A colleague said recently – “We need to concentrate on fighting the crocodiles nearest the canoe”, which is both useful to remember and also one of the reasons why some organisations struggle to address their data quality problems effectively.
When you do not have enough resources to deal with all the challenges and issues facing an organisation (which is typically the case for most organisations), you need to prioritise the resources you have on those tasks that are highest priority.
Although there are many ways to prioritise tasks, these methods can generally be simplified to an assessment of the Urgency and the Importance of the task. You will generally want to address the Urgent and important tasks first, followed by some mix of Urgent tasks and Important tasks. There can be a tendency for organisations to address the Urgent tasks ahead of the longer term, but potentially more Important tasks as they sometimes believe that more time/resources/labour will be able to address the problem later, or that the nature of the problem may have changed.
Unless things are really bad, improving data quality is almost never an Urgent/Important task to resolve. Typically, it is also usually not an Urgent task, unless auditors or regulators are involved! So improving data quality is one of the Important tasks which struggles to get resources allocated to them.
So, back to those crocodiles, if you were in a canoe with armies of ravenous crocodiles eyeing you up as their next meal, you would naturally choose to shoot those nearest to the canoe. However, what if all the crocodiles nearest the canoe are relatively small and young crocodiles, whose mouths are not very big, but over on the bank is the big bruiser of a crocodile who is big, strong, experienced at consuming unwary travellers and therefore takes their time to join the fray. If this big bruiser is also a female croc, then over time she will create many more crocs to deal with. In such a circumstance, you would try and devote some of your efforts to distract or otherwise put off this mean old croc, whilst still dealing with the closer young crocs.
Improving data quality can be likened to one of those big crocs – not close, difficult to deal with and likely to become a much larger problem if not dealt with. Strategies to weaken the big data quality croc can include:
- Gather the ‘low hanging fruit’ by allocating some resources to deal with those data quality problems that are easier to resolve and have a higher benefit
- Train key staff on the importance of gathering the right data and treating it correctly
- Ensure that standards and specifications for what data needs to be captured and to what quality levels are up to date and easily accessible in order to reduce the rate at which new problems are being created
- Piggy back on other business activities – if a surveyor is visiting all sites to assess building condition, say, why not get them to do some simple data accuracy checks and to fill in any data gaps whilst they are on site?
- Establish a data governance function, or similar, to agree prioritisation of activities and to ensure that there is better awareness of the nature and impact of the data problems
- Incentivise staff to find and fix data problems – perhaps a personal reward, or recognition of their work in internal communications?
What tactics do you recommend to deal with data quality crocodiles?
Please note, no animals were harmed in the creation of this blog post!