Loading content...
Pedro Arellano
Breaking Through the BI Barrier
It's time for the BI industry to finally break through the 25% adoption barrier. Here's how.
It's time for the BI industry to finally break through the 25% adoption barrier. Here's how.
Since its origins over three decades ago, the business intelligence (BI) & analytics industry has worked to help organizations derive greater business value through improved decision-making. Over the past decade in particular, virtually every vendor in the space rallied around the idea of data democratization, which implies an increase in data literacy and broad adoption of analytics tools.
These are worthy aspirations, but unfortunately, the reality is quite different. A recent study shows that, on average, only 25% of employees use BI and analytics tools. Additional research reveals that nearly three-quarters of people feel overwhelmed or unhappy when working with data.
It’s unacceptable that such a small percentage of workers use analytics to make decisions, but, at the same time, it’s not surprising. I’ve witnessed these problems many times throughout my career: people frustrated because they have to wait weeks or even months to get the reports they asked for, end users dumping data into spreadsheets, or business leaders paying expensive consultants to manage their BI implementation.
In this article, I’ll explore the reasons behind the industry’s inability to reach a wider audience and share a perspective on what needs to change. I’ll focus specifically on tools and the role we, as solutions providers, can play. That said, I acknowledge that BI adoption goes beyond technology; it also involves issues of organizational culture, corporate processes, data skills, and even human psychology.
When I started my career in the late 90s, IT departments staffed with data experts owned every aspect of BI inside organizations. All your data needs (e.g. which customer segment engages more with a particular campaign? Or what does my sales velocity look like compared to the same quarter last year?) went into a queue and you were forced to wait a long time to receive your reports.
Roughly a decade later, tools like Tableau, Qlik, and Spotfire enabled a new audience—data analysts—with easier ways to explore data visually and build analytic content like reports and dashboards.
The industry deserves credit for bringing BI outside the confines of the IT department, but this rate of progress plateaued. For every data analyst delighted with their data visualization tool, there are countless people who still can’t get the business insights they need.
BI adoption plateaued because the industry viewed analytics as its primary, and sometimes only, function; it built products for people whose main job is analyzing data: data analysts, data engineers, and data scientists.
But what about everybody else? What about those who lead very data-driven functions but aren’t data experts? Sales & marketing professionals, heads of operations, retail store managers, warehouse supervisors, and HR leaders—BI vendors have not designed tools for people like them, and functional applications like Google Analytics and HubSpot Analytics can’t deliver insights outside their specific silo.
A VP of E-commerce does not spend their day thinking of how to build a better data pipeline or how to design a better dashboard. Instead, they care about identifying the right mix of products, driving more shoppers to their sites, and ultimately, growing their business. So why does every BI product force business people to wrestle with dashboards and data pipelines?
I remember a conversation with the head of Analytics at a leading global telecommunications company. Despite significant investments in BI technologies and programs, they had only reached a fraction of their organization’s user population. Expressing his frustration, he said, “I already have 8,000 dashboards. I don’t need more dashboards!”
You know that phrase, “If all you have is a hammer, everything looks like a nail”? Well, the dashboard is the hammer of BI. The industry developed an over-reliance on them as a way to communicate insights.
To be clear, I’m not against dashboards. They serve an important purpose. A dashboard can monitor the current state or historical performance of specific business metrics. They’re good at answering predefined questions:
On the other hand, dashboards can’t help you if you don't know exactly what questions to ask. You end up with a very narrow understanding of your business and miss valuable opportunities simply because you didn’t ask the right questions or you couldn’t contort the dashboard to give you what you wanted.
For example, here at Wallabi, we recently launched our website and wish to understand how to increase traffic. We’re not digital marketers and may not intuitively know the right data to examine, but our own tool uncovered, without asking, certain referral sources that have a higher bounce rate. Using only Google Analytics it’s unlikely we could have figured that out. We would be losing valuable opportunities without knowing.
Second, dashboards are not necessarily the most intuitive method for consuming insights. If you have ever landed on a dashboard and thought to yourself, "what in the world am I looking at?" then you know what I mean. Dashboards expect us to make connections between the different charts, KPIs, breakdowns, and data points to draw useful conclusions. Often, this can create excessive cognitive load and leads to an experience that overwhelms you more than it helps you.
BI has worked well for some, but not for everyone. The conventional approach to analytics excludes many people. The industry will remain stalled at 25% adoption until we break with established ways of thinking about BI, its purpose, and the people it can benefit.
Rethinking BI begins with acknowledging and embracing the diversity of the people who need data-driven insights. Our individual data needs vary greatly depending on our specific functions, skill sets, and the types of decisions we make.
Vendors must design user experiences with this diversity in mind. They must recognize that business teams need more than pre-canned answers in reports and dashboards; and make it possible to deliver insights without requiring the skills of a data expert. BI experiences should support the various ways in which people consume information and adapt to how data-driven insights support different business functions.
We must also solve the problem of irrelevant insights. The overabundance of reports makes it nearly impossible to sort through all the noise to find the truly meaningful insights that matter most. It’s no surprise that so many people feel overwhelmed by BI tools.
The world has moved to algorithmic personalization: YouTube, Netflix, Spotify, and Amazon recommendations as examples. We have come to expect that our tools can anticipate our needs, but for some reason, BI has not caught up. There is no reason why our analytics cannot become more relevant and useful based on a user's specific job function, responsibilities, goals, and business circumstances.
A merchandiser for an apparel retailer, for example, will use BI to analyze past sales and attempt to predict what customers will want in the future. What if the tool also knew that the merchandiser is rolling out a new athletic clothing line? Or if it understood the characteristics of the target market and current fashion trends? That additional context adds substantially greater relevance and value to the insights generated.
Finally, traditional BI ignores the way people work together. The idea of collaboration has been reduced to sharing reports and dashboards with others, maybe allowing them to add comments or edit the content. But collaboration is more complicated than that.
Decision-making is often a collaborative, community-driven exercise. Dr. Linda A. Hill, a Harvard Business School professor and an expert in leadership and innovation, uses the term “discovery-driven learning” to describe how people with very different perspectives collaborate to drive change and innovation. We’re influenced by input from our teammates. We tend to trust information more when our colleagues use it. Humans are, after all, social animals, and we don’t work in a vacuum.
The industry has come a long way since its inception, but what got us to 25% adoption won’t get us to 50%, 80%, or 100%. It’s time for BI to break this barrier and finally reach all those people who have been underserved for so long. We must:
The BI industry must reimagine how it helps people, but this will require breaking away from deep-seated conventions and thinking in ways that go against the grain. The result will be more inclusive BI technologies that enable everyone, regardless of their role, background, or expertise, to benefit from data-driven insights.