open search menutoggle menu
Actualités t3://page?uid=116

Is your company really data driven?

A new generation of self-service business intelligence tools is democratising access to data. However, this appropriation of the business lines must coincide with the implementation of a governance framework to control data quality and help the organisation to mature.

As the new "black gold" of our economy, driving the digital transformation of companies, data is now omnipresent. From floor to ceiling, data infuses organisations. At the same time, we are seeing democratisation of access to data. Until recently, the business divisions had to call on a decision-making expert and wait several days or even weeks before having their reporting request completed.


A new generation of self-service business intelligence (BI) solutions such as Tableau Software or Microsoft’s Power BI has changed the game. Easy to install and take charge of, the business lines quickly adopted these tools to create dynamic reports for themselves that highlight data visualisation.

Self-service also means shadow IT. Some of these solutions in SaaS mode have been deployed under the radar of the ISD, their publishers not hesitating to contact the operational departments directly. This has not been without problems in terms of integration and urbanisation of the information system.

The dissemination of data throughout the organisation also poses a challenge for regulatory compliance with the GDPR. The democratisation of access to data can therefore only go hand in hand with a new distribution of roles between the ISD and the businesses. As business partners, the ISD provides them with tools to make them autonomous while defining the framework for their use.


Data quality control

More generally, the question of data quality arises. The trust placed in data quality is central. When a user publishes a dashboard, they must be sure of the reliability of the results. When is the data reliable and secure? Otherwise, what remediation policy should be implemented to remedy it?

Data preparation involves establishing cleaning, standardisation, classification and categorisation rules to ensure that the data used is complete, integrated, duplicated, consistent and up-to-date.

A governance project is both technological and organisational. First of all, it involves mapping all the data of an organisation to build a data catalogue. This repository lists the data but also provides a clear definitionof shared indicators.

A company’s departments all manage customer data but each has an often different definition of what characterises a customer. For a sales department, this will be a natural or legal person who has placed an order for less than X months. For the ISD, a customer will be limited to a code. We need to arrive at a common terminology.

At the organisational level, roles and responsibilities associated with profiles of data owners and consumers must be defined. Who is responsible for the data and will ensure its quality? Who will manage their development or even their end of life?


Acculturation to data

A data-driven company must also conduct an intensive acculturation policy that will be managed over a medium-term time frame. The aim is to make all employees aware of the value of data, a driver of transformation in their professions. While large accounts are mature in this field, mid-market companies are not yet.

Acculturation to data requires the appointment of reference persons with a strong knowledge of both issues and tools. These data ambassadors represent all functions of the organization, whether logistics, sales or management control.

Still called data champions, they find themselves in data communities that will work on upskilling and developing new use cases. Writing a manifesto defines the purpose of a community. This profession of faith is summed up in one sentence as "enriching our know-how to promote data and imagine its new uses".

These communities will look ahead and monitor the contributions of new technologies. Modern data platforms allow data to be processed in real time or in near future. Artificial intelligence provides a predictive dimension whereas BI above all offers an image of the past.

Another line of work is to help the business lines "talk" about data. Data storytelling uses, for this purpose, the principles of storytelling and a set of graphic conventions and colour codes to explain complex figures in a simple way and identify trends.

As we can see, for managers, setting up a truly data-driven company is the result of an extensive data policy and a set of controlled timeframes.