Monday, May 25, 2015

Calculating the Potential of Cloud Computing

A recent Bain and Company report claims a number of insights about the move to cloud computing
  • Enterprises are realizing only 35% of the value from their workloads already in the cloud. 
  • Leading enterprise cloud adopters have migrated nearly two-thirds of their workloads to the cloud, yet the average company has only 18% there. 
  • Up to 50% of the value of cloud investments is predicated on streamlining and improving company operations.

How do they calculate that, asks IanCohen (@coe62)

The Bain report asserts some analysis and shows some graphs, but doesn't explain its calculation. The evidence cited comes from a survey of future intentions, together with a couple of IDC reports on the benefits achieved by an unstated (presumably very small) number of companies from some very specific technological changes (implementing Salesforce, replacing AWS). This doesn't appear to be a vast amount of relevant historical data.

Elsewhere, a 2011 survey for the European Commission estimated that, as a result of the adoption of cloud computing, 80% of organisations reduce costs by 10-20%. A similar figure is quoted in a KPMG paper published in 2014.

Calculating the potential of cloud computing presumably means projecting forwards from historical data (how much has already been achieved) to estimating the future opportunity (how much remains to be achieved). This kind of calculation typically makes a number of simplifying assumptions.
  • For example, when a typical organization has moved 20% of its workflow to the cloud, it has realised 20% of the potential benefits. 
  • Furthermore, the 20% of organizations that have moved 20% of workflow to the cloud are assumed to be similar to those that haven't.
Based on these assumptions, we could estimate the remaining benefits for the remaining organizations by simple multiplication.

However, both of these assumptions are implausible. Obviously the projects that are carried out are not selected at random, but are precisely the ones that have the highest payoff and the highest confidence level, so we shouldn't expect all potential projects to have the same payoff as the actual projects. 

More importantly, the people making the technology adoption decisions don't have this expectation. Nor are they likely to be persuaded by a graph, however beautifully drawn. However, if they want to adopt Cloud Computing for other (strategic) reasons, a few white papers like this may provide some intellectual cover.

Meanwhile, the European Commission sees Cloud Computing as a way of both reducing ICT costs and increasing ICT jobs. I don't see how companies are supposed to employ more people without increasing costs, but then I'm not a politician.



Syed Ali, Steve Berez, Paul Callahan and Vishy Padmanabhan, Tapping Cloud’s Full Potential. (Bain and Company, 2015) A downloadable version of the study is available here (8 pp, free no opt-in).

Jeff Atwood, Hardware is Cheap, Programmers are Expensive (December 2008)

Louis Columbus, Tapping Cloud Computing's Full Potential (Forbes, May 2015)

European Commission, Unleashing the Potential of Cloud Computing in Europe (September 2012). See also European Cloud Computing Strategy.

KPMG, Cloud Economics: Making the Business Case for Cloud (2014)


Updated 25 May 2015

No comments: