A blog by Martin Erasmuson.
The last couple of years has seen the emergence of a new role in corporate information infrastructure, the Chief Data Officer (CDO). Do a Google search and you will find a bunch of articles forums and conferences dedicated to this emerging role. And before we assume that is just some DBA role re-branded with a flash job title; think again. Last November (2016) Gartner predicted that “by 2020, 15 percent of successful CDOs will move into CEO, COO, CMO or other C-level positions” which is indicative of how strategic many organisations see the role of the CDO and by extension, their corporate data. It’s only taken three decades to finally work that out; that maybe, its not about the technology after all? But, how did we get into our technology pickle? Well, mainly through commitment-creep.
We’ll all have friends and/or acquaintances in relationships that are the end result of commitment-creep. It likely started with a meeting in a bar or social gathering. Then a second date? Sleeping together – cool. A tooth brush at my house, sure. A microwave, should be handy, why not? And a couch? OK. A bassinette!! WTF – but bubs got to sleep somewhere……….……… until one day, as Talking Heads nicely put it in their song ‘Once In A Lifetime’ – “And you may tell yourself, this is not my beautiful house! And you may tell yourself this is not my beautiful wife! And you may ask yourself, well how did I get here”? That’s what is called ‘commitment-creep’. i.e. you end up committed, but there was never a time when you actually made a conscious commitment.
That’s how most organisation got themselves into their de facto-relationship with their Information Technology. There was no conscious, strategic decision, it just crept up on them. And it all started a bit like the bar meeting I mentioned earlier, a single one-night-stand in 1992 with ‘Omega’ (that was her stage-name but she was really Microsoft Access tarted-up a bit). Fast forward 25 years and you’ve got the IT equivalent of a poorly renovated villa (Microsoft) with badly integrated extensions (the Unix bathroom), Solaris spare bedroom and iOS rumpus-room. And let’s not forget the kids that turned up along the way; COBOL, MS-DOS, Apple IIGS, Windows (95-2017), SAP, Oracle, Esri, and some of the new arrivals, SaaS and AI. And, as Talking Heads put it, how did we get here? Because while it’s only been about 25 years since that ‘holy-crap, what was I thinking’ night with Omega, we’d had other ‘technology-partners’ before her. Why was she so different? Let’s strap into the DeLorean and go back for a quick 50 year traverse of the evolution of information.
Back in the 1960’s, along with Hippies, Neil Armstrong, and Richard Nixon was the data store, magnetic tapes with files stored on a mainframe and the hard-copy ‘print-out’. Back then they didn’t necessarily consider the distinction between data and information because data and the process to extract it were all wrapped up in the same application.
In the 1970’s we came up with the first relational data models. That was the key turning point in the evolution of information because for the first time you could distinguish and separate information from the source data. It was now possible to create a whole bunch of different information outputs from the same source data based on how the data was retrieved and processed. With structured query language (SQL), different applications could access the same data. Into the late 1970’s and 1980’s, relational database models become best practice.
It was at this stage that data management began to stray from the virtuous path. It started with some harmless flirtation, an application created here and there with their own, independent data, often a copy of the source data. But then that ‘regrettable’ incident with Dbase III at the 1986 office party! Before we knew it we were playing fast and loose, building applications and copying data everywhere. Worse, it turns out Dbase III was just a promiscuous pre-cursor to the main event. Databases and applications were so easy to create, every department and user began creating their own.
And just when the Data-Prudes thought it couldn’t get any more wanton, that Dame-of-Data-Decadence arrived in town, the personal computer PC) and with her, Omega (Miss Access or MS as she prefers to be called). She was so easy that users with only a passing technical ability had the opportunity to create their own personalised databases filled with all manner of new, copied or modified corporate data. And they did. It was the data equivalent of a Roman brothel on ‘Two-for-One’ Tuesday night.
Sometime during the 1990s client/server architecture was declared Queen and IT architectures focused almost exclusively on desktop PC’s. Microsoft became the data-Pied-Piper as more and more applications moved to the desktop, the data quickly followed. Data Management, looking about as sexy as a road-accident, just couldn’t compete with the short-skirt-sex-appeal of low-cost processing and hardware represented in this new PC world. As silos of systems and information proliferated on every departmental server, and on every desktop, issues of data management were forgotten. IT departments continued to manage the main databases and infrastructure but had no real idea of who had copies, who had created new data, how, when or if it was being updated. And as network servers fill up with a myriad of structured and unstructured data, the typical response was to install more ‘cheap’ disk space in an attempt to keep pace. Virtual-Servers, The-Cloud with every conceivable “(any letter of the alphabet here) as a Service” (XaaS) offerings didn’t help. We are drowning in information and our on-going infatuation with technology is not helping. Time to tune into ‘RealityFM’. That one-night-stand with Omega in 1992 and the 25 years of commitment-creep that followed has turned into a sea of pain.
The silver-lining? Perhaps the realisation that Data, like the no-frills, plane-Jane, girl-next-door is, and always was, the girl for you. Marry her, have a bunch of kids: Taxonomy, Data Dictionary, Metadata Portal. One source of truth, many users.