“That’s not a company, that’s a feature."  So goes a frequently heard VC comment on business ideas that are incredibly narrow and don’t seem to deliver enough value by themselves.  I have certainly made that comment in the past.  But I believe that cloud computing is changing what’s a feature and what’s a company: a feature may now well be a company.  How?  Well, there are many features of web apps and services that are hard to build, operate and maintain but are incredibly useful.  So if you can build that feature on a cloud computing platform, operate it very efficiently and offer it via API to lots of others you are likely to have an excellent business.

Here is an example.  Lots of sites and services would like to be able to process responses to emails.  For instance, disqus lets you respond to a comment on your blog simply by responding to the email alerting you of the comment.  Basecamp lets you add to a project discussion via email.  For endusers this is a great feature because it means you can often get things done without needing to switch from your email app to your browser, which is especially handy when you are on a mobile device.  But providing this feature is a bit painful because you need to separate these responses from junk your service might be receiving (including bounce messages), figure out who the original email was to and what it was in regard to, and then reliably extract the response information.  So a "feature” that lets you reliably process email responses would be a great service that a lot of sites could use and in fact pay enough for to sustain a business.

“That’s not a company, that’s a feature."  So goes a frequently heard VC comment on business ideas that are incredibly narrow and don’t seem to deliver enough value by themselves.  I have certainly made that comment in the past.  But I believe that cloud computing is changing what’s a feature and what’s a company: a feature may now well be a company.  How?  Well, there are many features of web apps and services that are hard to build, operate and maintain but are incredibly useful.  So if you can build that feature on a cloud computing platform, operate it very efficiently and offer it via API to lots of others you are likely to have an excellent business.

Here is an example.  Lots of sites and services would like to be able to process responses to emails.  For instance, disqus lets you respond to a comment on your blog simply by responding to the email alerting you of the comment.  Basecamp lets you add to a project discussion via email.  For endusers this is a great feature because it means you can often get things done without needing to switch from your email app to your browser, which is especially handy when you are on a mobile device.  But providing this feature is a bit painful because you need to separate these responses from junk your service might be receiving (including bounce messages), figure out who the original email was to and what it was in regard to, and then reliably extract the response information.  So a "feature” that lets you reliably process email responses would be a great service that a lot of sites could use and in fact pay enough for to sustain a business.