Our organisations have mountains of data - and with that comes many challenges and considerations. UKISUG Director, Alan Bowling tells us more.
One of the common themes over my years in IT has been the growth of data. Sadly, I’m old enough to remember getting excited by the launch of the 5.25 inch floppy disk with a staggering 1.44MB of capacity, and then equally as excited by my first IBM PC with a 10MB hard disk. But now I casually walk around with 128GB of data in my pocket just on my smartphone. Statistics say that the human race now generates over 90% of the entire data in history every 2 years (no idea how they get to that but nowadays it actually does sound believable). Pushing that growth is the ever increasing number of connected devices estimated to be in the region of 50 billion by just next year.
So our organisations have mountains of data. And with that comes many challenges and considerations.
Data that is correct
I’ve lost count of the number of times in software testing that you think you have a problem with the code only to discover the source data is wrong. I won’t use the word corrupt as some would, as that implies that it happened by accident. Sadly, it doesn’t happen by accident (well very rarely in my experience). In one organisation, although they preached “one version of the truth”, the business analysts actually uncovered over 15 points where different groups could change what we might refer to as master data. It’s a bit like having an old fashioned library and instead of one place on one shelf for authors beginning with A, you have 15 places on different shelves for the same book. And then you wonder why you can’t find what you are looking for! The starting point for everyone is their data – don’t blame the software solution as many do, blame the discipline and structure of the organisation.
Actually an organisation doesn’t really want data, it wants information upon which to base decisions. And this is why all organisations should have, or invest in if they haven’t, coordination and control functions for data aided by core data management solutions. If you don’t do this you get what you deserve. Apparently bad data in the US alone cost $3.1 trillion last year, now that’s eye watering…..and imagine what people could have done with that!
Data that is safe
Ownership of data, whether it’s your organisation’s or someone else’s, is a huge area for concern. With the advent of GDPR the consequences for personal data breaches are also eye watering, not to mention career limiting! So expert penetration testing, and thorough application testing are vital. And then there is your organisation’s data, the thing that makes you different to your competition - securing that is also essential. I find it odd that many organisations still do not classify and mark up their material to indicate how it may be used, and apply protective tags to sensitive material. It’s one of those things that everyone hates doing, but woe betide you if you don’t and it gets out.
Where to keep your data and where you send it is increasingly becoming a challenge, often the cheapest locations just don’t make the grade. Sadly, that frequently gets the IT people in an organisation into conflict with the financial function. I found that having a risk appetite statement, which defines the very top of the organisation is prepared to accept, goes a long way to overcome this challenge.
Data to make decisions from
I’ve already said that actually data by itself actually isn’t that helpful, but information is. So key is to have is some form of collation, comparison, and perhaps conversion solution in your organisation. This is where the fun starts, because everybody in this day and age can do that – thank you Microsoft for making all our lives easier! The trick now is to provide tool sets that enable the end user to source and analyse the data from one location, or multiple locations that fit within a controlled data arena such as I discussed earlier. This is actually, for me, one of the big reasons for attempting to keep the data core strictly controlled and something that SAP, if employed correctly, can do very well. In terms of data analytics tool sets, it makes obvious sense, if you can, to keep it in the SAP family but they have to have things that the end user can easily work with. Gone should be the days of IT departments reporting developers.
Data makes the world go round, especially as more and more is generated in our shrinking connected world…..but only if you look after it and use it wisely.
Keep up to date with latest news, views and opinions on SAP-related topics, with our all new SAP Musings category on the blog.