Why So Many Data Analytics Projects Fail (And How to Save Them)

By on

Click to learn more about author Wolf Ruzicka.

Data Analytics has captured the imagination of businesses, governments, and organizations of all kinds. The potential insights that analyzing large chunks of information can unveil are undeniably powerful. But in reality, most Analytics Projects fail to achieve results. A giant spreadsheet of numbers can look impressive, but you can’t expect to divine meaning from them out of nowhere. Worse, companies may waste a lot of resources in pursuing Analytics Projects that ultimately are pointless, souring executives on the whole idea and missing the opportunities for useful Analytics. The reasons that these projects fail vary. But by looking at misconceptions and issues at the outset of a project and those that may arise during the implementation process, a useful approach to solutions can be gleaned.


Data Analytics can help you predict the future. That is the core of the appeal of Analytics Projects, and it is certainly true. But, there’s often a misunderstanding of what “predicting the future” means in this context. Excitement can lead to rushing ahead without a plan. Not having an Analytics Roadmap leads quickly to project derailment, especially if leadership applies old methods of management.

Setting expectations at the start of an Analytics Project can help avoid another common mistake at the start, where the idea of Big Data leads to diving in with everything all at once. To get good results from an Analytics Project, it’s much more helpful, not to mention a lot cheaper, to begin with small data sets. Throwing a ton of money and other resources at an Analytics Project isn’t a shortcut. It’s better to focus on gathering the right data at the right time than to indiscriminately grab data. This minimizes the risk of failure, saving resources to try again.

That’s critical to remember, because Data Analysis Projects are not a one-off. They are a system that builds on itself and gets better over time and with regular feedback. While enthusiasm is great, it’s a mistake to look for certainty when doing Analytics Projects. Expectations have to be set for both the process and the results, because there are plenty of moments where things can go wrong even if there’s a realistic understanding of what may come of the project, starting with the data itself.


Figuring out what data to use presents a minefield. Data may be siloed in tens of thousands of internal databases, all isolated from each other, and with complexities limiting access to data. Pouring all of that information into one giant warehouse brings its own problems in cost and delays. APIs are a great solution, as they can automate data sharing and start showing results with the data sampling on-hand, rather than waiting for all of the data that you have to become available.

And, even when data is available, that doesn’t mean it’s helpful. An all too common problem organizations face is dirty or inconsistent data. Dirty data includes misspelled entries, missing information and other small errors, but they can add up to extra uncertainty for any insights. It can also be a more systemic issue of definitions. For instance, one system might define Washington, D.C. as part of the Mid-Atlantic region, while another may not include that as an option and place it as part of the Northeast U.S. That doesn’t matter when the two systems are separate, but merging that data would lead to serious confusion that may not even be noticed before problems arise.


Data Analysis problems present a mix of potential technical and human obstacles. Navigating them all can be complicated, but the results are worth the effort it takes. Breaking down the data silos is the number one challenge for many teams, in both aspects, but it can be accomplished. When data owners worry about losing control of data, the benefits of sharing information should be pointed out. Then, with the use of APIs, they can be shown how data can be both shared and retained at the same time. And, as errors, both human and technical are discovered during the integration, every team has the immediate benefit of newly cleaned data.

While the right data is discovered and integrated, insights can already be found throughout that process in parallel. Their usefulness may be limited at first, but by gradually building up how much data the system takes in, the insights can become more accurate and approach a broader spectrum of predictions.

The enthusiasm in the business world for data projects is not slowing down, despite the failures many companies experience when they first tackle a Data Analysis problem. Getting and keeping all of the relevant stakeholders on board while conquering the technical aspects of a Data Analytics Project may seem daunting, but it can give a business the edge it needs to survive and thrive.

We use technologies such as cookies to understand how you use our site and to provide a better user experience. This includes personalizing content, using analytics and improving site operations. We may share your information about your use of our site with third parties in accordance with our Privacy Policy. You can change your cookie settings as described here at any time, but parts of our site may not function correctly without them. By continuing to use our site, you agree that we can save cookies on your device, unless you have disabled cookies.
I Accept