Purpose of my work is oriented on preparing the best BI (business intelligence) tools to support achieving goals within the organization. In the last post, I described four initial points to start the KYC process. Today I’m going to focus on next steps. In this post, I’m going to share my experience and my best practices on how to gather requirements effectively.
I have seen many guides about what kind of questions is good to ask during business or data analysis. Most of them are useful and relative to the topic. However, my role doesn’t focus only on business or data analysis. The subject is much wider. Daily, I’m using techniques and methods from UX & UI design (user experience and user interface) and data visualisation to create the state-of-the-art BI product. In here I would like to clarify one thing I’m not a BI Architect, who is responsible for the back-end. My research and tasks focus on overall business needs and front-end outcomes. I create content and the “look and feel” experience.
So, you already gathered as much as possible information about the customer company, industry challenges and trends. You are ready to kick off the project and meet with the customer. And now what should happen? What a plan is? How to find out what needs to be delivered? Where to start?
There is no better way to gather requirements than directly ask your customer. But customer can be a broader audience. In fact, there are plenty of people with different needs and different objectives. It’s necessary to identify those separate groups of users. And at this point, we have reached a first question which should be addressed.
Who is going to use these reports?
The common approach which I encounter is to design One Report for All. In a nutshell, the main idea is to design one report, from one data set and try to bring together all expectations. How does sound for you? Do you think it can be achievable?
Do you know what is told about a compromise? That it is not satisfactory for any of the parties. The same result is when we offer one data view for different viewers. Still, we can have the same data set, but, e.g. Marketing Director will look for other information than Controller. Their fields serve for other purposes in the organization; that’s why they need to have tailored alerts and KPIs (key performance indicators).
Having in mind those varieties my first question to the customer is about how many different end-user groups we are going to have.
What kind of actions/decisions these people are going to make?
This question is my second crucial one to understand how these groups use and consume data. Understanding of actions and decisions giving me the most valuable information about the true purpose of having reports/dashboards in place. Thanks to that, I can design dashboards with higher adoption, because they respond to daily challenges.
Another benefit of this approach is to establish a better and stronger relationship with the customers, who feel that they are heard. What is more, sharing such an attitude we can become a partner, not just the provider.
Of course, those are only opening questions, because just right around the corner, the following questions are waiting:
” What business needs do those groups have? “
“How often they use data?”
“Which channels they use to communicate/consume data?”
“Which devices they would like to consume data on?”
“What is the data literacy level?”
“Is the data culture established within the organization”
“What are the habits and behaviour patterns regarding data usage?”
To collect answers to these questions, the best way is to hold interviews or workshops with representants of each group to learn their perspective.
Tangible outcomes of KYC.
The success of all projects lies in the preparation. KYC process in BI world is nothing more than a sequence of steps leading to create the vision of the final product. At the end of this stage we should have some tangible outcomes:
– well-defined groups of end-users. Those groups are called PERSONAS – fictional characters who represent types of users.
– description of product vision: scope, KPIs and alerts list, UX & UI assumptions and foundations.
Convince Me with Data 