Mary Meeker has been the leading financial analyst of the Internet for more than the last 15 years. Recently she left her long-time employer, Morgan Stanley, and joined the venture capital firm of Kleiner Perkins in the Silicon Valley. At a Google event in San Francisco last month she gave a brilliant review of the current state of mobile and explained how it is poised for explosive growth:
The presentation focuses on mobile devices as extensions of activities we've up till now used computers for-commerce, social media updates, games. The "location-specific" applications are still quite murky. How does a store offer discounts to people who are in close physical proximity to the store? How does a store engage with a customer in the store, standing in front of a product? Some of this requires retailers and other merchants to think through innovations.
On the other hand, we need to be learning new behaviors with our mobile devices as well. Foursquare "check ins" were foreign behavior three years ago, and now they're common place. Among some of my friends it's even jumped the shark-at least until the time when Foursquare offers a richer array of information and offers.
Location-based customer engagement has many facets. Here are a few I'm considering as I work with clients to extend their customer services into mobile:
Location seems like a concept that's easy to understand. Location is where you're at, right? If you're relying on GPS features to solve queries like "where is the nearest coffee house?," that is true enough. But location just as often means where you will be or may be in the future. Micro-local advertising must therefore be addressable based on queries, not just "here I am" check-ins, as with Foresquare.
I was going cross-country skiing outside Minneapolis a few weeks ago in a large park preserve. I knew afterward I'd like to get lunch somewhere-but I wanted it to be a great place to eat, not just as a Subway or other chain. Google Maps on the iPhone makes this easy enough, but it turned out my query was too small and out of the way. I wasn't interested in restaurants near the park I was skiing at, I wanted places along the way home.
There's not much free information about places (only a few have Websites) and zero advertising or coupon specials. I had to guess if a place was what I wanted based on the name. Was McGilligan's Irish Pub a good place to eat? I rolled the dice and stopped in on the way home. I got lucky-the creamy chicken and wild rice soup was terrific.
That's the mobile experience today. Success in the future is not about advertisers getting their coupons into the system. That would be a nice benefit. But the benefit to me, the consumer, is that I find information right away about what's at a place I'm going to.
Context affects the information that gets returned based on what the mobile carrier can determine about what I'm doing as I make the query. If I'm in a car travelling on the freeway and I use the voice command to ask "where is the nearest gas station?" the carrier should be able to tell me based on the freeway exit system, the traffic congestion in the area, and the direction I'm moving. That's a simple example.
But if I'm out on a date with my wife on Saturday night and we get out of a movie and want to get some food the system should understand what kinds of places we've gone to before, which ones we like a lot, and what's similar in the area. Also, are we walking or driving in the car? Context is clearly a huge problem that will require many refinements over time.
The big Kahuna is what will my smartphone feed to me when I'm shopping at the grocery store? Or when I'm standing at the entrance of Best Buy? Will the mobile phone turn into a hideous profit-eating monster that allows customers to browse in the aisles of a store looking at real-life products, and then from the floor order that product from Amazon for $50 less?
What kind of coupon strategies will provide real value to merchants because people actually get them in a timely context and decide to use them? The ROI models on these need to go beyond the gross proofs that Meeker offers of improved foot traffic. Those metrics are enough to encourage retailers to experiment, but we need to get some success stories about the actual coupon / loyalty / location-based discount process.
Social is obviously important-keeping my friends and family up to date on what I'm doing, checking on them while I stand in line at the drug store. But I have more than 200 friends on Foresquare and I almost never check in to a place where anyone else is near. (The sole exception are social media professional events, in which case more than 75% of my contacts are in the same room-so what?) I'm sure this function of Foresquare would be vastly different if I was back in my 20s and single. But as a married guy with two daughters the only juice I get from Foresquare is winning mayorships. I've never followed a recommendation, I've never bought a special and I've never gotten a drink as the mayor of a restaurant.
The accidental discovery of friends nearby is only one social application. Recommendations coupled with thoughtful specials and advertising can be very powerful, but they require a density that Foresquare and Gowalla just don't have now. Foresquare needs to buy Yelp or Yahoo needs to buy Foresquare and get some serious merger of a critical mass of recommendations with location based check-ins.
Foundation - Finally, as these mobile applications move out, designers will be challenged by what experience the average user brings to location-based computing. What applications did the consumer start using first? Blackberry? iPhone? Or something other than a smartphone at all? This will do a lot to determine their future use and to predict the "switching cost" a consumer will go through to get familiar with and have enough relevant social data to draw from.
Developing applications for the smaller, better educated crowd that owns iPads, Android and iPhone smartphones and the like will be very appealing. There's a lot more you can do with a crowd like that, even if they comprise a small minority of all mobile users. I wonder if a kind of "smartphone" gap will begin to appear shortly in mobile computing, with a smaller group of smarty-pants customers who will use a lot of sophisticated mobile services and a much larger group that will have static GPS and little else?