"We've made a significant investment in mobile search technology" -- Dipchand Nishar, Google

Jun 30, 2006

Google has stated that they need to provide services to mobile phones and other non-PC devices to ensure the future growth of the company, and Google has started to address this market in earnest. For example, in May of 2006, KDDI, a major Japanese mobile phone carrier, announced a relationship with Google, giving Google a better hold in the cutting edge Japanese mobile market. Google's Director of Product Management and one of the key players in Google's mobile strategy, Dipchand Nishar traveled to Tokyo for the announcement and spoke with Nikkei Electronics.

Q: What are the differences between the approaches Google takes with the various parts of the mobile ecosystem and how does that differ by territory? For example, in Japan, you are working with KDDI, a carrier, and in the US, you are working with Motorola, a phone manufacturer.

The difference is in terms of the services we provide. Motorola as a manufacturer doesn't really provide a phone service to any users. But they do provide a very important thing, the device that uses the phone service. With Motorola, we provide a Google icon on the top deck, so, when you open the phone, Google is one click away. On a normal phone, it's very, very difficult to reach a website. It's not like a PC where you type in the address in your browser and you're there. On a phone, you have to go through several clicks to get there, which is a burden on the user. So our partnerships with device manufacturers like Motorola, Sony Ericsson, BenQ, Siemens, are designed to make Google and Google services available for users with just one click.

With the carriers, mobile operators have their own content and we provide them our search technology as well as our advertising technology. We have relationships with KDDI in Japan, with BeeLine in Russia, Vodafone around the world, Telefonica in Spain, D-Mobile in Germany, and many, many other carriers around the world. For KDDI, we are using our core search technology, not just to provide full Internet search, but also for KDDI specific content.

Another part of our technology also deals with formatting this information so it fits a small device. With search information from the full web, as you know, there are lots of images, sidebar navigations and so on. They don't come out very well on a small form factor device like a phone. We're providing KDDI with what we call transcoding services so pages can be rendered to show well on the phones.

We also provide our advertising platform. As you know, Google is used by hundreds of thousands of advertisers around the world. What we do is, based on the search query terms, the right sorts of ads show up. When users search for, say a singer, we are able to provide relevant targeted text ads specific to that search query.

Q: KDDI said that KDDI and Google just started to talk last November. So I think you developed this system in a very, very short time.

I was here last November and we started the conversations along with our CEO Dr. Schmidt at that time. And we reached an agreement in a very short period of time. But the core search technology has been available for a long time. We have been providing mobile search for users. In Japan, i-mode based search was available since 2001, so it has been available for a while. Obviously, we have refined it over the years, and over the last 12 months, we've made a significant investment in the search technology.

"We are excited about Ajax on the phone"

Q: Can you provide any more details on your mobile phone transcoding technology?

Many phone browsers don't support Java script and they also don't have good CSS support. So what we have to do is make sure that those things render okay. There are some obvious things we cannot do today like showing Flash on a handset that doesn't support Flash. But if the user queries for something, and if a web page contains that information, we do our best to actually show the information that is viewable on the limited form factor of the device. That's really the core objective of our technology.

Surprisingly, there is a lot of heterogeneity in device. The same manufacturer can make the same model but, due to some changes to the operating system, the same application doesn't run the same way on two of the same devices. So device platforms are very heterogeneous, which is why you require a lot of innovative technologies to do the transcoding right. We believe we have really put in a lot of effort into and have a lot of intellectual property around how we do the transcoding.

Q: What are the advantages or differences between other transcoding technologies and yours?

There aren't that many other transcoding technologies out there because a) it's very difficult to do it and b) it requires a lot of resources in order to make sure that it runs not on one phone or five phones but on the hundreds of phones available out there. So the first differentiation we provide is really scale. Our transcoding technology works on hundreds of devices in hundreds of countries around the world.

The second is we have created a very comprehensive capabilities matrix of different devices. A simple example is that different devices have different cache sizes, which means if you are downloading a full Internet page on your phone, your phone might only be able to see 10 lines of text. If we don't understand that and try to send 20 lines to your phone, you will lose the 10 lines and never see them.A third example is say you are looking for a very specific term, such as a special kind of orchid. On the website that's coming to you, it's on page 5. On a PC, it's very easy. You just scroll down and you see it. On a phone, it becomes a little bit more difficult. What we do is to take you directly to the page where or the portion of the page where the information is, so you see that on your screen.

Another example is dynamic images. If you have dynamic images, then they really ruin the experience on the small form factor device, so we handle them differently than you would otherwise.

Our algorithms can also help you to find the most relevant result quickly. We are very particular to make sure we provide search as fast as possible, and on the mobile, we go one step further. Even if the network is slow, and the search was fast, if you had to click on something and wait for it to load, then ultimately it took you a lot longer to find the information and we don't want that to happen.

