Why Data Projects Fail – No Data Strategy

James Roberts & Shannon Ragland

Ever feel like the deck is stacked against you? Or the house of cards you’ve built is about to fall at
your feet?

We hear this a lot from companies who are frustrated with the state of their most recent data
initiatives—whether they are stuck mining through an overwhelming amount of data without a
clear sense of direction, trying to determine why their reports don’t pass the sniff test (i.e., the
numbers don’t seem accurate at face value), or coming to the realization that the time and resources
budgeted for a data project are overwhelmingly inadequate.

We frequently encounter companies experiencing these pains, which we discuss in our blogs on
Data as an Asset: Quisitive’s Approach to Data Management for Business Success and Data Insights
as Innovations: Six Keys to a Successful Data Initiative, where we indicate that a data initiative’s lack
of strategy, quality process and clear goals are primary reasons why only 13% of data and analytics
projects reach completion.

Despite its challenges, data is clearly a tremendous asset for driving company growth, as discussed
in our most recent blog post, Seven Business Goals for Deriving Value from Data, highlighting seven
meaningful ways to leverage data to uncover valuable insights for growing your business, as well as
some tactical solutions for achieving these goals. Therefore, don’t be discouraged. Instead, focus on
a well-defined business goal and simply be aware of the risks and potential pitfalls along the way.
Even still, beware; if 87% of companies are pulling the plug on their data projects before they ever
provide any value, further consideration of the reasons projects fail is warranted so you can move
forward with eyes wide open.

So why do data projects fail? The most basic answer is that they simply do not start right—they are
lacking the proper foundation to ensure a successful outcome. Whether it’s the wrong (or not
enough) team members on the project, inadequate technology, or the lack of a strategy with proper
goals, roadmap and success criteria, many data initiatives are at high-risk of failure before they’ve
even begun. Below, we offer the Top 10 reasons data projects fail before they are even set in motion.

1. Poor Business Alignment

2. Asking the Wrong Question
3. Incomplete Data Roadmap
4. Boiling the Ocean
5. Failing to Gain Stakeholder Buy-in
6. Establishing Inappropriate Success Criteria
7. Underestimating the Timeline and Budget
8. Short-sighted Data Solution Architecture
9. Faulty Data Analytics
10. Lacking the Skills and Expertise for True Data Insight

In essence, each of these failures is related to lacking a comprehensive data strategy for the data
initiative. Below we explore high-level details around each failure point.
1. Poor Business Alignment

Fundamentally, the most basic error involved in the Data Strategy phase occurs when business
leaders do not take the time to ensure proper business alignment across the organization. Proper
business alignment better positions the company to move from data-to-insight-to-action, thus
honing in on the most appropriate business questions to address.

2. Asking the Wrong Question

The most successful data projects start with a question that targets the specific data to be analyzed;
they typically don’t start with the data and hope for an interesting insight to magically appear. When
a data team is equipped with the question that matters most, they can more feverishly pursue the
collection, aggregation, and analysis of data that will best answer the primary set of business
questions.

3. Incomplete Data Roadmap

Specific business goals with well-defined questions and an established business case set the stage
for the creation of a data roadmap. At times, a proper business case cannot be made by a single
question alone. Creating a set of questions that utilize the same data or data sources maximizes
return on investment for data projects. In this way, the first activity essentially funds a second, more
promising revenue-generating activity.

4. Boiling the Ocean

Data initiatives seem daunting when they try to do too much all at once. For especially complex
projects, this would be analogous to trying to boil the ocean. When companies try to do too much at
once, they risk having the project sidelined for other competing priorities due to a lengthy
anticipated time to value.

5. Failing to Gain Stakeholder Buy-in

The data roadmap should be adopted by all business leaders who will be required to produce inputs
and/or take future actions. By ensuring that all necessary members of your team see the value in
the initiative—which is supported by a well-developed business case—you have a much lower
likelihood of project failure resulting from unanticipated stakeholder barriers.

6. Establishing Inappropriate Success Criteria

The solution should have a measurable impact on the business. Ultimately, the success of the project
will be determined by evaluating the impact against predefined criteria. This will establish the basis
for a Go/No Go decision for implementing the solution beyond this particular data project. As is too
often the case, however, data projects are evaluated against standard company KPIs rather than
success criteria specifically designed to measure the impact of the data-driven solution.

7. Underestimating the Timeline and Budget

Going over time or over budget is the single easiest way to kill a data project. For this reason, the
importance of accurate timeline and budget estimations cannot be understated. Given their
complexity, it should come as no surprise that data projects can be time-consuming to get from
conception to real insights. Despite their complexity, proper estimation techniques have much
greater probabilities of success when projects have a sound data strategy and avoid the other
pitfalls on this list.

8. Short-sighted Data Solution Architecture

Although most of the technology-related failures will be discussed in our subsequent series on Data
Preparation, it is worth mentioning here that the data roadmap and business case should include a
high-level data solution technical architecture. The goal of most technical solutions is to leverage
existing systems and functionality as much as possible while providing a flexible, scalable, and agile
framework needed for achieving the business requirements set forth at the outset of the data
initiative.

9. Faulty Data Analytics

Because data analytics are the least understood phase of the data project, given the marketing
efforts from technology companies making grand claims on the ease of use of their products, we will
be providing an entire series on the most common errors in analyzing data—ranging from poor
sampling techniques and experimental design, improper model selection, and erroneous inferences
and interpretations of results. The effects of these types of errors create scenarios where businesses
fail to uncover the value of the data when value exists, or perhaps even worse, businesses act on
information from faulty models that can’t be trusted once put into production.

10. Lacking the Skills and Expertise for True Data Insight

Many companies make the mistake of framing data projects as just technology initiatives. This
conception needs to be dispelled. Having the correct data infrastructure and analysis tools in place
is not enough to offer your company insight—the right talent is an essential component.

Each of these failures represents a key ingredient required for a cohesive data strategy. They can
have big implications for a company—lost revenue, wasted resources, and even undue stress on
team members. One bad data project can lead leadership to be more hesitant or unlikely to invest in
future data projects, which results in missed opportunities for critical insights that ultimately
impact your business in a big way.

So what is the status of your data project today—are you set up for success or for failure?

Most likely, you recognize that your organization is at risk of committing at least one of these errors.
You likely understand that improving your company’s capacity for growth and sustainability
through a strong Data Strategy can drastically minimize these errors. By leveraging the right tools,
alongside a strategic partner, your business can be well on its way to deriving the value from data
necessary to make transformative decisions.

About the Authors

Combined, James Roberts and Shannon Ragland—both formally trained in applied statistics during
their doctoral programs at the University of Texas at Austin and Arizona State University, respectively
—have over 25 years of experience in business and academia. In addition to publishing dozens of peer-
reviewed empirical research studies while teaching in higher education, they have over 17 years of
consulting experience working for small start-ups as well as large, international corporations. Their
clients have spanned across a broad range of industries, including retail/e-commerce, higher
education, financial services, entertainment, government, biotechnology and petroleum. It is James
and Shannon’s combined, expansive knowledge about how to manage and analyze data that make
them an asset to advancing Quisitive clients’ business goals that achieve sizeable returns on
investment.

Connect with James and Shannon on LinkedIn