We believe that data should be a matter for the bosses and the top management, in every company. Knowing what data is collected and from which sources it comes should have the same importance as the expenses and revenues. The bosses should establish tools to track the quality of their data. The easiest way to get up to date about the own data are these 3 steps:
- Set up a data model
- Data quality rating
- Improvement actions
1. Set up a data model
First step to take is creating a schema about the data you collect. Write down all the external and internal data sources you collect or potentially could collect data. This can for example be data from CRM systems, website, app, mail clients like Outlook, client lists etc. External sources could be governmental data about demography or society, social media, area based data like temperature and many more. Search also for open data sources like opendata.swiss.
2. Data quality rating
Once the data model is set up it is time to dig deeper into its content. Therefore some statistical analysis is needed. The special interest is in finding outliers and everything that is not consistent, accurate or invalid input. Based on that analysis you will be able to rate your data on a simple 5-scale-rating. Based on how homogeneous the data are you rank them from 0 (heterogen data) to 5 (homogen data).
3. Improvement actions
Based on the rating you now have to prioritize the data by its importance to be improved and define actions for the highest prioritized items. Set up an action plan and timeline and monitore it. You should pay particular attention to improving the data quality and keeping it on a high level, optionally with support of automated processes.