Designing a Business Dashboard like an Airbus Engineer

On all accounts, Richard De Crespigny documentary of Qantas QF32 was a testimony of his skills, leadership and the brilliant minds at Airbus coming together, thus making it possible for Richard to fly an almost crippled plane to safety. The cockpit system, provided the information that allowed Richard and his team to take appropriate actions to prepare the colossal A380 back to Singapore 110 minutes after engine No. 2 exploded 5 minutes after taking off from Singapore Changi Airport.

While Richard’s 35 years of experience is crucial, we cannot discount that the feedback system was key in his decision-making process. It was from this very metaphor that the business world has borrowed and created the concept of a business dashboard that allows business leaders insight into running the business.

While we don’t deal with love and death matter on a day to day basis, why do we struggle to design a business dashboard that makes sense to leaders?

A Bird’s Eye View

Since most of us are not pilots or airmen, let’s look at an example closer to the ground. We take our car dashboards for granted. It’s been around for a long time and has only been subtly changed over the years. What began various knobs and dials, then progressed over the years to LED and LCD displays that are designed to help a driver navigate the vehicles safely from the start to the end of his destination. They provide nothing more than enough for a driver to commandeer his vehicle safely. All the incremental efforts that has taken place over the years were geared at making the information on dashboard easy to see and understand. Dial size, illumination, position, colours. Every aspect has been revisited over many, many times.

It takes several generation of engineers to refine this; to put it mildly, a dashboard that is designed to help executive drive a business to success, will require much more effort.

Defining BI Dashboards

Business Intelligence dashboard has been around for a while now. It promises greater insight for business executives and timely alerts for operational staff. Yet, many organisations have not fully reaped the benefits that dashboards promise. Given that it takes a considerable investment in time and money to embark on a dashboard project, many organisation are lukewarm to an idea that all agree is great. How do an organisation embark on this? I think to begin, we need to look at the fundamental, and understand what are the various dashboards and what are the elements that need to go into it.

One of the biggest gripes raised by the many organisations is that BI Dashboards have not given them much more than what their reports have previously provided. Our assessment is that this assertion is fair. The dashboards have not given more because they have simply been a replica of reports that were used previously that has been moved to a new, albeit sexier, platform.

To gain more from dashboards, we should discard operational reports out of the window as a template, and begin with the minds of key executives as the blueprint:

Beginning with the End in Mind

Many companies have a set of drivers and circumstances that affect them uniquely. It will be a fallacy to think there exist a one-fit-all measures that can be applied to everyone. Identifying and highlighting these drivers and subsequently defining KPIs will lay a foundation on how dashboards should be designed.

The intent of tracking issues and outcome is  to identify areas that require attention so that an objective is met, choosing them are subjective. Subjective to an extend the men and women-in-charge of the believe it matters to their business. As such, these measures are likely to change when a new business head or leader of a company are appointed: how they make decisions are products of their own experience and everyone bring something different to the table. Their thought processes are reflective on the kind of metric being measured and displayed on dashboards.

The User

Once a metric is defined, we need to establish whom and what the dashboard are intended for. For a simple classification, we would call them, Strategic, Tactical and Operational dashboards. Some would argue that there is a fourth type known as Predictive Dashboards.

Strategic dashboards are targeted toward senior level executives who would be looking at the big picture. These usually relate to the results that each department are pursuing. What our sales trending, which market/segment is growing, what’s our market share, etc.

Next, we would be what we term as Tactical Dashboards. The tactical dashboards are the most interesting. These provide the necessary information for manager to take action and keep the dynamic equilibrium necessary so an organisation can keep its course and arrive its intended destination. Actionable information is important. Essentially, they bring all the elements together that support the results.

For example, in the case of a sales department, detailed ratings and reports are indispensable for making decisions. Therefore, thanks to tactical dashboards we might see not only the totals (which is common with strategic dashboards), but also information about a given product’s sale. We might check the value of company shoes sale in Paris on January, for instance. It’s all about the details. With tactical dashboards, managers might also decide to launch each product on market on a given date.

The last category of dashboards is operational dashboard. They deal with daily activities that employees within a company must handle. For a customer support call centre, it could relate to the number of tickets, they closed in a day, how long they take to close it, how many tickets are still pending etc.

Most dashboard implementations do not go beyond operation dashboard.

Putting the elements together for an effective dashboard

One of the most important features of a dashboard that that is often overlooked is data discovery. This more than anything else lends the greatest insight to on the cogs and wheels that keep the business chugging along.

Data discovery refers to how data are associated to one another. This is often done visually, by checking trends, overlays and scatter graph forth. It is this visual identification would probably lead a business to consult its data scientist to validate that relationship statistically.

The Elements

Putting it all together can start with a simple storyboard and perhaps even in Microsoft Excel. However, what make the magic happen are interactivity that enables a business user to “slide-and-dice the data. Our choice of BI dashboard, Yellowfin Business Dashboard, allows our client to do this and more.

When we approach a client’s project the following are some fundamental that we look at

• Compare and Contrast – For lack of better word, I would say everything has a cost. When we look at sales numbers for example, it only has meaning when COGS are taken into account. When we look at sales conversion, we need to know what’s the number if the funnel. When we check the number of new subscribers, we need to know how many has terminated or gone out of contracts.
• Similar Baseline – To make everything comparable is key. For example, a carton of milk in 12 packs cannot be compared to 12 packs of milk sold separately. This baselining exercise can be painful but is required.
• Time period – The practice of using Quarterly results are common in large corporation. However, it may not be relevant to organisations who business is truly cyclical. A shorter timeframe may be a good approach so that patterns can be identified.
• Drill-down – Can data into logical grouping? Are these by sales territory, customer industry or product family?
• Data Banding – Banding is a way to group and classify data together. One common approach of doing this is by undergoing a process of customer segmentation. We recommend that segmentation exercises are carried out. This may involve statistical modelling so that customers are classified based on their behaviour rather than demographic traits.
• Mapping and GIS – To draw more meaning, we also seek the opportunity to see if data can be mapped onto a map. For example, in a sales tracking, we also recommend that marketing activities and investment are mapped accordingly. This will allow businesses to evaluate the relationship between marketing dollars and sales conversion.

Rounding It Up

While we grapple with the answers, we really need to start by asking who our dashboards are intended for. Drop us an email so that we can tackle the issue with you together. We can be your best partners making your business decisions better.