Changing your database structure

Data, a resource to be exploited

When it comes to data analysis, enterprises still have some work to do. According to the barometer of the digital maturity of SMEs performed in February 2019*, only 36% of the French SMEs claim that they have a consolidated database. A few of them work with data scientists or data analysts (23%). This means that there is an important margin of progression to capitalize on the value of data!

“The impulse must come from the Management for the data to be part of the company’s culture”, says Clément Levallois, professor at Emlyon Business School and data specialist. This dynamic requires training all the employees on the value of data (as this may not be obvious for everyone!) and putting proper governance in place. “Ideally data owners should be systematically designated, this means operating advisers capable of identifying the data to build on and the data to acquire first, in order to develop the customer relation.”  A complimentary business vision to the technical skills of a data scientist.

How to consolidate the available information

Building a reliable database consists of putting together different types of information from different sources. It is a matter of consolidating the essential information (customer name, company name, address, phone number), the information related to the communication history (calls, e-mails, or chats), and the useful business information for the sales and marketing teams (preferred products, detailed customer information to provide the 360°  view).

Database quality is the most important factor in successful customer relationship management. It should follow simple logical rules (such as avoiding duplicate data) and meet performance objectives: “for example, the repository information system must allow easy access to the data in order for it to be easily indexable and fully searchable” underlines Clément Levallois.

With the implementation of the General Data Protection Regulation (GDPR) in May 2018, the management of personal data by companies has been reinforced: “ It is not a question of creating a new repository but of defining more accurately the data governance. The data to be kept and the data to be cleared, should the enterprise fail to collect the consent of the concerned persons for the use of the data”. In nine months, some thirty companies have already seen their data policy challenged by the CNIL (National Commission for Computing and Liberties).

For any type of processed data, special attention must be paid to the information required in the necessary fields to minimize the number of incomplete accounts, as they may not be found through a search. For example, if a search is done to find communications among targeted sales managers in the agri-food industry in the Ile de France region, then the results will only include the accounts that meet these criteria. If only two out of the three search criteria are met, it is possible to miss a business opportunity. Taking the time to qualify the information upon creating the database is essential.

Towards a smart data management

With the development of machine learning algorithms, artificial intelligence opens up new possibilities for data management: “Data has always been a support for statistics, but machine learning allows us to push data analysis further, by refining predictions, for example, says Clément Levallois. With machine learning, it is now possible to analyze what is referred to as non-structured data such as long and complex text, which was not possible before.”

Practically, the machine classifies information based on a set of pre-defined categories. It makes data more accessible and usable in an automated way, an additional asset that increases the potential of the customer database.

* Study conducted by EY and the Apax Partners fund