Congratulations, blog fans! We鈥檝e made it to the end (well almost, you have to read this one first).
Now, if you鈥檝e ever worked with or for a bank you鈥檒l know that over the years that banks data collection will be huge. There’s loads of it. An amount I believe is scientifically referred to as 鈥渙odles of data.鈥� It鈥檚 all stored somewhere, whether in data stores locally or offshore, or master data management systems in a private cloud, or a combination of most of the above.
Where鈥檚 your bank’s data? 聽And it鈥檚 in loads of different formats.
This is because making data consistent is hard, and matching data is hard; implementing new policies around the collection and storage of data harder still.
So, when the bank eventually sets up a major IT project to replace its thirty data management systems with a new one, in time what tends to happen instead is that the new system becomes number thirty-one.
But for any of our banks who really want to make New Year鈥檚 Resolutions they can achieve, it is time to stop thinking of data quality as something for IT to sort and realise that聽data quality is a business asset.
If the data quality is poor, questionable or unknown, it should be considered equally as damaging as, say, credit risk would be.
The theoretical jump to this mindset is reasonably easy; bringing it into reality is perhaps on the trickier side. This can be because IT departments are typically tasked with storing and securing the data, yet business units want to use, analyse and evolve it, opposing views which can cause tension.
It鈥檚 the same idea as committing to eat calorie-controlled lunches, but then getting home and eating the leftover cream-and-deliciousness dinner your flatmate or partner had leftover from the night before: unless the policy applies across the board, compromises will ensue and goals will be missed.
If the bank treats its data as an asset and makes perfecting data a responsibility of every business unit and individual, it means that everyone is focusing on addressing the core problem and making the data the best it can be. But in order to do this, the business units are going to need tools they can use themselves, without expecting developers or programmers to do it for them: after all, banks can hardly be expected to roll out a developer per team to code and script bespoke rule sets!
These tools have to be able to measure the underlying data in terms of absolute quality and also tailored to the teams鈥� individual requirements. In that respect, it鈥檚 no different to the importance of knowing precisely how bad one of those delicious doughnuts is going to be in terms of sugar, fat and carbohydrates: everything must be measurable if a beneficial and life-changing adjustment is going to be made.
In the EU there are regulations concerning the display on food packaging of key information concerning nutritional value to set quantities (100g, ml ~3.5oz, fl oz), to enable better-informed decisions to be made. Banks seeking an equivalent measure for data quality can use something like the聽 standard to give clear definitions of data quality so that a聽baseline measurement can be established聽into the underlying health of the data at hand.
This way, the bank can monitor what it鈥檚 eating聽补苍诲听what it鈥檚 already eaten; it applies equally to measuring data already in the system聽and聽the new information the bank hoovers up each day.
Lastly, these tools need to connect equally as easy with the bank’s data systems, because 鈥渞ip and replace鈥� is an extremely costly activity to get right and an . Fast plug-in connectivity also means that when plugged into a remediation loop, it sets up a natural virtuous circle that shores up and improves the bank鈥檚 data assets.
For this and all our resolutions 鈥� whether聽,听聽or eating more healthily 鈥� all that remains to achieve these in the near term is the classic final line from the bottom of any diet recommendation:
Willpower Required!
If you’ve read something here that’s piqued your interest, even if it’s just in doughnuts, biscuits or The Crown, 聽and we’ll have a chat.