Person checking reporting numbers on paper and laptop.

So close and yet so far. When information “at your fingertips” is still out of reach.

Jun 30, 2022

You’ve probably seen the marketing headlines:

Intelligence at your fingertips!

Smart dashboards take the complexity out of analytics.

Intelligent analytics that know exactly what you need!

Wouldn’t it be wonderful? The computer screen reads your mind and produces exactly what you were wondering about.

Since the advent of web-based statistical packages (and yes, I learned computer programming in the days of punch cards, so I can look back that far), vendors have been advertising painless, accessible analytic results. All the answers! Like magic!

Let me burst your bubble. No system will magically make your fingertips any smarter. 

Now, in case you think I am criticizing other professionals, let me reveal that I have been on both sides of these efforts.

I spent almost two years co-designing (what we thought were) simple-yet-comprehensive dashboards that allowed human resource and benefits leaders to dig into their data trends.

Our customers, those same leaders, had lamented frequently that if only they could manipulate the data themselves, it would help them be more informed about and responsive to the inner workings of their organizations.

Our development team agonized over language, structure, interface, and content to make our system as useable and compelling as possible.

This, we felt, would revolutionize how practitioners used information!

They could become familiar with patterns in their metrics, use the dashboard for their day-to-day management, and free up highly trained analysts to work on unique, complex issues.

We couldn’t wait to turn it on and watch the flood of queries begin.

On that first day, most of the leaders opened it!

Once.

Then, they assigned a subordinate to become proficient, AND insisted on getting immediate answers to urgent, random questions—because now it should be so easy!

So, instead of having insightful conversations with more-informed leaders, we spent time periodically rescuing subordinates trying their best—and often failing—to interpret their managers’ cryptic requests. It wasn’t what we’d hoped.

Admittedly, I had seen this same phenomenon dozens of times before.

But we were arrogant enough to think that other tools had the wrong data sources (or design, system, software, insights, whatever)—ours would be different! We could do it better, and as the saying goes, if we built it, they would come.

They didn’t.

In the end, our system did not transform how leaders sought information.

Even if we did have more comprehensive data, structured in more usable ways, there remained a disconnect between the brain of the person needing to know something, a clear definition of how to frame that question in data terms, and in-depth knowledge about how best to extract it.


Actually, about 90% of the time, our dashboards did contain the desired information. But only after we were able to help the business leader articulate what they really needed and help a user translate that into the appropriate query.


 

Right tools, right people

Now don’t get me wrong.

In many ways, the system we developed was successful. Clients bought it.

Eventually, a category of happy users evolved.

These were people familiar enough with data, with an aptitude for digging in, and an ability to decipher what was being asked.

They developed a routine for knowing what to provide regularly to their bosses and what might only warrant mentioning when it changed.

The right people became users of a tool that was right for them.


But the dashboards did not do their thinking for them. It did not solve their problems or answer their questions on its own.


This story—where companies buy easy, magical answers—will be familiar to others with enough battle scars and grey hair to have lived through the disappointment (probably multiple times).

Or worse, remember building an internal system themselves—at the request of their boss—only to have that same boss decide they don’t have time to learn SQL.

And when the queries are simplified even further, the boss doesn’t have time to look at those either.

A turnkey solution sounds great until we realize, like anything new, it might take time. Which is always in short supply.

Maybe we will get closer to the imagined mind reading as Alexa and Siri get increasingly better at interpreting our requests. But I think that’s far off.

Today, there’s still a human role.

 

Solving the wrong problem

While it took me decades to figure it out, I realized the real “problem” can’t be solved with only a better analytic interface.

Certainly, the newest tools make life easier for the right people.

But they don’t instantly transform non-data-oriented people into analytic wizards.  

Leaders (without hands-on experience manipulating data) get frustrated because they aren’t getting what they need from their analytic teams.

Even more frustrating, they don’t know how to define what it is that they do need.

So, because answers are not easily obtained from others, they conclude that such insights will only appear when they look directly at the numbers themselves. (They don’t). Or that a few easy clicks by an unfamiliar user will reveal important secrets. (They won’t).

Despite what enthusiastic vendors tell them, seeing it for themselves won’t lead to the amazing discoveries they seek.

While our egos convinced us we could solve the problem with better systems, technology, and design, we couldn’t.

Why?

Because it’s not a design problem. It’s not a data problem. It’s a communication problem.

Yes, I said a communication problem.

Not communicating/translating data into answers, although that helps, too.  It’s the ability to extract business needs and convert them into the specific questions analysts will answer.

 

Let’s take an example.

Business leader says: I need you to investigate our turnover trends. I need to see what groups have been quitting lately. Can you help?

Analyst says: Sure! Goes to the computer and spends time extracting turnover rates by age, gender, location, tenure, job type, front line, back-office, performance ratings. Returns with a series of graphics rank-ordering which populations have been quitting most in the past three months. Looks like new workers on the front line are quitting most often.

Business leader:  That’s interesting. But how do these relate to historical trends? 

Analyst: I’ll check! Returns with year-over-year trends, highlighting groups that have changed the most. Actually, the ones whose turnover have increased the most are high-performers, mid-tenure in the home office.

Business leader: That’s helpful. Can we tell if those people received a pay increase since last year? We’re going to use this for a compensation strategy meeting next week.

Analyst (sigh): I’ll check. Returns with summaries of the percent of quitters who received pay increase since the previous year.

Business leader (getting frustrated): Got it. But were they more likely to get pay raises than the high-performing workers who didn’t quit…?

Business leader decides maybe they should learn to get the results themselves….

This is a typical exchange. Requests tend to be transactional, short, and continuously changing. The analyst feels like they are chasing a moving target. The business leader feels they aren’t getting what they need (and starts to wonder about the mind-reading, fingertip solutions they’ve heard about).

 

What’s Missing

What’s missing here is (at least one) conversation.

Leaders need a chance to think through their business need more thoroughly and identify not only what they want to know, but also how the information will be useful and what decisions it will influence.

When analysts or other team members learn to have constructive dialog  about what’s behind requests, rather than assuming that the initial question is complete, the interactions become less transactional and more strategic.

When leaders get the information they need, they are less likely to search for their own “magical” ways to get information at their fingertips.

Wendy D. Lynch's headshot photo.

Wendy D. Lynch, PhD

Wendy Lynch is an experienced sense-maker and data scientist with over 35 years of research experience, primarily in business settings. She has played the role of Analytic Translator for hundreds of companies, from start-ups to Fortune 100 corporations. Her expertise is both in data analytics and effective communication, combining the two into a framework for optimizing the value of analytics in a business setting.  Connect with her through LinkedIn or email.

Newsletter: A Matter of Translation

Making sense where business and analytics meet.
Training professionals to speak both languages.

We won't send spam. Unsubscribe at any time.