We Need To Talk About Dashboards

We need to talk about dashboards

An Unintended Controversy

A few weeks ago, I suggested to a colleague that I may write a blog on the decline (I might have even said death) of dashboards. Of course, this elicited a degree of shock. My heresy, it seems, strikes at the heart of anyone who can’t envisage life without their business intelligence (BI) dashboards. Let’s face it; it’s how the corporate world has managed its performance for years.

Even in my company, we have many dashboards. The phamax team have created a fair few over the last decade, but their true efficacy is something I’ve been thinking about for some time. In my view, new technologies are pushing hard to make these corporate shibboleths a thing of the past.

But for the time being, I again pose the question. Are dashboards still the answer to your information needs?

It’s not just me

When I talk to healthcare leaders, many express their frustration with how hard it is to get on-the-spot business information to make timely decisions. Having to rely on a team member, SharePoint or a dashboard to access critical data for a report or meeting takes too much time and delays making a decision.

In truth, leaders don’t actually need dashboard data. Instead, they want to know the overall story the data is telling them. How is the business progressing? Do I need to concentrate on specific things? What do I need to say to shareholders? These are the real corporate stories, and in telling them, future progress lies.

A recent 2021 Gartner Analytics & Business Intelligence poll showed that 25% of leaders view the skill of corporate storytelling as one of the most critical when selecting a new analytics solution.

While I would caution that data generated by BI algorithms are not comparable to the narratives humans create. But they can help us find and understand the insights we, as leaders, need to craft our stories.

I’m not sure dashboards do this as well as we might hope. So we need something else, something more.

We’ll talk about this later.

The dawn of a new (Data) age

Let’s start by getting our ducks in a row and defining what a dashboard is. The name derives from the idea that busy corporate players need complex business information laid out efficiently in visualisations like graphs, infographics and charts to promote understanding and get the evidence they need to prompt appropriate responses.

The name ‘dashboard’ derives from a car cockpit’s dials and gauges. As you drive, you have all the information you need to assess the quality of your driving and the vehicle’s health. In addition, the layout makes it easy to read and understand, allowing you to make any required adjustments as you progress to your destination.

It’s the same principle here. Performance data is captured on a bespoke digital dashboard and delivered quickly and simply to users.

Today BI dashboards tend to fall into four broad categories:

Strategic BI dashboards help senior management to get an overall view of how the business strategy is performing and derive new plans for the future.

Operational BI dashboards concentrate on real-time metrics such as website analytics, marketing response rates or call centre performance. These dashboards allow operational staff to make situational decisions.

Analytical BI dashboards tend to be more complex, using many data types to perform various corporate analytical functions.

Tactical BI dashboards blend operational and strategic data and indicate how functional teams contribute to strategic aims.

Naturally, many sub-variants like ’emergency’, ‘new-because-we-don’t-like-the-original’ or ‘a rosier outlook’ supplement these definitions too.

Tech to the rescue

As access to computing technology expanded during the 1990s, the use and complexity of digitised BI dashboards, allied with advances in data visualisation, transformed these tools into the ubiquitous management tool we know today.

Microsoft’s Digital Nervous System was an early example of a tech-led approach to corporate performance data. From there, automated BI dashboards spread throughout the corporate sector, driven by academics that honed the concept by introducing new-fangled ideas like Balanced Scorecards, RAG reports, KPIs, heat maps and bubble charts.

Dashboards soon became the currency of executive oversight and, for converts, were seen as a panacea—an all-seeing eye into the corporate heart.

Now we know everything, we thought.

Even as I write, dashboards still sound utilitarian. As a technologist and business leader, I can see precisely why dashboards have become so popular. Today they offer dizzying degrees of sophistication since their inception in the early 1970s. And, on the face of it, they seem to solve many informational dilemmas in an automated easy-to-access format.

But that’s not the end of the story. For all their understandable appeal, BI dashboards are not infallible Delphic oracles.

If you build it…

Will they come? Not necessarily. Building dashboards can be a lengthy process. Scoping and prioritising conflicting user requirements is only the beginning. The build takes weeks as datasets are sourced and reformatted, developers scratch their heads over a raft of formulae, and filters are hastily cobbled together. And after all that, user testing goes over schedule after discovering many problems.

Even after launch, corporate dashboard use remains limited. Despite everything, no one knows how to filter their data from the dashboard, and some find that what they need is no longer available, so the requests for further dashboards begin.

Marketing needs a dashboard for this. The Clinical Team requires a new dashboard for that. Drop everything! The MD needs a strategy update dashboard for next week’s board meeting. Soon enough, different iterations circulate ad infinitum through the corporate structure feeding data-hungry appetites wherever they go, but satisfying few.

As an analogy, we can liken dashboards to swans: seemingly elegant as they glide serenely over the water but with much frantic thrashing below the waterline to keep things afloat.

The definition of truth

I must also mention some conceptual issues with dashboards that worry me. I recently read this excellent article by Angela Meharg on LinkedIn and found that I agreed with much of her thinking.

Meharg notes several fundamental issues: ‘you don’t know what to measure in your enterprise.’ You might think this is an obvious point, but I wonder how many dashboard builds set off without a clear idea of what they are measuring and its use for effective decision-making.

