Location based services and GPS

Following my initial write-up a few days ago about consumer adoption of location based services, I've been doing some further reading and thinking on the topic.

For location-based services to be truly attractive to consumers, there needs to be technology that can accurately pinpoint your location on the map. Accuracy needs are different for different applications, but in an urban environment at least a 2 city block accuracy range is necessary.

There's an article on Venturebeat today titled Three Deadly Sins of GPS that covers this issue of location tracking accuracy. It's written by the CEO of Polaris Wireless, a wireless location company. The last three paragraphs of the article are a plug for Polaris's technology, but the rest of the article is quite useful as a primer on the shortcomings of GPS on the positioning accuracy front.

One last observation on this topic: When Google Maps launched their location tracking feature (beta) a few months ago, I enthusiastically turned on GPS tracking on my handset and waited for results. More than six months later, it's still off by several blocks. The point is, Google Maps is using hybrid technology, not just GPS, to locate my handset. And it's still not accurate enough for a pedestrian.

Dear Google, I am not Italian

Yes, I have shared an office with Italians for years at a time. I am addicted to coffee. I have even tried to learn Italian. However, I am not Italian.

Mystified? So was I, since Picasaweb decided a couple of days ago to start speaking to me in Italian. All other Google products seem happy to continue treating me as my boring old English-speaking self, but not so Picasa. It had invested me with a mysterious Mediterranean alter ego, with which it strenuously tried to communicate thusly: "Fotografiile mele". I decided to leave it alone in the spirit of "Picasa, heal thyself". The dam broke today. A friend, after months of procrastination, decided this morning to share pictures he'd taken over the last year or so. To view said pictures, I had to converse with Picasa. Which said: "Cine vede porecla dvs.?"

At this point, two things became clear:
(1) I was in for a small spell of diagnosis and (hopefully) fixage.
(2) This language can't be Italian. Exhibit A:"Ce reprezintă Căutarea publică?" Count the diphthongs.

Off I went, navigating by approximate translation of menu text, and hoping I didn't end up deleting all my pictures (and believing that dear Google would display large blinky "Are you sure?" dialogs before letting me do anything truly destructive). A short, but apprehensive silence followed. At the end of which I discovered that Picasa had decided to speak to me not in Italian, but in *Romanian*. I can barely pronounce Ceauşescu, and my idea of famous Romanians is Vlad III a.k.a Dracula and Nadia Comaneci. Oh well, I'm now back to speaking English with Picasa, which said brightly "Your settings have been updated!"

In the spirit of the day, here's a little more about Romania. Oh, and I don't think Google is mining my data to update my language settings.

Location based services and consumer adoption

I was recently part of a conversation about location-based services, and the cited "cool company" example was Loopt.

Loopt enables you to connect with your friends on a mobile phone, know where they are on a map (assuming they're letting you track them), send them photos or messages (and your location is automatically included), and look at your friends' recommendations for places to check out around your current location. This is a great start, but currently, the onus is on individual users to make the service useful. If you have a lot of friends on Loopt, and they are also active users, then there is value to be derived from it. If not, users who sign up for this service will be prime candidates for social media fatigue. Loopt as it stands today is a platform, but not quite a service. They are well-positioned though, to use the data generated by their users to offer a variety of interesting services. Before we examine that angle, let's take a quick look at another startup that is offering an interesting service.

Sense Networks uses massive amounts of real-time location data to provide aggregate consumer traffic trends. This can be really valuable for users if they're looking for something to do on a weekend ("hey, why's everybody in Golden Gate Park today?") or trying to discover the trendiest spot to dance the night away. It also offers a great way to explore a new city. One of my friends, on arriving in a new city, would go to city center and look for crowds of people to follow. By switching crowds every so often and spending the day doing this, he would create for himself a great walking tour of the city. I'm sure he'll appreciate the data Sense Networks' service can now provide him.

With enough traffic, Loopt could create more sophisticated variations on this approach (routes through the city by interest group, for example). It also has the opportunity to provide other relevant services, like:

• Show nearest street corner with highest likelihood of finding a cab
  
• Show "best bet" public transit stop, given a destination
• A "remember your umbrella" pop-up message if it's raining outside and you're leaving home. Or a "10 minutes to next train" message as you're running out the door.
  
