Friday, December 12, 2008

If You Build Will They Come?

Though it now is apparent communications service providers will have to become managed service providers over the long term, the way the need for viable applications is discovered, thrid party applications can be developed and sold remains a thorny problem. 

And the problem is measurably harder on the mobile side of the business, if only because applications have be tweaked for every handset the apps are supposed to run on. For this reason, some developers may well find it is easier to work with fixed line providers, as crazy as that might sound. 

Nor is it going to be especially easy for independent developers to get business deals done. "For two guys in a garage to make five different code applications, it's very hard," says Mark Kvamme, Sequoia Capital principal. 

The dream is to have any application run on any device and over any network. Ideally that allows developers to concentrate on what engages end users, instead of how to develop and deliver the apps. Platforms with large user  bases will help. The Apple iPhone is the best current example, though many have hopes for Google's Android OS as well. 

But business models remain a challenge as well, as it is doutbtful advertising will support most of the new apps developers expect to make available. That means subscriptions, which in turns means a really-compelling value proposition and serious willingness to pay. Few apps so far have that sort of status. 

For that reason along, a focus on business apps would seem to make sense, though the thought probably is unappetizing for many developers. 

All of which suggests the managed services business has a rather large opportunity before it, if some of these obstacles can be surmounted. Namely, make the process of aggregating demand, then authoring and delivering services--with huge scale--and simply. 

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...