In the first Industry Insight in this series, I looked at the history and evolution of cloud computing, showed there was nothing new under the sun, and began looking at a South African approach.
In this, the second in the series, I'll examine how South African companies can exploit cloud computing, and how they are in fact doing so uniquely.
The starting point is to recognise that the way software is being delivered today, and has been delivered for decades, just doesn't work. For a few years, the delivery rate of software projects improved incrementally, from an 80% failure rate to around 60%, but in the last two years it has slipped again to around 80%, as measured by the Standish Group in its Chaos Report 2009.
Malfunction
To understand what Standish defines as "failure", have a look at five measures:
* The software project exceeded budget;
* It ran late;
* It did not meet requirements;
* The business abandoned it; or
* It was signed off but never used.
While the Standish Group tends to focus its research offshore, I can confirm that many software projects - some of them high profile - easily fit into one of the brackets described above.
One was a banking project that was abandoned after R200 million expenditure; another, widely publicised was the Cape Town City Council ERP project, which chronically over-ran budget and caused huge embarrassment as salaries could not be paid; yet another was the Gauteng drivers' licence project, which over-ran budget quite dramatically, and has since crashed repeatedly; and another was that of a 'top four' bank, which was unable to deliver on its commitments to an offshore credit card provider.
In all of these cases, the risk resided 100% with the customer, which had to pay up-front for software, including licences. The solution provider in each case, as far as is known, assumed no risk. Its principal was paid. If the project ran into problems, as we know all these did, there may have been penalties, but the developer or solution provider by and large earned its cash, and the customer was left with egg on its face.
A far better approach is to defer up-front payment on licences and to share risk.
Richard Firth is CEO and chairman of MIP Holdings
A far better approach is to defer up-front payment on licences and to share risk. Here, in essence, is what I suggest (and I do not imagine for one minute that the bulk of the software industry will follow suit):
* The principal defers up-front payment. This is typically on the basis of software licences not being charged for before value is derived. And does this not make perfect business sense? Why should a customer pay millions for something from which they have not yet had any benefit, and may not ever have?
* Rather, defer payment for the software until the solution goes live. Given the quarterly pressure to deliver results to shareholders, it is unlikely that most software principals will agree to defer up-front payment, but in so doing they would be short sighted.
* The reason for this is that the current model prescribes money up-front, whereas the new model requires payment amortised over many years - in fact, over the life of the contract. And payment is dependent on value delivered - the success of the customer's project (the growth of value delivered in line with the customer's strategy, as per agreed deliverables). The greater the value the customer derives, the higher the monthly payments. This gives the solution provider and principal a direct and enduring reason to remain involved and acutely interested in the customer's business success.
The industry model of cloud computing says software services must be delivered via the Web. Nowhere does this model say the customer's business must benefit from this new model. Surely it makes more sense to bring the customer's requirements, satisfaction and benefits in line with solution delivery?
* In the next Industry Insight in this series, I'll explore this approach to cloud computing further.
* Richard Firth is CEO and chairman of MIP Holdings.
Share