Why Do Analytics Projects Fail?

     Analytics has become an important part of the digital world. Analytics helps decipher information that can be crucial to a project’s success. However, many analytics projects fail; leaving a tainted image to a growing industry that could potentially mean much more to how people predict success. If there were more engagement and trust in analytics, more projects would succeed; luckily, there are methods to increase the chance of success.

     In reality, reports demonstrate that many analytics projects failed caused by management resistance, internal politics, and additionally, lack of communication and empathy. With management resistance and internal politics, analysts will have a more difficult time changing a company’s or person’s core values. Several companies may still utilize traditional methods simply based on how it may have worked in the past or do not trust to go a new route since complex analytics can be risky to assess. Additionally, the lack of communication and empathy are huge reasons why many analysts struggle with finding success with projects. The outcome of whether a recommendation is taken does not matter if the client does not understand the reasons behind it. Without communication, an analyst will not be able to read the client’s mind to clarify objectives; without empathy, an analyst could solely focus only on their technological tools, leaving any relevant information out. For example, I had difficulties persuading the board at a nonprofit organization, revolved around the film industry, to expand into monthly articles that would provide future workshop dates, interviews for up-and-coming filmmakers, and useful guides. I wanted to increase audience engagement, but the board wanted to raise awareness in order to gain more funds for films. If there was a difficult time to persuade a couple of people in a startup organization, the situation for larger companies that could have hundreds or thousands of employees could be seen as almost impossible.

     Therefore, the sizes of larger companies are a major reason why there’s a bigger percentage of failure for analytics projects. Analysts are going to have to answer to not one person but to many people. The possibility of one person to disagree with an analyst is almost fifty-fifty, depending on their values, but the possibility increases as more people engage and more opinions collide.

     However, there are five methods that can reduce or eliminate the chance of failure: effective questions, influence diagrams, agile analytics, uncertainty and sensitivity analysis, and interactive decision support. Asking Effective Questions can be useful such as asking open-ended questions, what can the client control or influence, and collaborating to get a grasp of relevant data that can fill in gaps between what the data can provide and what information is needed to improve decisions. Influence Diagrams help clients clarify their objectives, decisions, and uncertainties with the simple tool where the analyst will understand how decisions will affect the outcome. Agile Analytics can help analysts visualize their step-by-step process with clients to first identify objectives and refine them until satisfied. Uncertainty and Sensitivity Analysis predict how decisions will affect the future, discover gaps in mental models, and find underrated factors that can be important to success. Interactive Decision Support can be useful for clients to further understand what decision needs to be made with alternative scenarios. The models and methods should be simple yet detailed enough that clients’ experiences feel authentic to understand the reasons and possible outcomes to decide on the right one because, by the end of the day, they have to make the decisions.

     Ultimately, analysts need to understand that in order for the industry to change its perception toward them, they need to understand their clients and create simple yet effective models that can predict the project’s future. If analytics were used confidently and accurately that corresponded to the relevant objectives, the industry could massively pave the way for more usage and more success. Those who are in the field understand how vitally important analytics is, so the time to engage is more relevant than ever to help those receiving the information see what analysts see: the power of analytics is just beginning to rise.