Criteria for Cloud Service Selection

I received an email about Paymo v.3 which was recently released. Paymo is a time-tracking, project management and invoicing cloud-based app (with native applications for desktop and mobile). Paymo has been around a while (hence being on their mailing list), and they have an impressive feature set and really thought through usability. I would recommend using their app.

However.

However, while they are a cloud service, and very reasonably priced, the data is in their system. It is in a format that is neither open nor usable in a format either offline or with another system. And that is the crux, for me at least.

Cloud Systems Best when no Lock In

Cloud systems are best when there is no lock in. For example, Dropbox does not lock me in (other than perhaps being more widely available with third-party app integration). I could turn off and delete my dropbox account, and my data would still be mine, and still be usable by me (though some people with shared access would be annoyed).

The same with Github. My data remains mine and in a format usable by me, locally or with a different cloud provider. In a highly transparent fashion.

It is once the cloud becomes proprietary in terms of data that the problem begins. This is why Google Drive is fine as a file service, but not when thinking about actual documents, which require a completely manual process of saving-as and downloading to be able to use.

This is a key feature. The data must remain in an open format or be easily translatable into such a format. This is where Paymo could innovate as well, and I hope they do so.

Cloud Services - Viable Options

In sum, viable options for cloud services include:

  • Operating system hosting (open source operating systems)
  • Application hosting (open source applications)
  • File hosting (non-proprietary file formats)
  • Messaging hosting (open formats)

Now, this isn't to say I follow all of these guidelines, but these help identify the kinds of undesirable lock-in (where financial decisions would be based on lock-in and not on service performance).

Things like Gmail, Google Apps, Line (messenger), and even iOS and OSX have severe drawbacks that only performance and reliability can overcome. But when performance and reliability start to degrade, the pain and expense can become much more than anticipated. Too much risk.