Mike Rowehl: This is Mobility

I posted a mobile portal list a little while ago. I’ve heard a lot of people in mobile say there needs to be “something like the early Yahoo for mobile”. Which is to say a manually moderated hierarchal list of all the cool stuff that’s out there. There are a few points that go into that which I’m not sure everyone explicitly acknowledges or would agree with if they heard it stated directly:

• The mobile web is still small enough that it can be organized by humans instead of machines.
• Mechanisms such as tagging and other collective action systems aren’t the right way to introduce a user base to a new medium. Just give them a top down tree of what’s out there.
• That there should be a winning system providing the directory.
• That linking out to other resources instead of trying to collect information is the right way to expose the mobile web.

What of that do you agree with and what do you disagree with? I think that the mobile web evolving along the same lines as the desktop web is not a foregone conclusion. I remember the evolution of putting a website online back in the day and sending out links, and then adding metatags instead, getting others to link to you, changing your own link structure and tuning for searches to get users. The stuff online evolved in a particular way because of the environment it sits in. Link structure analysis became the smart way to determine relevance because there’s a lot of cross-linking on the web. It’s human behavior analysis at root, cause most pages are either generated by humans are automatically generated based on the actions of humans. Search became the default method of interaction because you can extract keyword based relevance from the link structure, and at a PC you have this big keyboard sitting right there in front of you.

However in a lot of areas of mobile things have tried to hop ahead and jam themselves straight into the end behavior of the wired web. Mobile search for instance is always an interesting one. Take a look at most sites meant to be used from mobile devices. How much cross linking is there? Not too much. How do you extract link structure information from those sites? You don’t, so if you’re going to rely on link structure you have to go with sites that were meant for desktops and transcode them if you really want the results to be “relevant”. Which means that if we cede search the way it is now as the defacto method of interface with the mobile web we’re really giving up most of the mobile optimized sites there are out there now. “Superior user experience” or not, they’re not going to end up with users based on search traffic from most of the current services.

Maybe that is the right way to go. As much as just about anyone in the world I’ve ever met, I would like to see the world of mobile services grow. I would like it to happen in a way that the services are all the things that are particularly appealing about mobile (situated, always available, immediate, personal, etc.) But maybe the way to get there is to swing the pendulum out the other way for a while. Right now the mobile world is pretty small, particularly on the mobile web side. Few people publish mobile content when compared to overall web content, few users browse the mobile web on a daily basis when compared to the number online or the number holding mobile handsets.

Just thinking out loud here, I have my own take on what needs to be done to get the mobile services area to explode in terms of growth. I got the end of my newsreader today and was tired of everyone blathering on about services and products yet to be released, so I figured I would spew drivel in some other direction. In my opinion the stuff that we’re really missing is mobile publishing tools and toolkits for building mobile specific services. It’s still one of the most difficult areas to address. There are very few mobile specific offerings out there. Here’s what the stuff needs to be:

• It needs to be open source. New developers in evolving markets need to be able to self-service and innovate at whatever level of the stack they want to address.
• It needs to work symmetrically across all areas. Mobile services have traditionally been highly regionally segmented because they’ve always just broken down along carrier lines. The mobile web is global, developers need to be able to hit that global market by default.
• It can’t drag around a lot of web baggage. There are tools out there that allow you to make a mobile version of web content, not horrible for some circumstances. But there needs to be 100% mobile versions of the common tools. Forum and discussion tools that I can setup and use from my phone without ever touching a PC, content management systems that work completely from a phone, blogging tools that don’t treat mobile use as a second class citizen.
• It needs to allow for users who have no knowledge of mobile specific features without forcing them to dive in too deep, but not restrict users with domain knowledge from doing special tuning.

There’s a lot out there that lets you get your content up online, but that’s only part of the problem really. At the Mobile 2.0 conference we spoke a lot about growing the ecosystem. And this appears to be one of the areas that’s really missing. Having an ecosystem doesn’t mean letting people toss content into the mix, it means people coming up with their own services, building on what others have done, and contributing meaningfully to the base set of services available to others.

Russ put up a wiki for publishers looking to use Mowser. It’s got all the info about how to swap in your own ads on the mobile version, bookmarklets, how to setup a site so that Mowser is the default mobile version. I think the most important part is that the steps used in adapting content are described, so that you can know what to expect from the transcoder. The Wordpress Mobile plugins can be a pain, so I setup my AdMob site id in my headers and m.thisismobility.com is the mobile version now. I also turned off mobile sitemaps though, lets see how that affects things.

