Best-of-breed vs mobile application suites

Many companies developing mobile applications are currently deliberating whether it is better to deliver one or more best-of-breed point applications vs. a suite of integrated applications.

The iPhone AppStore is dominated by best-of-breed applications, where each icon on the iPhone desktop is a gateway to a focused, single-purpose application. Pandora is a great example, where a simple icon opens a killer web-radio application that is receiving rave reviews.

Examples of integrated suites include Lightpole and Where for the Blackberry Curve. Focus on location-based services, each of these applications acts as a portal into a suite of related, location based applications. Lightpole features 'channels' that offer me location aware versions of Wikipedia, Yelp and Yahoo! Local, while Where offers me a suite of location aware applications like StubHub, Zipcar, Flickr and Yelp and a set of location aware widgets like BuddyBeacon, GasFinder, etc. Unfotunately, it seems like much more time was invested in the portal and the suite, rather then the individual applications, and most of the applications and widgets in both these service are disappointing;

I'm firmly in favor of the best of breed approach, for a couple of reasons:

• One, the vendors focused on standalone, best-of-breed applications seem to be intelligently focused on building applications that will delight consumers and drive high levels of usage and engagement (See TechCrunch's review of Pandora). I'd prefer to see platform vendors and location vendors like Lightpole and Where  focus on enabling application vendors building best-of-breed applications, rather than trying to build suites.

• Two, we don't need any more portals like AOL. The beauty of the iPhone and new mobile browsers is that they offer unfettered access to an open internet, without the need to selectively port it onto different mobile platforms and carrier decks. Walled gardens are bad, and are finding more resistance than ever. The biggest knock on the AppStore is that it doesn't practice what it preaches, and creates a walled portal that is the only place to download iPhone apps:

"It's easy to argue that the App Store just trades one walled garden for another, but what the hell, we'll happily take Apple's over AT&T's." (from Engadget)

It's a sharp contrast to the open experience of downloading plug-ins to Firefox, and one I hope ends sooner than later.

 

---

Bottom line: I'm betting on best-of-breed.

Comments

[Doug Klein]

I would never have thought to compare our system to a portal. I don't like portals. They’re closed little worlds in which someone else decides what you can or cannot see. This is exactly the opposite of what we are providing at LightPole. We are building enabling technology that allows a much wider variety of content and applications to be delivered to the mobile phone. You could think about it as bringing web 2.0 to mobile (versus previous attempts at bringing “the web” to the phone). Let me explain.

There are lots of attributes to what people like to refer to as “Web 2.0”. The one that we are leveraging is the separation of presentation and content on most modern web sites. Web sites are now built with a presentation front end (e.g. RoR application running on/in an Apache httpd) talking to a database backend. As these system have matured, standardized external services have evolved out of them. These include RSS feeds and web service APIs. Our system takes advantage of these to create a “friction free” vehicle to deliver that content to mobile devices. Rather than face the painful (technically and financially) task of dealing with the multiplicity of handsets, operating systems and carriers, our content partners merely “publish.” We take it from there.

By utilizing modern web standards and our high performance content switching infrastructure we are able to make it possible for all content players to participate, allowing for the scale of a Yahoo! class publisher while creating the low entry economics that opens the market for the individual blogger. This is far different from a portal. We do not set the content or control the user choices. Indeed, our system already supports addition of geo-tagged content by the user with a simple cut and paste of the appropriate URL. Hardly a closed portal.

We do believe that our content partners will want to customize the content experience for their users, however. We have crafted our platform so that our partners can deliver content with the graphics and specific user interaction that fits their brand. But unlike a portal, this is open to all of our content owners.

That said, I do still believe that there are plenty of arguments for dedicated applications. You can absolutely tune the experience for the utmost consumer pleasure. But what if everyone on the web had to write a custom application to reach the user? Do you think that blogs such as this would have ever reached critical mass? I doubt it. The idea of services such as ours is to open the mobile opportunity for new ideas and applications. While possibly compromising the ultimate experience, we allow the type of innovation that brought us so much of the rich experience we call the web today.

[Antony Brydon]

@doug klein. You commented "Rather than face the painful (technically and financially) task of dealing with the multiplicity of handsets, operating systems and carriers, our content partners merely “publish.” We take it from there." I think this approach is great; the platform should have great value to web sites, services, and publishers, who should not be in the game of porting their presentation to every carrier and handset out there.

On the portal point, it's because I'm asked to open the Lightpole application on my Blackberry and select a site from a list of 'channels' that makes Lightpole feel like a portal, similar to the AT&T browser installed by default on my device. If I could access Lightpole enabled sites (i.e. Yelp) outside the Lightpole application simply by browsing the web on my phone - this would be the opposite of a closed portal. This may already be the case, I'm not sure.

[Doug Klein]

Ok, now I see where the portal analogy comes from. This characteristic (opening up a specific application to get to the content) is really more an artifact of the weakness of the current model of porting a standard web browser to the phone. It's a limiting approach since by definition a phone is a very different beast than a standard laptop/desktop computer. Simple things like the ability to dial a phone or call up a fully integrated map are missing (unless, of course, you are in control of the platform like Apple on the iPhone :)

Our current services include a purpose-built application so that we can deliver a more meaningful consumer experience but in reality we have a pretty straight forward API to our content network that we're always willing to open up to other applications and/or devices. Over time we believe that mobile browsers will embrace more "mobile relevant" capabilities and we'll be able to deliver our services to generic browsers. We also believe that there will be some very specific clients that utilize our network services on the back end while delivering a completely optimized, single-purpose applications (as you describe in the original post).

Over time the mobile world will include a spectrum of consumer facing solutions, ranging from very specific applications to very generic browsers. Our current mobile application sits somewhere in between, although when we call it a "geo-browser" we are very consciously thinking of it as closer to the browser than a specific application.

With respect to the idea of just surfing from the phone and finding content I would predict that over time, as the web converges to more standardized web services and reasonably geo-tagged content it will actually be reasonable to let the consumer "surf" from the phone. Today we accomplish this indirectly by allowing the user to select whatever content they want by "surfing" from our web site. Due to the fact that these interfaces still need a bit of hand-holding (i.e., just typing in some random URL typically won't give you the results you expect) it is still too early for the full generic browsing experience.