• Automatic "shop together" invitations if you and your friends are in a mall at the same time.
  

All of these are relatively small services that nevertheless would provide value to the average consumer. These might be enough to keep users engaged and appreciative, but maybe not quite enough to generate revenue from consumers. What types of services might a consumer be willing to pay for?

• Tickets to a cool concert (or reservations at a great restaurant) near your current location when you're looking for something to do
• If you have tagged people on your "friend" list as "want to meet soon", proximity notification and suggestions on where to meet when they're near you
• Pedestrian route recommendations that can be customized for weather or consumer preferences. There's a company in Japan, Navitime, that provides exactly this service. I don't speak Japanese, but the idea is quite clear from that page! Having said that, the US is not exactly pedestrian-friendly, so it is unclear if this will be a mass-market product here.
  

LBS can build and retain consumer traffic, assuming they don't emulate the social networks' stance of providing the platform and putting the onus on others to build services. Here's to consumer adoption :)

Commercializing technology

In which we ponder the gulf between technology and product

I was having a conversation this morning with a friend who mentioned a startup he knew of. The company has developed a new technology that converts biomass into fuel (don't ask me how, I don't know the details yet). A similar company in Eastern Europe has been successful, so the investors believe that given the right GTM (go to market) strategy, this company could be a commercial success. The problem? The technology they have today needs economy of scale i.e., it needs to process a LOT of biomass to make fuel, and this means you get a BIG machine that costs a lot of money. The investors are interested in affordability i.e., make me a small machine that one family can buy and use. Result: a startup team that is focused on the cool new widget they invented, and potential investors who are (rightly) worried about commercial viability.

As a thought exercise, let's see how this might play out.

This is not an unusual situation. In fact, you see it in Silicon Valley about a dozen times a day. Startups are almost always about cool technology. They're not thinking about commercialization i.e., making money. At some point, the startup team realizes (or the investors help them realize) that money needs to be made. How? You can go both ways: you can take a product and build a business around it, or you can identify a market need and adapt your product to it. Which approach is better? That depends on your particular situation. A machine that converts biomass into fuel can be either (1) a small gizmo that you sell to every farmer in India for very little money or (2) a fairly large installation that is sold to large farms, co-operative societies or even joint families and villages. (I'm over-simplifying here. Obviously, there is a spectrum between these options and the final product can be anywhere on that spectrum).

In (1), you want a truly retail product that can be manufactured, shipped, marketed, sold and supported at pretty thin margins. In this case, your "big money" is going to have to come from scale. Assuming, of course, that you make that commitment to go out and sell the thing by the tens of millions. If you're trying to cover a country like India with a small retail product that will be sold to farmers, what should you worry about? You should worry about being able to find and reach that customer base. You should worry about the logistics of getting your product from manufacturing to customer. You should worry about how this product is going to be maintained and serviced. And how you're going to do all this at very low cost per device. Is the retail product going to be robust enough that you expect a very low failure rate? Is it simple enough to repair that every village cycle shop guy can repair it and you don't need to worry about building a service organization?

In (2), you have a little more wiggle room. Each installed product has to be profitable on its own: so you need fairly hefty margins (remember that you're factoring in the cost of getting it there, setting it up and probably servicing it too). Now you're less worried about finding new customers and more about scaling carefully and maintaining your customer base. What is the lifetime of this product? How long before you have to replace it? How often do you have to service it? Do you price lifetime service into the installation or do you take the product + service fee pricing route?

Neither case gives you easy problems to solve. These are fairly standard business principles. The trick is to remember these and apply them rigorously when you're looking at that shiny new gadget that's _just_ short of where you need it to be for the business model to work.

The long tail of technology

The editor of WIRED magazine talks about the long tail of technology. Briefly, his point is that every major technology goes through four stages in its life cycle:

• Price falls below a critical amount
• Adoption gets to critical mass
• Replacement of another (previous incumbent) technology
• Price approaches "free"

This talk is from 2004. Four years later, it is interesting to sit back and see how many of the trends he pointed out in this talk fulfilled their potential in the interim. Accuracy of prediction nothwithstanding, the "stages" he outlines in this talk is a useful takeaway.