We talked a lot about technology in the Data Lab and the POC articles, let’s talk about people. When we think about Big Data and artificial intelligence, we think about a data team with a Data Scientist, a Data Engineer, a Data Steward… But let us not forget about Marketing or R&D profiles as the business vision needs to lie at the heart of the project in order to make it succeed.
What Are the Reasons to Involve Business?
Big Data and AI initiatives are driven by business needs. The project aims at addressing business issues, that is what makes it so valuable.
A Data Lab is initiated for two reasons :
- The company understands the stakes when it comes to data and does not want to miss the opportunity
- The company has a business issue and the Data Lab can address it.
In both cases, the business vision needs to drive the data vision, not the other way around. Creating a strong relationship between business lines and the Data Lab is how you make sure the project is relevant and allows you to make it happen even faster. The first step is to identify use cases and KPIs. The reason why the business needs to be involved from day 1 is to avoid the Data Lab from exploring without any guideline, which would be a waste of time.
How to Involve Business in Your Projects?
Business people need to be involved in the decision making process when defining the use cases. An appropriate way to do it is using the Divergence / Convergence:
Divergence: It is about finding as many use cases as you can think of such as internal process improvements, customer satisfaction, customer knowledge or supply chain.
Convergence: Now that you have plenty of use cases ideas, this step consists in defining whether they are adapted and if they can actually be implemented. In order to do so, you will need to assess the company’s data and find out whether your teams have the skill set to manage it.
What you need to know : It is better to focus on few use cases to quickly demonstrate their relevance and profitability. The main risk is when the Data Lab is not able to show results because you addressed too many use cases and teams were way over their heads. It will weaken your teams’ motivation and may even jeopardize the whole project.
Last but not least, here are a few questions to ask yourself before you go head-first into your project:
- Does this use case can create value to my business?
- If it does, do KPIs can be defined to evaluate it?
- Will the POC I am about to have integrate operational business processes?
Do not forget : 95% of a Big Data / AI project is Design Thinking. No need for you to rush and pick a project that will not serve your company’s interests. Once the use cases are defined with the business, you are left with data process, POC and Data Viz and believe us, it does not always go as planned.