Grijzegrubben 17 6361 GK Nuth, Netherlands | +31621146995

Business Analytics


Yes, I know. This picture is wrong.
Gathering (big) data is a good idea but is has nothing to do with Business analytics.
Most companies don’t know how to start Business analytics.
They'll approach it the traditional way and write business cases.
“We have to know what the results are, then we know if it’s worth investing”.

For Business analytics you can’t predict the end results. Sometimes it takes you weeks to figure out that your model doesn’t apply and you have to start all over again. Sometimes your model produces results in a matter of hours and you discover a new way of doing business. I can only promise you that it’s worth the investment.
Companies also need a new way of judging employees. It's a culture chock.

The base for all Business Analytics is your Business Intelligence. You need an Agile Data Warehouse as a solid base for Business Analytics. A Data lake is an excellent place to park unstructured data or 'you don't know yet how to use' data. An even better idea is to park data in your agile data warehouse, integrate it into your reports and then decide if you want to load it into your data warehouse or data lake.

Predictive analysis is totally different from presenting data in cubes and you need a different type of employee and toolset.
You (preferably) need a Data Warehouse and a Data lake that enables time traveling.
You need tools that show you to the most influencing attributes in your models.

Depending on your current environment I can point you to alternative tooling and help you to attract the right kind of employees.
Remark: You need 95% of your employees working in BI to enable 5% of your employees to perform Data Analytics.

Now, how did I start Business analytics in a company that was willing but didn’t know how to start.

  • Gather data.
    We just parked a lot of (different) data into a Data lake (Hadoop).

  • Organise a Hackathon.
    I sponsored two days in an Hotel for my team, Business users and Tool experts.
    We just looked and combined data, used tools (like ASTER NPATH) and found 9 interesting use cases.

  • Implement selected use cases.
    We productionised 3 use cases in 6 weeks including compliancy with privacy regulation.

  • Be aware of new business requirements.
    Business users had participated in our Hackathon and now frequently came up with new ideas for new use cases.
    We appointed a ‘data scientist’ to support.

  • Organise Hackathon’s regular.
    Always finish a Hackathon with lots of drinks and socializing. Don't forget to plan a next one.

You need a lot of political influence to succeed, but it's possible. Even without a lot of budget.