No, most leaders do not know how to fix the 85% analytic failure rate.

Nov 04, 2023

You’ve heard the numbers before. And they are awful.

  • 85% of analytic projects in business do not provide value.
  • 90% of digital transformation projects fail.
  • 70% of online dashboards do not get used.

It boils down to a huge amount of waste, rework, frustration, and burnout.

Lately I have been interviewing analytic leaders to confirm that project failure really is this common (yes), hear about their perceptions (it’s pretty much always this way), and understand what they think will solve it (more on that below).

When asked, mostly they describe technical and procedural issues, caused primarily by business professionals having unrealistic expectations, or inadequate knowledge. Analytic leaders report that business teams want results to be:

  • Produced instantaneously (even though data preparation will be required)
  • An answer to their primary question (even though they can’t articulate what that is)
  • Definitive, with a single right answer (even when the project models likelihood)
  • Simple (even when there are many nuances)
  • Displayed in a slick report (whether it fits that format or not)
  • Confirmation of their assumption (even when the evidence is contrary)

Obviously, few projects can achieve all these desired outcomes.

On top of these mismatched expectations, too often, the project begins with a vague request, that analysts don’t clarify.

They report that the business question and the analytic answer are not aligned. Consequently, in most companies, business and analytic teams develop a dysfunctional dynamic. Business teams experience consistent disappointment and frustration, not getting what they need. Analytic teams frequently feel set up for certain failure, knowing what they produce likely won’t meet business needs.

The financial implications of this dynamic are obvious: huge waste and inefficiency.

Projects rarely produce actionable results. Answers are unsatisfactory. Results go unused. Work is redone multiple times. This erodes trust between the teams, often leading the business team hiring outside consultants, believing they can achieve what internal analysts cannot.  

But like in any dysfunctional relationship, there are serious ramifications for the people in both teams. Analytic leaders report very high levels of burnout, dissatisfaction, stress and anxiety in their team. An underlying defeatist attitude expands with every perceived failure.  Similarly, business leaders report ever-increasing pressure to be “data-driven” and provide evidence to support their strategies, combined with resentment about how hard it is to get the answers they need. Not surprisingly, we find these teams retreating to their corners, steeped in animosity.


What analytic leaders THINK the answer is

Faced with the alarming rate of failure, leaders of analytic functions naturally try to fix it.  Here are a few strategies I see.

  1.  Fix it by making non-analytic people more like analysts. This is a common theme among analytic departments. They emphasize that if business leaders could just have a better understanding of what it takes to accomplish analytic tasks, then their expectations would be more realistic. Plus, if they understood analytic terminology, their requests would be much more accurate and appropriate. To accomplish this, companies are launching data literacy programs for all employees, from entry level to the C-Suite.

What’s good about this:  Of course, having more data-savvy employees in every department will lessen misunderstandings and promote better communication regarding analytic needs. 

Why it may not work:  In many companies, a large majority of employees – including executives – have low data literacy.  Getting 70% to 80% of employees up to speed on data concepts may not be realistic.


  1.  Fix it by requiring exhaustive specifications and signoff before work can begin. For analytic teams who are tired of being told they answered the wrong question, some choose to update their processes. They implement a system requiring that any requests be defined in great detail, confirming methods and metrics before the project can begin. This facilitates a narrowing of the project. Business teams must sign off so that any misunderstandings will be documented, (and analysts can point out that, yes, this was exactly what they asked for). Analytic leaders report that business teams hate this process, but insist it is necessary to avoid being blamed.

What’s good about this:  Like universal literacy, having agreement about project parameters at the outset is always preferable. This helps set expectations and align mutual goals.

Why it may not work:  Because these two teams use very different terminology and have different levels of understanding of analytic techniques, very often the initial request does not reflect the real business need. Requiring exhaustive detail about the wrong question forces a level of specificity that is uncomfortable to non-analysts and discourages communication, brainstorming and discovery.


A better solution

 While some degree of education and structured agreement can certainly be helpful, the underlying problem is more human than that. These two teams need a reliable communication method designed to achieve mutual understanding about business goals, rather than detailed agreement about the minutia of the analytic process.  

Analytic translators are trained to have discovery conversations centered around questions that uncover and clarify the actions and decisions that the business is facing. The discovery process is not about narrowing a question to tiny specifics, but instead about, first, expanding the question to understand the overarching purpose. This is the opposite of the “detailed specifications” approach above.  Rather than compressing a request into the tightest box possible, it encourages generative thinking to understand the big-picture objectives before selecting the best way to achieve them. 

Also, rather than creating a dynamic designed to protect oneself from future blame, Analytic Translators work to create mutual appreciation and respect between the teams. These professionals report a drastic reduction in failure (and frustration).

If you are a leader concerned about the high failure rate in analytics, and you are open to investing in skills – rather than mandatory data classes and detailed processes --- to fix it, consider a kinder, more constructive (and effective!) approach. 


I’m Wendy Lynch and I help companies transform their Analytics into Actionable Value.

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.