Get Profitable Insights from Your Analytical Dashboards
Business intelligence dashboards make an impressive appearance at presentations. Their real value results from the insights you gain and the actions you and your team take based on those insights.
There are four key steps you should take to gain the most profitable insights from your investment in the data mining and data visualization that go into your analytical dashboards. These are steps you should take when you are planning and implementing your dashboards.
- Choose an experienced provider
- Identify the key metrics that are critical to your business success
- Define dashboard prescriptives to guide your team
- Nominate a data steward to guide your provider through your data stores
Choose an Experience Provider
If you have an existing team of data scientists, data architects and analysts, and business intelligence dashboard designers, you may have the team you need to build your business intelligence dashboards in house. This presumes that all the team members are free from other projects and available for duty. It also presumes that they are up to speed on the tools needed to collect and transform the data, maintain an operational data store of aggregated data from you ERP and CRM systems, and skilled working with dashboard tools needed to design and build your analytical dashboards.
You’ll usually enjoy a shorter delivery schedule and better payback by engaging a skilled team of business intelligence specialists from outside your organization (the dashboard team). Such teams will have already worked dozens of projects like yours, and will have the expertise needed to guide you through the process. Experienced teams will more quickly put the correct data, cleaned and transformed, at your fingertips so that you can answer critical questions and guide your organization in the right direction at the right time.
Identify Key Metrics
Why Reports Fail
How many cute reports and dashboards have you seen that are basically useless? Great colors and design don’t always give you clear answers to the questions you care about. In some cases, you may be getting your most helpful information from some rather unattractive tables and reports, or assembling your own answers in Excel. Such reporting shortfalls result from:
- use of reports that are based on a single data object or table, with data from other objects or tables merged in
- use of tools that cannot integrate data from multiple ERP, CRM, or billing systems into a single, coherent, display
- reliance on canned reports that aren’t built from the ground up to answer your specific critical business questions
Selecting KPIs for Success
To get the greatest business value from your business intelligence dashboards, you have to first conduct a thorough and honest analysis of what you really need to know to manage your company. You have to separate fluff and noise from each critical key performance indicator (KPI). Answers to questions like these below will help you outline your analytical dashboard requirements.
- Which numbers am I accountable for, and which go to the bottom line?
- What information do I really need to see to make decisions and manage my business?
- What do I need to see to spot problems and to find valuable opportunities?
- What key metrics do I want to highlight to help focus my team?
Perform a sensitivity analysis on your data to help decide which numbers you need to see to reach the KPI goals for which you are responsible. Is it costs, sales, gross margin, days sales outstanding, backlog, close rate on deals, average deal size? Doing this as an exercise in Excel will help you select the most important information. The outside consulting team that will deliver your dashboards should also be able to help you perform this exercise in algebra to put the right information into your new reports and dashboards. That team will also be able to walk with you through existing reports to identify the helpful parts that you need to expand and carry forward.
KPI Metrics Provide Leadership
The key performance indicators that you choose will inform your team about what is important and provide guidance for your team. Your team will work to the key performance indicators that you select, and this will influence the performance of your team and the results you get. In other words, if you focus on sales and gross margin, then the sales and marketing team
Provide a Guide for Data Discovery
Your analytics dashboard team will perform an in-depth discovery of data sources and uses so that they can identify
- authoritative data sources and data owners for each data element,
- subject matter experts who understand the semantics of your data, and
- report and dashboard users who rely on the data to run your business
You or someone you assign will serve as a data steward will help connect the dashboard team with these data sources, subject matter experts, and users. The data steward should be able to work with IT to arrange connections and credentials for secure connections to existing systems.
The dashboard team, with the help of your data steward, will develop a data dictionary, identify the system of record for each data element, and choose those data elements that are needed to correctly provision your new business intelligence dashboards.
Nearly all business intelligence dashboards provide statistical and historical data. Many provide forecasting tools that allow prediction based on current trends and current status.
A newer class of business intelligence dashboards, served by business rules and machine learning algorithms, spotlight less-than-obvious trends and make actionable recommendations based on the data and predefined business rules.
Reports and dashboards with this capability can not only highlight red, yellow, and green conditions, but can also recommend a course of action in the case of red light conditions, for example. This can help guide you and your staff as well when an alert condition arises.
In the portion of an example dashboard shown here, an alert condition is shown for a group of customer accounts, along with a selection of five actions prescribed to investigate and resolve the issue.
Prescriptives can be defined by combining trigger events with resolution steps, as shown in the example.
We have hundreds of these dashboards in our “day one ready” portfolio set by your role or by your industry to get you to Maximum Profit in the Shortest Time.
About the Author
Bruce Brownlee is a data scientist and product manager with an undergraduate math and engineering degree and an operations research graduate degree with a focus on stochastic processes and mathematical optimization. He’s worked as an industrial engineer, software developer, marketing director, data scientist, and product manager across a broad range of industries where analytical skills are required. His special interests include financial modeling and machine learning applications.
Product Manager at Blackstone+Cullen