigndesign could partner with you, or you know someone there, DM me!