Imagine the scene, you are a newly appointed/ promoted Data Governance Manager or Data Manager with a remit of ‘sorting out the data mess‘. Colleagues are impatient for results and some are starting to blame you for the poor quality data. This does not feel right to you, but you are unsure how to keep motivated through these daunting challenges and the opinions of colleagues.
A number of our clients have been in similar positions – we have helped them in part by explaining that their role and challenges are similar to those of buying an old house to renovate.
Read on to understand why.
|House renovation||Data quality renovation|
|You’ve just bought the house, it is in a poor state so there is lots to do to get it to the standard you are hoping for, so you start drawing up a list of jobs to do||You’ve just got the role, the data is in a poor state so there is lots to do to get it to the standard your colleagues are hoping for, so you start drawing up a list of jobs to do|
|Even though you now own the house, it is not your fault that it is in the poor condition that it is in||Even though you now are the key person responsible for data quality, it is not your fault that the data is in the poor condition that it is in|
|Before the surveyor has completed their survey, you decide to engage a roofer to patch some areas of the roof that are leaking||Before you have analyse the quality of the data, you decide to initiate corrections to obvious errors in the recording of office addresses to be undertaken by the facilities team|
|You decide that the priority should be tackling the rewiring||You decide that the priority should be tackling product master data|
|You engage an electrician to assess the wiring, replace end of life sockets and circuits and any wiring that has degraded||Your team use data profiling to understand product data quality to identify areas to address first|
|You decide to get the lounge usable, so engage a decorator to redecorate||You decide to get customer contact data fit for purpose, so get the team to review requirements and assess quality against these requirements|
|When stripping the wallpaper from the walls, large pieces of plaster fall off. The decorator has to stop work until you have got a plasterer to replaster the walls||When reviewing customer contact data quality, the team have identified many potentially duplicated customer accounts. De-duplication of customers has to be a priority, before updating contact data|
|An area of roof that was not patched has suddenly started leaking. All other work stops whilst a roofer fixes this leak||An overnight batch file load has corrupted 50% of product master data. All other data related activities have to stop while the product master data is corrected|
The above scenarios could be extended significantly, but you should by now have understood some of the key messages for a Data Governance Manager or Data Manager:
- You are not responsible for historic data issues and errors – you can provide guidance for resolving them
- You have a complex ‘brownfield’ situation with many issues to address – you help understand the importance of these issues and prioritise remedial actions
- Many remedial data actions may provide unwelcome surprises or require significantly more time/ resources than forecast – you can help provide clarity on the impact of these issues and reprioritise improvement actions
- You are not personally responsible for all data improvement actions – you coordinate resources of your team and other teams to improve the data
- Some data issues cannot be fixed until other data issues have been resolved – you help the organisation understand data dependencies to determine the optimum order of activities
- Fixing data is not enough – you should be ensuring that root causes are determined and actions taken to reduce reoccurrence of such problems
- This is a marathon and not a sprint, you are not going to sort things out overnight – make sure you get the ‘data foundations’ right and that the organisation understands it is a long term job
- Data will degrade if not managed effectively – maintain awareness of areas where data quality may be degrading
- Just because you have identified data quality issues, does not mean you have to fix them – support your organisation to prioritise corrective actions where they will deliver benefits
- Data on historic transactions (e.g. sales, customer calls) may be impossible to fix – don’t get hung up about it
So, if ever you are getting run down by the scale of your data challenges, remember, this is like renovating a house:
You are not responsible for historic data issues, they were created by others.
You can help your organisation prioritise beneficial improvement activities and secure resources, budget and support for the onward journey.
Above all, look after yourself and if you are starting feel overwhelmed or run down – reach out to others in the data community – we are a helpful bunch who may well have dealt with similar issues.