By Jaap Scholten, Head of Group Hybrid IT Strategy at Datacentrix
JOHANNESBURG, South Africa – There have been many plays, books and movies produced, featuring an anti-hero; a lead character who, despite perhaps lacking the qualities of a conventional hero, is at the centre of the story and thereby becomes the hero.
It can be argued that cloud fulfils exactly this role within the modern technology world. There is no IT conversation currently that does not at some point mention cloud. Yet, without good connectivity or proper security, cloud is of no use. So yes, we must talk about cloud, but all the supporting roles must be scripted into the story for it to make any sense.
Looking at it from this perspective, it becomes clear that data remains the golden thread that keeps it all together, so instead of cloud taking centre stage, these conversations should rather be data-led.
With real-time access to your company’s data – specifically unstructured data – machine learning tools can provide insights that are very specific to your business. Whether it is customer behaviour, procurement trends, stock levels and even staff behaviour, you can now have information insights at your fingertips that allows your business to adjust to these dynamics, hopefully also in real-time.
And where is all this heavy lifting done? In the cloud of course… This is where the business tools are located and from where they are able to access the data securely, with all your data from disparate systems aggregating in one eco-system – one cloud ecosystem.
So, a data-first strategy provides an end-to-end journey, or roadmap, to digitalisation. This includes the three pillars that support Datacentrix’ Hybrid IT strategy, being data hosting, data transport, and data security.
And yes, it is possible to put a timeline on a cloud journey. While tried-and-tested projects, such the moving from Microsoft Exchange to M365, are well-known and well-scripted, taking a business on a digitalisation path is a multi-faceted process. The size of the projects, as in how many apps, users and servers it involves – effectively the nuts and bolts of the IT aspect of the business – hugely influences the time required.
That said, proper transformation via digitalisation for an enterprise is typically a five- to seven-year plan. A good strategy takes time, even if you are a cloud-first business. Referencing Uber, the organisation took a year and a half to offer its first ride in San Francisco, and another year to activate in New York. The app arrived in South Africa four years later, and it took them a total of 10 years to list. That is a long ride.
Ultimately, you would want a digitalisation plan to help your business become more competitive, that almost goes without saying. You need to be able to learn from the data you have, about your customer behaviour, your vendor purchasing habits and more. If your aim is simply to get rid of hardware and virtualise this in the cloud, it can become a costly exercise. Also, that is not a business strategy, but rather an IT plan.
A clearly defined business goal – perhaps to improve competitiveness, drive down manufacturing costs, or move toward e-commerce – will allow for massive cost optimisation to take place. And by having all your data and subsystems in the cloud, you will have a more integrated data set on which to apply the insights that machine learning and outputs that artificial intelligence will deliver.
We have seen a trend where the net IT costs increase initially, but these are offset against market research, or even against the traditional discounting methods. For example, when you ask a data scientist to research your data in order to see what the company’s sales trend was last season, you have already biased the outcome.
The scientist will provide you with accurate feedback as requested, whereas a machine learning tool will not give you that answer. Machine learning will instead look at all the unstructured data and provide unbiased insights of trends that you may not have requested. Yes, it will provide last season’s sales trends, but may also highlight that logistics costs were at an all-time high, your cost of sale was greater than average, and because discounts were introduced, your margin was rock-bottom.
And cloud or no cloud, a result of not investing in business systems is the inevitable need to sunset business applications, often due to a hesitancy around spend. Yet, IT and business are intertwined – we are seeing that everywhere – and by separating the cost of a server from the application that runs on that server, we create these silos.
Closing down legacy apps and migrating the business functions served by them elsewhere is not an impediment to the adoption of cloud, as these processes can comfortably run in parallel. In fact, the hybrid cloud falls squarely into the philosophy of Hybrid IT, meaning that you can run what is needed on-premises, some co-located in an off-site data centre, and some in a hyperscaler or cloud.
IT costs and business costs are blending more and more, and the fixation on the IT component in isolation effectively misses the point. What is important now is to elevate the IT conversation to a business dialogue. Cloud needs to be part of that discussion, and while it might not take centre stage all of the time, it will certainly be waiting in the wings.