I sat in on the Co-Evolution of Marketing and New Media panel at Play while I was there, my coworker Niren was one of the participants. There were a few mobile specific points:

• Niren pointed out that one of the biggest impediments to mobile advertising right now is misunderstanding of the service. Most people equate mobile marketing with unsolicited SMS messages. Sure, that’s something that happens. Unfortunately it’s the attention grabbing part so it’s what people hear about disturbingly frequently. But mobile web advertising has nothing to do with that.
• Data tariffs are one of the other major blockers. High rates for data services keep away users, and the audience really needs to hit critical size in order for it to be a viable channel for major advertisers.
• Phil said that 9 months ago AdBrite was seeing about 2 million pageviews a day coming from mobile devices, and now they’re seeing more than 25 million. Fantastic growth. He doubts however that mobile as a distinct segment will remain forever. In three years time it’ll just be “advertising” without a distinction between mobile or desktop.
• To the specific question of “does mobile advertising go away” Niren said it could once the networks hit broadband speeds. But for most markets that doesn’t happen any time soon. And given that your mobile is a pocketable device there are constraints that probably dictate a different user experience.

Overall there was a lot of talk about engagement and interacting with brands. Examples of sharing and commenting on YouTube videos, running user generated content competitions, product placements. But most of the folks from large advertisers or agencies that I spoke to think that set of discussions is misplaced. Sure, there might be some cool stuff that an advertising channel can offer that differentiates it and allows for some interesting campaigns. But stuff of the sort tends to be once off runs meant to capture some buzz and ride the coolness factor. The core of the money needs to follow the audience, and the real story is about what mediums can get a large repeatable audience who respond well to offers. Provide that and the advertising dollars will follow.

One of the notification systems at AdMob went insane yesterday and as a result this morning I was looking at just over 19 thousand messages in my inbox. Not that I like dealing with my inbox to begin with. And Thunderbird was choking trying to deal with the volume. Not sure exactly why, but instead of figuring it out I just pulled up the Ruby interpreter and used the IMAP support in the standard library to delete the messages and expunge them. There are probably about a dozen additional ways deal with the issue, but I thought it was a cool bit of hackery and a quick practical usage example.

I moderated the Beyond Mobile 2.0 panel at the >Play conference yesterday and just wanted to get down some of the thoughts from the day. The conference sold out this year, great job by the organizers! The discussions were fantastic, the panels worked out really well.

Some of the points that came out of the Beyond Mobile 2.0 panel:

• Rick from Xohm clarified that the service is currently focused at other devices, PCs and PMPs and handheld game systems, not phones yet. I didn’t know that.
• The cost for mobile data services is still really one of the major limiting factors. Sure, we have flat rate data in the US, but it’s not available all over the world. And even in the US, the perceived value of data services is below the cost. It’s a kinda catch-22, where people don’t develop apps cause the data rates keep users away, it’s not worth figuring out how to lower data rates cause the users don’t have apps to use. The iPhone is helping to change that, not just the bundling of the service with the hardware, but that the browser opens up the services accessible so that the data rate doesn’t seem crazy.
• Everyone on the panel was into location based services in some way. Very much into them. No real answer to the common problems though, cost of getting user location on most networks, exposure of the info is very non-standard. But everyone sees the potential now.
• There was a general consensus that building more situated apps, apps that pull information about their context and the user’s intention and use it correctly, is the way forward in terms of evolving mobile applications. It was in that spirit that there was so much interest in location. But of course there’s other stuff like learning user behavior and habits, and lots of domain specific behavior based on the application. Jeff from CBS in particular was enthusiastic about using additional user info to build better services. He gave some examples from their sports properties, where the info the user is interested in is probably different depending on the day of week it is, if a game is currently going on or not, what they typically look for, etc.
• In terms of location it only really makes sense when location is always just there and available. Otherwise you can’t really base an applications behavior on it. And the way the location infrastructure and cost structure lays out, that doesn’t seem too realistic yet. Evan from Loopt says that they really need to concentrate on carrier by carrier rollout of their service because it really requires close ties to carrier infrastructure. While that’s the case I don’t really see mobile apps being able to evolve in the organic “remix” style that we need to a revolution in mobile the way we’ve seen on the web.
• Rick from Xohm raised a point about privacy vs personal information available for apps to build on. In general folks in the audience said they were willing to share some additional info with service providers so that they could provide better mobile apps. However there were also quite a few that said they were already somewhat uncomfortable with the info being shared out and didn’t want to release any more.
• The killer component of mobile has been and probably always will be communication. Data services that give you access to info are great, but the ones that are really going to change user behavior and affect people’s lives are the ones that connect people with each other.

