Q6FSA Approach - Mastery of Data

Forget Master Data or MDM domains, follow the money.

I'm probably going to ruffle a few feathers in the master data management industry with this one...but here it goes...you're all missing the point...let me explain why.

The real business value is in recording, understanding and managing provenance from the staring point of acquisition and subsequently throughout the entire end-to-end data value chain in all the variation and entropy of the complex enterprise.

Therefore, the management of mastery should be where information and business architecture disciplines intersect and neglecting either can actually result in diminished value, a solution that neither manages business data nor processes effectively. A race to the bottom of the minimum agreeable common denominator...but leaving the exceptions and variations unreconciled.

The issue here is in thinking that consolidating data into a single centralised data model, location and set of processes solves the entire enterprise scope provenance and interoperability problem, even within a single domain, any better than managing the data better where is it, keeping data close to the tribes that understand it the best. The business has silos for a good reason, they support essential localised business variation and tribal communication.

Striving for a single view of the truth is understandable, data modelling theory for decades has been telling us this is the best thing to do. But this is counter intuitive to the way that businesses generally operate, in functional tribes. It is surprising that such approaches have lasted as long as they have, given the litany of MDM failures...

The root problem is and has always been successfully combining structural and semantic interoperability...enabling localised tribal, consolidated and enterprise view points to be interoperable whilst retaining their unique language, value and process differentiators...and we at Q6FSA believe that data need not be either consolidated, moved or re-shaped to solve it. Here is how...

Part 1 - Not just "Master Data", ALL data starts somewhere.

The MDM industry focus on the enterprise consolidation of the specific category of "Master Data" is typical of the way that data practitioners and vendors try to contain scope in order to justify the cost, activity and the value of "doing something now" by introducing even more technology, creating another silo and, in the end, not necessarily cleaning up the entire mess as the business expected.

"Master Data", in theory, is an easy target. Slow moving, low volume with some overlap of structure and use across the enterprise. Low hanging fruit. "We have to start somewhere!", I hear you say.

But this category of data has exactly the same need for interoperability as every other category, there is no difference. And surely, more strategic impact and business value would be gained from improving interoperability across, say, transactional data, because there is much more of it and it relates more closely to business value chains, such as procure to pay.

So let's not keep inventing jargon to slice up the data industry, data is data. Period. And all data needs, is to be made semantically and structurally interoperable, right?

We need to work the whole problem and find a common approach for the enterprise, starting with how to identify and manage all the points where any items of data first physically manifest themselves, get to the full picture of data provenance...find the mastery of all the data.

The Q6FSA approach and platform does not differentiate between "Master" and any other type of data. All data is treated equally.

Part 2 - Provenance, Quality and Governance must be Upstream

If you believe Google Images, this is the source of the River Dee, the place where the river...becomes a river. The fresh and mostly unfettered water is tumbling straight out of the moorland, its short term purpose is clear, to make its way down this hill, letting gravity take its course. Some rivers have multiple sources, each contributing to the passage of water as they join together. All rivers adapt to the laws of physics and their immediate environment as they progress. The long term purpose is always the same, to carry water back to the sea and continue the age old water cycle.

The characteristics of this river source, however, are different to the same river down stream...here, its quality can be easily measured, it has a clear point of provenance and its governance and purpose has little ambiguity. These characteristics are much more difficult to determine downstream as rivers join and humankind adapts it to harness its core purpose for wider use. External factors, like geography and weather play their part also in confusing the picture.

Each of those downstream adaptions also has a specific intent and starting point, merging with the river to create both cost and value and by the time that this river gets to the sea...it is transformed, augmented and complex.

Such is also the way with data, at its point of acquisition it is easy to define and govern and as its journeys progresses, other data and processes shape its context and purpose. Some characteristics will remain common, others augmented and re-purposed as it flows.

Those common shape, semantic and purpose characteristics are, therefore, best defined and governed only once at the first point of acquisition and persisted downstream for as long as they remain relevant, which should be forever if consistency has been fastidiously observed.

This reduces the overall governance burden by not constantly redefining data that has not changed its characteristics or purpose in all the places it is found downstream. Quality is assured at the point where the data is most pure and meaning and ownership is clear.

This is the Q6FSA approach also, to ensure that effort is made to define and govern atomic characteristics, purpose and meaning at the earliest point of acquisition in the data landscape...or where we at Q6FSA assert, the mastery of data occurs.

Part 3 - Follow the Business Value Chains, follow the money.

This is an area that I really struggle with...the approach with which most practitioners, integrators and vendors attempt to solve the mastery of data through information domain consolidation and centralisation, for me, is virtually impossibly to justify in terms of holistic business value...or at least does not relate and combine well to basic business value chains.

This is why ERPs continue to sell so well, they talk the language of integrated business, end to end value improvement and arrange themselves to core processes and operational functions. Or in other words, they are business and not information domain aligned...they follow business value...they follow the money. An ERP captures, combines, augments and enriches key business functional journeys across domains, they may get many things wrong, cost masses and are prone to poor implementation, but this they get bang on...every time. Could you image what an ERP would look like with no concept of an end-to-end Procure to Pay flow?

Data is acquired, generated and combined at many points in such a business flow from many information domains, transactional and reference...new structural combinations are created and business ownership and governance needs to adapt with them. With this focus, mastery is augmented everywhere and closely linked to the business value chains...it follows the value and therefore the money...it can be justified as directly supporting business goals.

So, taking a specific information domain away into a new silo, attempting to consolidate it to a least common denominator structure based on a compromise of enterprise consensus, then trying to rebuild all the localised process and business language variation...losing its direct link to business flows, to us at Q6FSA, makes the exact problem that it is meant to solve actually worse.

Part 4 - Mastery is better left federated in the tribes

Enterprises are tribal, this you can not change. Despite our deep desire as data practitioners from our traditional theoretical training to make them all talk the same language, they won't...ever. Better just accept this is the case and save yourself many wasted years and wasted tears...it took me 20 such years to stop banging my head on that brick wall and now I just refuse to do it that way.

Let go of the single version of truth and you will be released from this world of pain...it does not exist...stop pursuing it...let the tribes do their own thing as they are closest to the data than anybody else, but more importantly, they own the variation and localised interpretation of it. We at Q6FSA assert that trying to remove that localised connection is a fools errand that, however well-meaning, has not worked in the past and will not work in the future.

Instead, you should federate mastery and its related quality and governance activities to the tribes themselves, give them the tools to do it and focus on then making their tribal view points of structure and language interoperate with others and the enterprise as a whole. This is not hard to do, strip away most of the technical IT jargon and most of the language of business has not changed for decades, possibly for centuries.

Q6FSA asserts that mastery can not be centralised, it exists everywhere and can not be contained in a single consolidated structure, it is a continual process of augmentation of cross domain data.

Spoon Boy had it right all the time... "Do not try and bend the spoon, that's impossible. Instead, only try to realise the truth...there is no spoon. Then you'll see that it is not the spoon that bends, it is only yourself."

Sorry! As I said at the start, you're all missing the point.