Meharg also thinks many dashboards concentrate on the wrong metrics focusing on outcomes (effectively the past) instead of ‘measuring the actions you have pre-determined ought to produce those results.’ The eternal battle between lagging and leading indicators…

I also found this helpful summary in the Harvard Business Review by Joel Shapiro. In his view, there’s a tendency to overstate the capabilities of dashboards to include the ability to make accurate predictions of the future and help to shape ongoing business strategy. He eloquently states:

“Moving from description to prediction to action requires knowledge of how the underlying data was generated, a deep understanding of the business context, and exceptional critical thinking skills on the part of the user to understand what the data does (and doesn’t) mean. Dashboards don’t provide any of this. Worse, the allure of the dashboard, that feeling that all the answers are there in real time, can be harmful. The simplicity and elegance can tempt managers to forget about the all-important nuances of data-driven decision making.”

The message is clear, despite their alluring utility, we must use dashboards carefully and critically to ensure they help us in our efforts to develop our organisations and the stories we want to tell.

Plus Ca change

Despite all the above, dashboards remain an essential part of corporate analytics. But rather than being a quick-fix solution, their value depends on their quality, how easy they are to use, and where a company is on its data journey.

For example, a specific website tracking dashboard has much to offer to marketers grappling to make sense of how effective their efforts are. But a more dynamic approach might be recommended when setting a high-stakes corporate strategy.

Maybe it’s time to look at dashboards as a contributory part of new data discovery technologies. Ones that give a clearer idea of what corporate data is saying so we can make informed decisions before taking action.

On life support?

So let’s return to my theme: the sad, slow decline of the BI dashboard.

Maybe I’m being a bit preemptive because, for all their faults, dashboards remain a valuable way to assess performance. This is especially true when used by experienced professionals with great instincts, natural curiosity, and the ability to look further into what the data is saying so they can craft stories appropriate to the business circumstances.

Happily, there are many such individuals in modern commerce.

And this is where advances in AI and conversational digital assistants, like Ariya, will help breathe new life into BI dashboards, enhancing their utility to everyone in the business.

This process will advance in stages as AI technology develops and embeds into tech ecosystems and leveraging existing data infrastructure and toolsets.

The first stage is adopting a well-integrated AI-powered conversational layer into an existing BI dashboard suite. Preferably a domain-trained industry-specific product like Ariya designed to work within a specific business context (in Ariya’s case, it’s healthcare).

Now users can find BI data using an intuitive UI on their phones, tablets and PCs. This becomes additionally powerful for team and project work when the AI is installed in group collaboration tools like MS Teams, Sharepoint or Slack. For example, a team can ask for supplementary snippets of info to address their ongoing queries and concerns during their discussions.

Information access and use become increasingly fluid as users freely explore corporate data repositories to find new ideas and connections and share them seamlessly with colleagues and stakeholders. Whether in business meetings or client interactions, the data is always there.

The next stage is where digital assistants become trusted everyday partners, simplifying how we work with data and business information. The reliance on complex static BI dashboards wanes. Users now use fewer dashboards but with increasing speed and accuracy to extract, filter and present disconnected corporate (and none corporate) data to achieve their objectives.

On this theme, the phamax team is now exploring use cases that integrate Ariya into existing BI tools offering a dynamic user experience where the dashboard plays an increasingly minimal role.

Instead, a domain-trained algorithm proactively directs BI data straight to subject matter experts based on their habitual data needs. Imagine the convenience of having all the metrics you need to meet your objectives delivered daily to your inbox by Ariya – no effort required.

How much more productive would you become as a result?

And now the end is near?

Readers know how times have changed over recent years and have challenged how and where we work. Likewise, the advances in AI mean how we use dashboards will also change. To my mind, this is a growing truth.

But as we have seen, there’s still a place for at least some dashboards.

In effect, we need to cut the number of dashboards, increase their quality, and amplify the ease of use of what remains. The key is to create as few as needed and strengthen their utility by using new access tools, like Ariya, to manage the interactions efficiently and creatively.

The phamax team is working on a dashboard-lite approach to BI. Using Ariya, our team has made it easy to create a set of predefined user-specific dashboards that require zero maintenance and operate using Ariya’s sophisticated conversational layer.

So we can conclude, at least for the time being, (some) dashboards will live to fight another day. Indeed with AI tools, they have the potential to perform better than ever.

Until…

A visit to the museum

Finally, let’s cast our minds forward for a moment. A couple of colleagues from a future company are visiting the Museum Of Old School Business Practice. They stop to examine the BI dashboard exhibit, studying how it worked. They look at each other, and simultaneously, they both say:

‘Wow, working with data was difficult back then!”

And yes, by then, the BI dashboard will have sadly left us for good.

Contact the phamax team by email or via the contact section to discuss how our conversational digital assistant Ariya can transform how you work with BI data using cutting-edge AI.

References:

  • https://www.linkedin.com/pulse/6-reasons-dashboards-dont-work-what-do-angelameharg
  • https://hbr.org/2017/01/3-ways-data-dashboards-can-mislead-you
  • https://www.forbes.com/sites/brentdykes/2018/10/30/the-real-reason-most-dashboardsdont-tell-data-stories/?sh=19877bae1abb
  • https://towardsdatascience.com/dashboards-are-dead-b9f12eeb2ad2

Get free practical tips, case studies and quotes for the pharma industry

for inquiries write to us ariya@phamax.ch

ariya-sales