So we focus a lot on making sure that the first result you get is really the right result, and you might not even have to click on it to get the information. For example, if you want to find a certain ramen joint in Shibuya, you can just say "name of Shibuya ramen joint" on Google search. You will not get a link, you will actually get the answer with the ramen joint in Shibuya and the information you want. So you don't have to click on anything. It's almost like you asked the question and you got the answer. I can't change slow network speeds but I can definitely change your searching experience.

Q: There are a couple of browsers, called full browsers, that enable you to access a PC oriented website. Will this lower the value of your mobile services?

There are two parts to this. One is that for a phone that supports it, a full browser actually is a good thing because then you don't have to make a compromise as a user. And for Google that's a very good thing because, at the end of the day, all we want to make sure is that the user gets the information they want.

The flipside of that depends on which part of the world you're in and what kind of data plan you have. A full browser downloads a lot more than a transcoded portion does because it's rich content. And if you have a plan where you're paying by the packet and you just want one tiny piece of information, that's a lot to pay. Now, KDDI, for example, has a flat rate plan and this might not be as important. But in Europe, a lot of people pay per kilobyte of download. Then it becomes a little bit more expensive for them. Even if it's a flat rate plan, it still takes time to load the images.

Q: In terms of rich-media contents, what do you see the potential for Ajax content on mobile?

That's a very interesting question. The way we provide Ajax-based services is heavily dependent on what the browser capabilities are. Until very recently, mobile browsers did not support Ajax, but we have started seeing some new Ajax browsers coming out. I think some of the high-end phones will start supporting Ajax in the next six months and then, anywhere between the next six to twenty-four months, a lot of phones will start having it. We are very excited by that opportunity because now, all of a sudden, you can provide many more value-added services for the users and the experience becomes a lot richer.

A very simple example is let's say you had a box where you fill something in on your phone and you had a link. If you don't have Java script-capable browsers, what you put in the box when you click on the link is lost. If you have Java script, then you can do that. If you have Ajax, you can provide on the phone some of the richness of interaction that we provide today with Google Maps on the PC. So we're always very excited when the phones become more capable from a technology standpoint. If there was a graphics chip on every phone, then you could actually have a Google Earth experience on your phone. That would be very interesting and very useful. Combining Google Earth with GPS on a phone would be fascinating.

"Japan is a leading indicator of what will happen in other parts of the world as higher speed networks become available"

Q: What unique traits do you see in the Japanese market and do you have any unique services to address them?

For instance, the Japanese user doesn't use text or SMS at all whereas text and SMS is very common everywhere around the world. But the usage of e-mail here is very, very high. One thing we provide in the Japanese market, is searching via e-mail. People want to have a lightweight way of searching for information and we actually provide that using email. You don't even need to open a browser. Open an email client, enter a search query, send an email to g@google.com, and we will email the responses to you with three results. It is also hyperlinked to our services to show you the results. So this is for people who might not be familiar with a browser so they can actually see the results quickly. If they want to see more, they only have to click and see the browser content.

In Japan, 25% of Internet usage happens on the phone. A lot of Japanese users tend to use their devices while commuting because using public transportation is a key part of their experience. These sort of cultural or social factors are not necessarily replicated everywhere else around the world.

However, factors such as the high penetration of the 3G network in this country and the availability of the 3G network means that people tend to use a lot of media-rich applications. So a lot of users here see TV on their phones. That's not as prevalent in many other parts around the world.

So to make a long story short, what we learn in Japan is 3G networks tend to promote a high usage of media-rich applications. And I see this as a leading indicator of what will happen in other parts of the world as higher speed networks become available, say, for example in parts of Europe as well as in the US.

In terms of cultural notions, the high usage of public transportation in Japan promotes certain types of services that will become less prevalent in the US, since people tend to use their phones when they are in their cars more than half the time. But in other markets such as India or China, where again the norm is more pubic transportation, as the markets grow, I expect them to follow a similar path to the Japanese market.

Q: What kind of activities does the Japanese R&D center undertake?

The Japanese R&D center is very much like all of our R&D centers around the world. We have one in Beijing and Bangalore and Sidney, Australia and London and Zurich and four different places in the US and so on. We treat all of our R&D centers in the same way, which is they build products that provide the maximum user value and the products can be global in nature. People in the R&D office in Tokyo can work on products that will be used everywhere in the world, and we also encourage them to build products for their local markets.

An example is the send-to-phone feature where you can do a local search on your PC and then send it via email to your phone. If you look for directions, you don't have to write it down on a card, you don't have to remember the map, you can send it to your phone by email. That was actually done as a joint development between our engineers in the Tokyo R&D center and engineers in California.