CTO Blog

An unusual blog this one as it's not based on, or linked to online sources, but instead comes about from a series of meetings, including my participation in a roundtable event with a number of companies on Master Data.

I have been struck by the vehemence of business managers against the IT department on the topic of ‘master data’ which they see as holding back essential changes in ‘processes’ that they want to make to respond to business opportunities. Indeed if there has been one reason above all others that I have found at the heart of the ‘separation’ of the users doing their own thing with ‘shadow IT’ it’s been this issue. The whole idea of Master Data seems to be an IT department concept that they don’t recognise. Talk about the quality of the information they have, how up to date it is, their desperate need to be able to recognise events and react with decisions support and they are there with you. Mention the need to identify and work on managing Master Data Management, and they start to question the value (to them) immediately.

The more I think about this, the more I realise that this is at the heart of the whole concept of Information Technology. The change in name for the operation of ‘computing’ to ‘information technology’ was to focus on the ‘mastering’ of data. It was to bring back a reliable, single version, of the truth after the disastrous effects of decentralisation and lack of management of ‘information’, read data, as the new and innovative use of PC technology took hold. After all centralised mini and mainframe applications don’t have this problem in anything like the same way. And before anyone flames me, no I am not suggesting that the control of data is, or indeed the need for ‘master’ data, is in any way wrong, it’s just not at the heart of the web-based technology that is now driving innovation in the business.

What we, meaning the IT department, set out to do was to ensure that our systems contained accurate master data so that the many users could work from a common reliable data base, something that remains absolutely vital today. These systems are ‘transaction’-oriented and therefore data centric, so we must pay attention to the principles of Master Data. It's just may be the old challenge of securing the back office systems around common data used by various Enterprise Applications, is not something that matters in quite the same way to users in the front office who are looking to improve their decision making in response to events, frequently from a Web-based environment.

The new, and additional challenge, is to enable user, and business, ‘interactions’ around communications, content, collaboration, etc. over the top of the existing layer of IT of ‘transactions’. As someone nicely put it at a recent meeting; ‘the challenge is how to work out to ignore the nine ‘interactions’ and to find the one ‘transaction’ in the string. Quite right too from the perspective of the role and responsibilities of the IT department, but let's focus on the nine interactions for a moment. Firstly what would the nine interactions have been? Anything from instant messaging to email; looking up a web page to reading a WiKi, etc etc, as there has never been so many people, devices, etc available to the user. There are only two common principles; the first is that an interaction is based on the challenge of diversity in terms of people, events and content that need to be worked through, and second that this is all unstructured data.

I think there is another way of looking at this. The internal back office ‘transactional’ model with structured master data is a ‘push’ model for information, whereas the externally focused front office ‘interactional’ model is a ‘pull’ model searching not just for information, but for context. Hence why data quality comes up so highly on surveys about Business Intelligence, the challenge is how to verify the sources and accuracy of the information found. Data, or content, or collaboration, or contact governance are becoming the issue, it's not can I find it, but how can I verify it, before I use it as the basis of a decision that is concerning.

I see the term ICT, standing for Information and Communication Technology appearing in the press and blogs more often, but I think we have the wrong C, it should be Information in Context Technology, and that’s the same principle as Master Data, but a wholly different proposition in every other respect. One looks inwards and restricts users around a set of master data, the other looks outward and supports users with context around which they can make decisions. Controversial stuff eh?