There have been drastic changes to how companies perceive data over the past few years. The sheer volume of data that’s available from multiple applications, the speed that it gets processed, and the variety of data that is available for consumption, have combined to create an overwhelming scenario for most companies. At the same time, they have an additional challenge: to make the data accessible, meaningful, and digestible to their decision-makers.
Before 2018, Business Intelligence (BI) was an optional software instrument for many organizations. This is not the case today. Now, organizations understand the value of business intelligence, be it for competitive market studies, internal process operations, efficiency of business or to present data to stakeholders in a consumable format.
Most of today’s organizations recognize the value of data, focusing on building a data-driven culture and developing data-driven strategies in their business projections, as reported by BARC. The research firm conducts annual surveys of more than 2,000 executives in a variety of industries and sectors in order to gain an understanding of the most important or most discussed trends in business. In their 2020 survey, data discovery— the process of collecting company data from multiple sources and consolidating it into a single source to be accessed—received a score of 6.9 out of 10 in importance to business leaders. The #1 spot, with a score of 7.3, went to master data management, which is related to data visualizations and data modelling BI space.
At ClicData, we often engage with clients who are either in the process of adopting BI or want to expand their BI adoption strategy within other departments. As we consult with organizations, we invariably find a gap in people’s understanding of standard procedures and the optimal framework to approach a BI/data visualization project within their organization. We find that invoking data-driven decision-making with a structured approach to building a BI dashboard to serve the organization’s information needs is the most effective way to work.
We use a six-block DataViz framework:
Organizations often wait until the last stage of BI project development to consult with their audiences. However, we find it most effective to think about your audience first as they are ultimately your users. In our framework, in fact, all elements revolve around the end user. This kind of bottom-up approach allows the project manager, as they collect data about project requirements, to keep in mind the decision-makers who will be using the dashboard regularly for their daily/weekly operations, and therefore to develop a solution that responds well to their needs.
The first step to understanding your audience is to identify major categories (roles) for them and to develop a description of the activities they perform. This information will eventually allow you to help them define objectives, intended actions, and KPIs for their dashboard. The next step is to categorize each audience type as either internal or external users. If they are external users, you might want to give your dashboard additional filters or styling requirements that match your external client’s perception of the company. Finally, after categorizing your audience, find out the frequency of access and get a brief description of the expected decisions your audience will be making based on the dashboard. Here’s a sample audience collection sheet:
Sample Audience Collection Sheet:
After defining your audience types and gathering a little information about the decisions they are trying to make, begin to define a coherent objective for the project. One of the most common issues we see with project managers is their non-stickiness to the objective of the project. As they engage in conversations with data teams, end-users, and executives, it’s understandably easy to get overwhelmed with the information gathered and the myriad possibilities of what information the dashboard might present. Objectives can be a lot like Pandora’s Boxes; if they aren’t clarified and addressed appropriately, they can cause much more confusion down the line—not to mention the impact of increased budgets, unmet timelines, or, even worse, the completion of a project that no longer fulfills the needs of the business.
According to Barney and Griffin’s “Management of Organizations,” objectives deliver four key benefits:
- Providing clear guidance and direction
- Facilitating planning
- Motivating and inspiring
- Evaluation and control
We recommend writing down your project objectives for all to see and refer to as needed.
3. Intended actions
Not to be mistaken for objectives, intended actions are the specific actions you want your end-users to take based on the information the dashboard provides them. They also form the basis for KPI discussions and help refine the means and purpose of your dashboard for your audience. Intended actions are chosen to provide highest output and to prompt an immediate response. Here are some of our recommendations for defining your intended actions:
- Don’t assume:
Don’t assume that people will know what actions they can or should or even want to take from getting information from their dashboard. Talk with them, document expectations, refine your solutions and effectively communicate what actions to take.
- You don’t need it if there is no action objective:
As mentioned, intended actions will help you define what KPIs to make visible on your dashboard. Every KPI should have a corresponding intended action. Any KPI on the dashboard that doesn’t have a corresponding follow-up intended action, serves no purpose. Your sales team, for example, should be able to take action on how they can perform better based on the KPIs.
