Project analysis

Let's imagine a case where a company asks for basic and quick information about the possibilities of launching a business intelligence project. The aim is not to be as precise as a complete analysis but rather to provide general information in a technical-commercial approach, taking into account that to answer this question we have only thirty minutes, not one more!

The situation:

The Company, a leading health insurance firm, would like to optimize the use of its data to fulfil several practical needs, for instance:

  1. Reporting of activities and performance: accurate data refreshed on daily, weekly and monthly basis
  2. Listing of clients for marketing campaign execution
  3. Development of predictive and prescriptive modelling to better address questions like acquisition, up-sell/cross-sell, retention
  4. Market studies enriched with external data
  5. Near real time actions: to react immediately based on client's profile and actions

To support these requirements, the Company can count on the following data sources:

  1. An ERP system that captures all contractual and transactional data with the clients
  2. A CRM tool that encompasses all marketing activities information
  3. A third party database that collects instantaneously all the logs coming from the online activities: website, customer portal, client app...
  4. Publicly available data related to socio-demographic indices, and data that has been scraped on competitor websites

Currently, the Company only leverages the use of an ETL tool, some machine learning algorithms and the competences of a few business and data analysts.


Questions, followed by answers:  

1. How would you organize the technical infrastructure?

  1. Keep the existing ERP / CMS / DB External systems as they are.
  2. Add a dedicated server in DataCenter, assuming the insurer has multiple locations.
  3. Engage the services of a BI developer who will be able to give precise advice to the company and free up the time of traditional IT teams.
  4. Creation of a "Data-Lake" in which the data from each system is sent in order to constitute a centralized data source.
  5. List the needs of each department by providing information on the project and asking that everyone can consult their teams in order to define their needs, both globally and by indicating the specific cases, because if the "Data Analysts" are certainly the most precise, the whole company is able to use certain data but they do so in an anarchic way, which has various impacts.
  6. Set up the systems that manage the data (SQL / ETL), define the dataset and make it available via the existing network and ACL.

2. How can you motivate your choice based on the following dimensions?

a. Matching the business needs

Surcharges for additional servers as well as "BI Dev" and third party software, on the other hand, can be offset by a reduction in IT staff costs and an increase in the performance of staff who no longer have to search for information, also adding recuperated sales efficiency that can be improved on a case-by-case basis through better responses provided to customers.

b. Implementation and deployment

The implementation and deployment do not have a negative impact on operations because the system is set up globally in parallel with what exists. At the same time the staff is pre-informed and pre-trained on the upcoming changes, so they can submit their needs and ideas.

c. Maintenance

The functioning of the other systems remains identical at the IT level. The specifications of the " Business Intelligence (BI) " project are managed by the new developer who can also include the IT department in the project in order to have support when needed and increased efficiency from people who are already familiar with the infrastructure and the internal functioning.

d. Future proof

Use of standard and proven technologies that have been on the market for many years and are therefore expected to continue to be common and stable enough to guarantee its use or evolution in the future.

e. Security/GDPR

See more precisely with the legal department in order to get all the information. Other systems that have already been checked are safe. The new infrastructure will also require verification, but it can already be said that the GDPR security will be increased: indeed each department receives access to its own information and in a more secure way than the current open door to everything.

f. Connectivity/flexibility

Centralized provision via "datacenter" allowing a more serene and efficient global management, and above all a standardized provision of information regardless of where the user is located.

g. Budget

  • Additional personnel costs for the developer, compensated by cost reductions for IT, optimization of personnel utilization, improved sales.
  • User licenses to be defined according to the final needs, to be quantified according to a choice "Table" or "Power BI" which can be induced in Office 365, as well as according to the options.