Data democracy or data anarchy?

Data is more important than ever for decision making. More of it is being created than ever. And silos between those who need answers and those who look after it must be broken down. By moving from the traditional model to something more democratic we can level the playing field.

Data Monarchy
Data Aristocracy
Data Anarchy
Data Democracy
Data Citizenship
Further Reading

Data Monarchy

Using data to make decisions is nothing new. The term Business Intelligence (BI) gained traction in the 1980s with increased volumes of data generated and decision-makers needing information.

More data and more pressure for answers meant data warehousing, ETL processes, and visualisation became more common. Teams of data analysts were created to support those processes and get answers to those crying out for them.

This traditional model created issues getting the information to decision-makers. Each new report required a fresh look at business logic and long lead time to create custom static reports.

Data Aristocracy

Organisations who have recognised the benefits of moving away from a data monarchy may have a self-service business intelligence tool. This is a good step forward but just getting trained in the tool presents a very steep learning curve.

Some organisations have broken down the walls and embedded data analysts in the teams who need data the most. Analysts report to their functional manager and have a dotted line to the centralised data or business intelligence function. By being embedded analysts gain an appreciation for the stakeholders they serve, and the subject matter they are dealing with.

The tradeoff in this model is that the tools are still in the hands of a few. Analysts are empowered to provide excellent service but may become a bottleneck with increasingly complex requests.

Data Anarchy

When communication is flowing between the analyst, the functional manager, and the business intelligence manager work should flow smoothly. Analysts should come together with their analyst peers to form a community while solving problems and becoming subject matter experts.

The risks ultimately come down to communication. When this communication breaks down there is data anarchy. The volume of requests could reach a level where more analysts are hired by existing or new teams. If there is no relationship with the centralised business intelligence team there is the risk of bad practice creeping in. Analysts could be moving data out of approved systems, releasing statistics without approval, or becoming a ‘shadow BI’ function.

Problems will also arise if the organisation’s management or structure doesn’t give analysts what they need. If training isn’t provided or tools are restrictive analysts could build workarounds. At best this may be bad practice, and at worst could pose a security risk. Complying with privacy legislation is more important than ever so getting this right needs to be a priority.

Data Democracy

The goal is to get insights to those who need them quickly but with security and governance in place. This is accomplished through training, tooling, and trust. With the data analyst educating rather than simply being a ‘resource’ or data vending machine.

This is not to say that everyone should have access to everything. Teams must commit to playing by the rules, judging when to share results and when to seek guidance.

Data Citizenship

Get started by creating governed datasets and training power users, or data citizens, in best practices. Teams can then come up with creative solutions while gaining an appreciation for where their data comes from.

On the spot training can help new users through the first few months but it simply doesn’t scale. Training power users to confidently champion the use of data in their teams is much more scalable.

  • Grouping datasets so teams can access just what they need.
  • Create short videos with clear instructions on how to get started and use the datasets.
  • Making use of a data dictionary with a description of what the dataset does and examples using it.
  • Train power users together and update them with new information regularly.
  • Inspiring exploration with data visualisations which can be repurposed.
  • Clear direction and a framework in place on the expectations around data governance responsibilities.

Organisations with a need for insights should consider this shift in culture a priority. It may not be a quick process but the move toward data democracy will be worth it.

Further reading

Achieving data democracy without sliding into anarchy is possible. Breaking down data silos means teams can learn more about their customers, products, and industry with governance in place.

Photo by Maria Orlova from Pexels

Comments are closed, but trackbacks and pingbacks are open.