What challenges do you face and how do you resolve them?
Staying relevant means a digital transition. One way to achieve this is to work in an agile manner, and to use the possibilities offered by the cloud. But working innovatively and new technologies can throw up challenges.
Jan Aril Sigvartsen
of Basefarm lists the most significant of them.
1. New way of working
The transition from mainframes to x86 occurred some two decades ago. It heralded another major change. ‘Mainframe specialists regarded Linux and Windows more as toys than as real IT,’ recalls Jan Aril Sigvartsen, Vice-President Consulting and technology evangelist at Basefarm. ‘These discussions are currently underway again now that we’re transitioning to the cloud. It is essential to realise that as a company, you must be able to roll out features quickly, and to continue innovating. This keeps customers satisfied and takes on the competition.’
Contrary to tradition, nothing has to be completely ready immediately in the first iteration. That happens later. This requires a different way of working. ‘Many companies have difficulty with this. They regard AWS or Azure the way they would a traditional data centre, and re-use the tools and processes from their existing IT environments.’ Except you are then not using the cloud the way it is intended. ‘This way you come up against major problems. Because you are not using any native cloud tools, and you are not benefitting from the many new possibilities the cloud offers. A public cloud is not simply another virtual environment. It changes the way you work fundamentally.’
2. Separating standard IT and innovations
Alongside existing IT environments that are stable and standardised, with innovations you are dealing with new environments that have not yet been standardised, and that will only move to production gradually over time. ‘This means it’s extremely difficult to innovate on existing IT environments that have been standardised over the years and that have since become a commodity,’ Sigvartsen emphasises. ‘In the many companies I have talked to, this has gone wrong in 95 percent of the cases. So it would be better to do this in a cloud environment.’
That it almost always goes wrong is not because the required competences are lacking. ‘The reason is that for most companies, IT is only there to support the business. With innovations where you are deploying new technologies that have not yet been standardised, you have to be able to work through the entire innovation cycle.’ It must be possible to fail time and time again, to adapt to this, and then to keep trying again. Just as long as it takes until it works. ‘You can only permit yourself this if IT is your core business. If IT is only a support for your company, it makes more sense to do this alongside a specialised partner.’
3. Managed framework and new operational model
Standard environments that are stable in production are often accommodated with a service provider, releasing time and capacity for innovations and employees who are no longer distracted by daily operational concerns. You deploy cloud features strategically to achieve new business advantages. Consider an algorithm, for instance, to make your customers more satisfied than compared with the competition. Sigvartsen: ‘At the beginning you are probably not prepared to outsource such an innovative new environment to a service provider. You want to stay on top of it, and to retain technical control over it for at least a year.’
That is not to say that you might perhaps want to work with partners who will help you to build an operational model gradually. ‘What is still an innovation now must become stabilised and standardised over time. So that it can shortly be transferred to production.’ This is because the smart employees working on innovations are scarce, and you will need them again soon for new projects. So above all you should not bother them with operational issues like incidents. ‘Except that you cannot just re-use the operational model of your traditional IT environments. Supplementary to the PaaS platform, a managed framework is also needed which has been developed specifically for the cloud and working in an agile manner, and which contains among other things an operational model with which you can move to production at your own pace.’
4. Old finances new
Many companies decide that over the short term, a digital transformation is needed to be able to survive. ‘Here the existing traditional business always finances the transition to the new business,’ emphasises Sigvartsen. ‘So it is not the case that everything old suddenly becomes boring or superfluous and only the new is now hip and counts. It is also the expectation that we will be dealing with a hybrid world in the next ten to fifteen years.’
For instance, many business-critical environments are still running on mainframes. ‘You certainly would not place your new integration hub on this, keeping an eye on all your cool new applications, but mainframes do still play a significant role.’ So you should not opt for just one or just the other. ‘Many people forget that they are now able to experience this digital transformation, only because they have been successful with their traditional business.’
5. Separate development teams
Is it sensible to let a development team work on both existing IT environments and on innovations, or would it be smarter to deploy separate teams? Sigvartsen: ‘In principle both are possible, but companies where development teams are doing both come up against extra challenges. Because in the existing environments the developments simply continue and problems have to be resolved. This very quickly clashes with working on innovations.’
Initially new features do not yet deliver any value for the business. After all, money is earned through the traditional environments. ‘This means that innovations can quickly be assigned a too-low priority,’ warns Sigvartsen. ‘If management is not aware of this, it can lead to innovation grinding to a total halt.’ Because instinctively all attention and energy is focused on everything that has to occur this year or this quarter. ‘In my experience the companies that are the most successful are those that are assigned a budget and deploy a dedicated team specifically for innovation.’
6. Think outside the box
Many business owners are used to producing requirements based on what is currently possible in an already standardised IT environment. Sigvartsen: ‘Logically, that is what they have been used to doing over the past ten years. But to really generate new value from a cloud or Big Data context, you need technical expertise.’ Developers have this in-depth knowledge. They know, for instance, everything that is possible with the data that is publicly available, how this can be combined with one’s own data, and how you then translate this into business benefits.
‘It is specifically your developers who can come up with splendid innovative ideas, because they possess the technical knowledge of whatever is possible. Companies that are able to make use of the innovation and business skills of their development teams are therefore the companies that will gain a leading role in the market over the short term.’
7. Baby steps
Companies wanting to make the move to the cloud often hit the hiccup of how they should begin. ‘The best way to start this journey is to first carry out a simple iteration which immediately delivers a small benefit and already takes you a little way forward,’ suggests Sigvartsen. So above all, do not immediately transfer your entire organisation to the cloud. ‘Take baby steps when you are at the beginning of the journey; that is much more sensible. If you are lucky, you will immediately identify something that can produce a significant direct benefit for your business.’
For financial companies this might for instance be a smarter way of detecting fraud, or resolving a budgeting problem by using Azure’s analytics. ‘The cloud offers an enormous range of opportunities. As long as you perform small iterations, and in doing so you aim for a high value per iteration, my experience is that the journey will succeed.’