What if you have many different types of data, but don’t use it enough to stay ahead of the competition and improve your processes? What if you have various companies, each with their own data structure and software systems, and within those companies you also work with different software systems? How are you going to restructure and combine this unstructured data without spending months and hiring expensive external parties?
The solution: accommodate different software systems in one data warehouse. The software systems to be combined often contain various data from customers and processes and are not approached uniformly. For example, the first conversations with a customer of ours showed that in some reports more customers were reported than cars were actually sold.
Inventory
Often a company does not have a central data warehouse or data lake, but there are various software systems with valuable customer and process data. For example, our customer worked with three different CRM systems, two ERP systems, specially tailored rental software, lease software (both B2B and B2C), the Email Service Provider, RDW and Autotelex. So a whole bunch of systems, with a lot of valuable data.
At first glance it doesn’t seem to fit together at all, but if you look at the structures behind the data, you see how these tables and data can be linked to each other.
Integration and creation golden record
It is important that the data is made uniform. You then have to determine with the customer which business rules are needed to ensure that the data can be accommodated in one system. You then look at a combination of different factors, depending on the reliability. In the first iteration, you mainly look at company name and Chamber of Commerce number. This combination is then expanded further, for example with telephone number and zip code. By combining the different systems with each other, an increasingly complete picture of the customer is created, and which products this customer has purchased. By bringing together the data from different systems you get:
- A complete picture of each customer; including purchased products, customer profile and customer value;Â
- A reliable picture of the actual number of unique customers; so that reports could be set up reliably and uniformly;Â
- An overview of the pollution and poor data structure in the various source systems.Â
Dashboarding and reporting
For the first time, our customer was able to perform reliable analyses again, and the organization was able to share the analysis with its customers, so that the My Environment went live a year earlier than originally expected. The dashboards that provide real-time information to the management provide more control over the organization. Even the accountants, including the external ones, currently use the data from the qualitative data analysis software. Thanks to the close collaboration between the data experts and the business, dashboards were created that are reliable and functional.Â
Marketing automation
While setting up the dashboards, the customer journey was also worked on. Because a complete picture of the customer had been created, one-to-one campaigns could finally be sent. By connecting different data points, you can fully automate this process. If you also measure and provide insight into the results of the e-mails, you can quickly decide to improve, expand or disable certain e-mails.
Another interesting example is retention. By informing lease customers in a timely manner about the expiration of their contract, the percentage of customers opting for a new contract with the same organization increased by 18%. Of course, a pure A-B test was used for this.
Machine Learning & AI
Our expertise was also requested in the next step in the data maturity of this company. First of all, the question was whether we could contribute to a way to increase sales effectiveness. A prediction model has been created for this that gives a score to leads. This score automatically determines whether a lead is called by sales or followed up by an e-mail journey. After a few iterations, this process had become so successful that 70% of the FTE on sales could be used differently, because the program knew exactly who would convert and who would not. After all, why would you follow up on leads that don’t convert anyway?
The second question was whether we could make an automated offer to the lease drivers who had indicated that they no longer wished to lease from the company. An integration with the Autotelex database makes it possible to make people an offer to take over the car just before they hand in their car.
The result: yourself at the controls
The final phase involved handing over the activities to the organisation, because they wanted to be able to take up the regular activities themselves. The customer insight tool is designed to easily perform analyses, send campaigns and provide dashboards with the correct information. As a result, this company is currently able to carry out the desired activities itself and to expand with even better processes. Â