- Be reasonable
While discussing and gathering a list of the actions you expect your end-users to take, be sure to verify that they can indeed perform them. Do they have the power and authority to take those actions? Are those actions controlled by them or someone else? How quickly can they take action?
Here’s a sample Intended Action Collection Sheet:
4. Data sources
The first three steps of the DataViz framework are foundational building blocks that support the development of the rest of the project. The last three blocks are concerned with planning for and gathering the technical requirements of the project and executing the final dashboard.
Today’s organizations are complex and so is their data usage; it’s rare for an organization to rely on a single application to do all the work for them anymore. To stay competitive, to be agile, and to adapt to changing norms, organizations must adopt a multitude of applications and integrate them into their daily business organization.
For example, a manufacturing unit that’s running a simple process to produce a single type of zippers might use the following applications:
- CRM application to collect customer information
- LMS tool to manage internal operations and processes
- HR management tool to manage HR-related operations
- Sales procurement/ERP tool to manage production efficiency and inventory
- Supplier relationship tool to manage suppliers and distributors
- SalesForce tool to manage deals, calls, and other sales activities
- Document management tool to manage purchase orders and proposals
- Financial tools to manage company accounts, receivables, etc.
- Digital marketing channels, such as Google Ads, Facebook Ads, and LinkedIn Ads to generate online ad presence
An organization’s use of disparate data sources adds to a project manager’s challenge of implementing a dashboard solution. Project managers need to have a macro view of all the systems being used, identify which applications are most likely to provide the details required about the objective and intended actions set for the project. Along with understanding the data sources, project managers should also notice if there is a need to create a data warehouse and if there is a need to standardize or normalize the data for more seamless utility.
5. Building Key Performance Indicators (KPIs)
The terms “facts,” “measures,” “dimensions,” “goals,” “objectives,” and “KPIs” are sometimes used interchangeably, but they are not the same thing. A KPI is a metric (which can be visualized with an indicator) that can help you determine whether or not you are on track to achieve a goal. It’s often said that KPIs need to be SMART (Specific, Measurable, Attainable, Relevant and Time-Bound); however, in our DataViz framework, we are least concerned about SMART KPIs as we have already identified intended actions. If any KPI doesn’t directly relate to an intended action you need to take, then it is not necessary.
At ClicData, as we consult with organizations, we usually focus on two types of KPIs, simple and derived. Read Telmo Silva’s blog about the differences between simple measures and derived measures. In brief, some numbers can be displayed as absolute aggregations, such as Total Revenue, Total Budget, and Number of Calls. These are simple measures. But when you compare regional sales revenues with another region that is smaller or has a different number of field reps, you realize the two aren’t comparable. That’s where derived measures come into play. Derived measures allow comparisons of ratios such as Sales % Achievement —i.e. Total Sales / Total Budget—and provide a better and more informative story about what’s happening in that particular region.
Here’s a KPI Collection sheet you can use internally:
6. Mock your dashboard
Once you have defined all the requirements and have gathered the information, it is tempting to get right into dashboard building. But doing so too early can lead to disappointment; your audience might still not be happy with certain aspects of what you build. With our DataViz Model, we recommend creating a mock-up of your dashboard as you would do for any other project. MVPs are always a good starting point for discussion to ensure that end-users are aligned with your deliverables and that you get a soft approval before investing time and resources on building a dashboard.
You can use virtually any method to create your mock dashboard—sketch it on a paper, whiteboard it, or use online tools such as Canva, Balsimiq, or resort to traditional approaches such as a PowerPoint and Keynote.
Here’s a sample mock for Company ABC, Inc.:
Undertaking a data visualization project within an organization has many layers, requires that you communicate with multiple groups and that you ask the right questions. Going about it without a structure or process will likely set you back in budget, time, and resources. Our Six-Step DataViz Process will help you approach any data visualization process in your organization in an efficient manner and achieve the result you are looking to achieve.