Tokyo Forum; November 21, 2002

The Emergence of Context: Consumer Behavior on the Mobile Internet

Philip Sidel (Assistant Professor, International University of Japan)

This is a summary of Prof. Sidel's presentation at the Tokyo Forum on November 21, 2002.

Mobile Internet Hype

I think most all of you would agree that the major advantage of the wireless/mobile platform is its ability to free us from the constraints of location and time. This is something new and creates a very powerful platform. It is not very difficult to say that the Internet hype has happened and we are through it. Now, we are at the point of mobile hype. There are many people looking at the advantages that freedom from location and time bring, and some have suggested that a new form of marketing, or "contextual marketing" takes advantage of this "anytime, anywhere access". We are starting to see the hype build again, focusing now upon the business opportunities for location-based marketing, and "just-in-time, just-in-place" information. The excitement is definitely brewing. We also have seen some very interesting attempts at business models. For example, from a Harvard Business Review (Nov/Dec. 2000) article, we find large companies such as Unilever trying to provide mobile recipe content.

I won't make any comments about the success or failure of this type of application, but I would submit to all of you that as of yet the killer application, the application to take advantage of all of this technology that we've heard about today has still yet to be provided. So for those of us in the business world, who are looking to create viable, long-term, sustainable businesses and business models, I don't believe we have yet seen a framework emerge.

Testing Contextual Marketing Hypothesis

So what is "contextual marketing"? Scholars tend to focus on three specific areas: we got "content," "location" and "time." Some authors also mention customer demographics or psychographics. So if these authors are correct, then we should see a clear picture of how different people are using mobile Internet at specific locations, for example, from home or from work and the specific times of day, in the morning, in the afternoon and in the evening.

So we created a survey for mobile Internet users here in Japan to test this hypothesis, and determine if there really is a story here.　Our survey was designed to understand how time, location and content interact with one another to try to form this picture for all of us. We sent a survey out in conjunction with a company, iShare, Inc., to 250,000 current mobile Internet users in Japan, and received responses from approximately 5.6 % of this population, a little more than 14,000 people responded to our survey. Although the survey sample was skewed a little bit more towards male than female, these respondents represented 47 prefectures and wards across Japan, across all age demographics from people 10 and 12 years old all the way up to 70, and male and female from all professional categories as well.

Survey Results

When we look at these survey results, we did find some interesting stories. So, for example, if you look at mobile Internet usage from the home you do see that people either in the morning or the evening are accessing mobile Internet most from their living room or dining room. Second comes the bed room, third the bathroom, and finally the kitchen.

There were some interesting stories that emerged from these survey responses. Through statistical significance testing, actually in the morning we would have expected more people to be accessing mobile Internet from the bedroom than actually did. And conversely in the evening hours, more people that we would have expected were accessing the Mobile Internet from their bedrooms. So we can see here that some interesting stories are emerging here.

We also went a little bit deeper. We surveyed across home, work, school, leisure time and commuting time mobile Internet usage, and probed a little bit deeper than previous surveys have. We asked what people were doing in specific locations at specific times of day. For example, here we see Mobile Internet access from "the work bathroom or toilet". We've all heard the stories about "trading stocks" from the bathroom, but our research showed that this was not the case. That isn't to say that people aren't going to the bathroom and trading stocks on the Internet. But what they are doing most often is mail and chat. And if we look at every other location, the story looks pretty much the same.

Again inside in specific location and day part interaction we find specific, statistically significant differences, where actual results are different than expected results. But again, mail and chat followed by news and information, banking, trading and entertainment were the things that were leading.

So in adding all these up, we looked at a number of different interactions between location and content and a number of different interactions between day part and content. And we did find some interesting stories. Mail and chat is dominant. But when we try to step back and aggregate these important findings, we find very few cases where actual patterns emerge. We find little bits of interesting things here and there, but, to be honest with you, in all of our data modeling and in all of our data analysis we could not find a clear pattern.

So the question now is: Why are we seeing such differences and similarities? Why are some stories interesting and why are some not? And further yet, for any statisticians in the room, why do statistically significant results fail to show a clear picture? So we believe in looking at these results that the people who have been propagating this hype about location-based marketing, location-based advertising, "Just-in-time, Just-in-place," etc. only have part of the story. The story is too simple right now, and the businesses that try to implement solutions just based on where somebody is working or where somebody is sitting at specific time of day will potentially fail.

Future Research Direction

Rather than content, location and time of day being the key descriptors, we are focusing now instead on that the unique psychology of each individual customer. What they bring to that location at that time of day. Specifically, their "location embrace," "time usage" and "information privacy." So, we plotted these in three dimensional space. To define context, we define it in these three terms.

The first axis is "location embrace." Let me explain a little bit. If you walk into a store and you have your "keitai" with you, your mobile phone, then you feel like knowing about where the latest coupons are and where the latest sale is. That is a very strong location embrace. We see those people highly embracing their location. And you can use your keitai, your mobile phone, to do that. On the other hand, there are people who can walk into that same store or you yourself could walk into that same store the following day and not be interested in that sale or those coupons, but instead you would be interested more in mailing your friends or looking up some entertainment information. And in that respect, you are not embracing that location. The environment does not matter so much to you. So, that is one axis.

The second axis is "time usage." Do you want to be productive or do you want be consumptive? We found in our survey that you could have two people in exact same age demographic, sitting on theoretically the exact same bus at exact same time of day and one has an interest in getting some thing done, while another has an interest in just escaping, passing time.

The final axis that we have is "privacy" or "self-disclosure." Within that context, within that time of day, at that location, how willing and interested that person is to share something about themselves with that location or with the content provider that they are accessing.

So we plotted these in three dimensional space and created eight potential interactions here. So our efforts now are looking at this model and trying to expand upon the actual behavioral characteristics of customers within this space.

So, we are now looking for partners who have access to some of the background usage data so that we can fully populate this model. And at the same time, we are also looking to create some panels across large user segments in order to get this picture straight. Because what will emerge from this effort is diametrically opposed to the prevailing Internet "Opt-In" mentality: Once somebody says "yes" to marketing, it is ok to give them our advertisements and promotions. On the mobile platform, customers might not be in the mood at the specific time of day or in the specific location that you think they are, and you have to take that into consideration.

So we are looking forward to embarking on this next phase of our research so that we can share those findings as well as to how businesses can truly adopt and embrace mobile Internet consumers today and not worrying so much about the hype, all the projections, graphs and charts of what's to come but what people are doing now and how we capitalize on that. Thank you very much.