Launching a new BI project or initiative can snowball into a problem, if not planned carefully. It’s important to understand the end result – what your organization hopes to achieve, what are the business objectives of the initiative – and work backwards to design your project. This will ensure that you don’t miss out on key requirements as well as eliminate unnecessary details.
I’ve seen organizations go the other way around – they gather as much data as they can, use BI tools to slice & dice data, calculate tons of KPIs and build dashboards no one will use. This happens because they jump into data analysis & visualization without questioning if these calculations & KPIs are even relevant to your current project.
Every BI project is unique and should begin with the business end of matters, instead of being centered around data & tools.
Here are 4 steps you can use to plan your next BI Project.
1. Clearly identify relevant stakeholders
Stakeholders are the target audience for your BI Projects and it’s important to identify them clearly, when you begin planning a BI initiative. They are the executives, managers and analysts who you will use your BI solution to analyze data and view the various deliverables (reports, dashboards, charts, etc) regularly. So you’ll need to build a solution tailored to their requirements.
In today’s collaborative work environment, information flows across the organization in various ways (as emails, spreadsheets, presentations, etc). So if you implement a BI project without understanding who will benefit from it, you might miss out key requirements and be forced to make changes at a later stage of your project.
For example, you may build a dashboard for a manager but realize that it is being used by your CEO as well, who wants to add a couple more metrics to it. This may force you to go back to the drawing board – where will you get the data for these metrics, how to calculate them, how to present them on your dashboard.
So it’s always useful to create a simple document or spreadsheet with a list of all the people who will be using your BI solution. This helps a long way in defining the project scope and requirements.
2. Interview key stakeholders
Once you’ve identified the key stakeholders, then the best way to understand the business objectives of your BI projects is through face-to-face meetings with relevant stakeholders. If that’s not possible, then phone or skype are the next best options.
Avoid skipping this step or making do with written specifications over email. These few hours of meetings will simplify your job a great deal, save you a lot of time later on and greatly increase the chances of your project being a success.
Here are some of the questions you can ask during the interview:
- Pain points: What problems do you hope to solve with this project? What information are you looking for? Why isn’t it already available or why is it difficult to get this information?
- Current & desired state of affairs: How are decisions made currently? How would you like this to happen in the future? What are the barriers to accomplish your vision of future decision-making?
- Wh-questions: What are the different dashboards needed? Who will use each one of them and why? What needs to be shown in the dashboards? When and where (on which device) will they use it?
You should spend 1-2 days trying to ask the above questions. At the end of this stage, you will have a clear summary of business requirements for your project.
3. Design a data workflow
Once you’ve got a good idea of business requirements, it’s time to draw out the different processes required to accomplish them – transforming data into final deliverables that provide useful insights to the stakeholders. We’re not talking about the actual dashboards and data visualizations but a flowchart diagram that shows the different business processes you need in your BI Project.
This will enable you to break down your BI project into easy-to-understand blocks that can be implemented step-by-step. It will tell you which business processes can be leveraged (to get data, use existing reports & analytics, etc) and which ones need to be built. You will also get a clear idea of the various dependencies among the processes.
At each stage, you’ll understand which input data is needed, what transformations are required, what are the different deliverables (charts, reports, etc) and who are their recipients. By the end of this stage, you’ll know how the organization makes decisions and how it evaluates those decisions. For example, marketing folks are focused about campaign performance, customer support is concerned with customer satisfaction, etc)
A top-down approach works well for this step – you start with high-level KPIs relevant to the senior decision makers – growth and overall performance metrics. The lower metrics should elaborate the high level ones, making it easier for users to get more details about individual KPIs.
This is a pen-and-paper exercise and should be done in a few hours or a day, at most.
4. Create Mockups
At this stage, you create the initial draft of dashboards & reports delivered by your BI project. This is where you’ll need to start going deeper into the implementation details: what data sources are required? where will the data come from? what metrics are readily available? which ones need to be calculated? which charts to use for each kpi? how to arrange them on a dashboard?
You should always get a sign off on the deliverables from the key stakeholders. Sometimes, this step can be quite iterative simply because there are too many decision makers involved or due to indecision. In such cases, it’s a good idea to limit the number of iterations and communicate it to the stakeholders, in order to avoid delaying the project.
Once you’ve created a mockup, you should go back to the stakeholders and get their approval. They’re the ones who’ll be regularly viewing your dashboards, so it should contain all the information that’s useful to them in an easy-to-understand manner. Refine your mockup based on their feedback and finalize the dashboard designs.
Once you have their approval, you’re good to go.
Some BI projects may take a few days, while some may take months. However, it’s important to treat each project differently. Avoid falling into the trap of using preconceived notions from your previous company or project to plan your next project. You can apply these steps to plan your each BI project and ensure it is executed successfully.