Driving Technology – theory versus practice

Keeping control of your organisation’s IT estate is a fundamental competency for any CIO.  Most end up centralising control of technology to better manage costs and leverage an economy of scale of sorts. An extension of this is the concept of the ‘dominant stack’. This is where a set of technologies from (usually) a single vendor and can be leveraged across the entire Agency.  

The theory is you have less technology integration (therefore less costs) and it also enables you to standardise on the skills of the staff needed to support the environment. In practice however, integration is still required, and getting staff with the skills you need to support your typical enterprise class implementation are expensive, hard to find and even harder to keep. 

The management challenge this presents has driven many organisations over the years to outsource their support and development capability. But all this has achieved is even higher costs and de-skilling of their internal workforce. Certainly in the Public Sector, I don’t think outsourcing has achieved what the theory promised. In practice, it has stifled innovation, actually increase costs and worst of all de-skilled its workforce.

The dominant stack theory is used by most large organisations in both the Private and Public sectors. But in today’s world, is such a strategy still valid? Many are indeed starting to question the logic. If it’s expensive, inflexible and not able to meet the demands of modern business, naturally questions are going to be asked. But what are the choices when companies have invested millions of dollars into legacy technologies?  

In many cases, I see it analogous to the decision about when to trade in your car. Eventually the car is no longer fit for purpose – your needs have changed as life moves on. It also doesn’t have the ‘mod cons’ that makes driving more comfortable, safer and more reliable. You are undoubtedly going to lose money on it as well. You can continue to maintain your old car but parts and skills to fix it is becoming harder to find and your costs inevitably increase.  

There comes a stage where you start to think about pouring good money after bad. This is usually the point where other options need to be considered and a decision on future direction needs to be made. There are many emotions involved as people rely on having a car and would find it difficult to live without it.  But there finally comes a time when hard decisions need to be made. As times and needs have changed this is a logical point to investigate new approaches.  

With cars there are so many options to choose it can all seem overwhelming but if you break it down into what you really need, and what would be nice to have it does become easier. It also helps in taking some of the emotion out of the purchasing decision. 

When looking at an IT investment, there is common attributes that all businesses need: flexibility, agility, stability, and increase in speed to market. It goes without saying that these capabilities must also be delivered cost effectively. At the end of the day, like buying a car, there are trade-offs, compromises and marketing hype which all need to be considered before making future investment decisions. There are also purchasing models that need to be considered. For example with a car, do you lease, hire purchase, personal/business loan, extend the mortgage or just buy it outright? Or do you even need a car now and just use Uber and/or public transport instead!

At least in the technology space, one (relatively) recent change that has helped in the decision making process is the advent of the Cloud. A Cloud model provides a more cost effective way of trialling software change without the overhead and build costs of traditional software systems. It can be like taking a new car home for a period to see if it does, in fact, meets your needs. This will help in making informed decisions in an environment of many options, approaches and models. It can also be used to help put more clarity around user requirements through a process of live ideation. I have seen this being used to great effect in recent months using the Oracle cloud. 

But like any technology, using Cloud needs to be managed and as it is different to the traditional on-premise model, new processes (including governance) needs to be implemented. This will enable the CIO to keep a level of control by minimising technology spread and associated spend.