There are a bunch of posts over at the DMEC blog as well:

• Beyond Mobile 2.0
• Who Goes There, Friend or Foe?
• Irreverence
• Reflections on Play

No matter how many times I see the stuff and never get a chance to play with it, it always makes me happy to see things like Samsung detailing plans for a wrist wearable cellphone hybrid. I don’t know about all of you, but I think it’s pretty hot. And I would try it even if people pointed and laughed at me in the streets. I know, I know, if Steve Mann has managed to assemble a wearable system I should be able to do so. But I’m looking for an option that doesn’t take a lifetime of research to assemble. You can only drop out of society to explore the fringes of technology so many times before people start to ask questions.

The N800 is a fantastic device for web and IM, but I’ve been poking around with how to make it more useful as a production device. Or something more like a personal dashboard. I would like to have the device sync with my server and other services out on the net when it goes online and pull in a bunch of info. Pull down stats from the web server, comments and links from the weblogs, a few slices of info from my inbox, stats from AdMob and Adsense, juicy stuff from the newsreader.

The way I’ve been dealing with that stuff in general is with an IM bot, which works out pretty well from the device. The IM support in Maemo is pretty sweet. However I’m having some information density and interaction issues. So I’ve started fooling around with the Python port to see if I can make something useful. Just need a few independently updating ‘windows’ and some nice small buttons to change views. Should work till I can get data glasses.

One of the big items for yesterday was the release of the Facebook Platform for mobile. Lots of people have picked up on the news, occasionally followed by “So what exactly is it?” It’s cool, but can be hard to understand if you haven’t already read up on the Facebook platform. And most of the news I read about the release makes me think that many of the people who have talking about it have done little besides see that Facebook released something mobile and immediately start gushing. So of course Russ and I had to hack around with it a bit and figure out what was in there.

The first thing to understand is that one of the main forms for Facebook apps (at least what I hear about and have seen most often) are the applications that put some info on your Facebook profile page. One of the things that Facebook has started allowing is markup within the snippets normally displayed on users profile pages to indicate what should be displayed on the users mobile profile page. Got that? You don’t setup a mobile web application that makes some calls to Facebook for services, which is what a lot of the postings so far make it sound like. If you’ve built a Facebook app before, or even read a bit of their developer docs, it all makes perfect sense. But if you’re normally doing mobile stuff and this is the first look you’ve taken, it can be hard to grok.

So what this allows you to do is build applications that have a presence on the Facebook Mobile side as well as the regular web side. If your next question is “why the hell would I want to do that?” you should really go and read about Facebook apps on the web side. The Facebook developer site has all the info about how to setup simple applications and get them running.

There’s also support for SMS interaction now as described in the wiki page. That seems to be pretty straightforward keyworded shortcode support with requests from end users getting sent to the callback URL for your application.

It’s interesting, but after digging in and checking out what’s in there most of the info I’ve seen online so far seems to be somewhat off base. Maybe it’s just the vagueness of journalistic reporting on tech, or maybe I was looking for some info to be in the writeups that I just shouldn’t have expected to see. Really what happened is that Facebook has added support for including user applications on their mobile site and opened up SMS interaction. Something I would love to see but have been having issues with is adding applications from the mobile version. Anyone tried that out and have it working? Because there’s no UI for it on the mobile version that makes me think it wasn’t meant to work, and trying to visit the app URL directly from my phone seems to verify that (I keep getting errors).

After we released the first version of AdMob Mobile Metrics some people took a look at the numbers and said “You served more than 650 million pages in the US in a month? There are only 240 million subscribers, that number seems really high.” Oh does it? How about those 240 million subscribers generating an average of more than 4 messages per user per day? Did you expect that? Think of all the people you know that have mobile phones and don’t use SMS. For me it’s my mother, my father, two of my sisters, my aunt. Think about who you get daily SMS from, for me it’s only 3 people. With about a dozen that I get a message from once every week or two. Everyone else out out there that does makes up the slack to make the average more than 4 per day per person. Now think that the average across all users for the mobile web is 3 pages per user PER MONTH! Still sound weird?

Take a look at Russ’s take on the numbers too. US subscriber count larger than what the census predicts for users of phone-carrying age? Where do all the extra users come from? Whenever people see numbers that don’t match up their first reaction is normally “Oh, the numbers are wrong”. The other possibility is that something is going on that isn’t expected or understood. Getting to the bottom of it is usually a pretty nasty process, but well worth it in the long run.

AdMob is hosting Lunch 2.0 on Halloween. Essential details:

• Hosts: AdMob US Headquarters
• Address: 60 East 3rd Avenue, Suite 225, San Mateo, CA 94401
• Date: Wednesday, October 31, 2007, 12:00pm-2:00pm
• RSVP on upcoming

There’ll be food, treats, and a presentation by Omar.