A Short Post
I moderated a panel discussion this morning, something that goes with the job I do. It was with the Massachusetts Technology Leadership Council (Mass TLC) and the event went really well. People were engaged, there were good audience questions and the panelists gave well considered answers.
The title of the session was “Finding the Right Way to the Cloud: Measure, Manage and Monetize for Success” and I was told it would focus on the financial and business model aspects of becoming a subscription company which it did for the most part. What interested me and what motivated this piece is the quality of the discussion about technology.
In a discussion that was supposed to be about business there was a lot of technology talk. We were ostensibly talking about metrics for managing the business like churn, deferred revenue, recurring revenue, renewals and more. But the technology discussion was also about monitoring API calls in a public cloud and other things that never come up in discussions about companies like Salesforce, NetSuite, Oracle, SAP, and Microsoft because those SaaS applications hide all this complexity form the user.
I don’t pretend to understand the significance of an API call and the costs involved, but the panelists seemed to. In the middle of this I began thinking that this is the price you pay for going for an infrastructure as a service solution. I suppose there are situations when a company might want or need a bare bones hardware solution. But it seems counter intuitive to me that anyone would give up a data center only to take on this level of complexity. What is saved or gained with this down in the weeds approach?
You must be logged in to post a comment.
Renting MIPS?! Where are they now?
Interesting discussion on your panel. I agree with you that discussing API calls is in weeds that shouldn’t concern business decision-makers. Reminds me of a panel I was on 8-10 years ago on the subject of “utility computing”. I was representing salesforce.com, and other panelists were from IBM, HP. etc. So the entire discussion ended up being about “renting MIPS”, as opposed to the business or higher-level technical advantages of utility computing.
I think the focus on API calls (and how to limit or “throttle” them through “governors”) reflects limitations or at least fears about limitations of the vendors’ own infrastructures, and any potential customer should ask themselves whether they should do business with such a vendor if the vendor has so little confidence in its own capacity to handle volume.