Skip to main content

Proving value in Big Data projects is not easy

Typically, when talking about ROI, leadership expects to see hard benefits, like profits, sales, or savings. However, for the majority of companies, Big Data is still a rather new endeavour and as a result, measuring benefits, ROI and results is still a challenge (according to our recent survey, 47% of the respondents said it’s not easy to prove the value from their Big Data projects). This is accentuated further, as the transformative nature of Big Data technologies requires time to experiment and test, in order to gain not only the understanding of how to measure the results, but also of how these technologies can be best used.

Proving value in Big Data projects

Another critical factor that adds to the challenge, is the constantly changing landscape of these technologies. From on-premise, to cloud, to edge, IT teams often struggle to keep up with how fast technology is evolving and may rush into choices that may not be the best for their needs and can hinder ROI. In fact, based on our survey findings, the changing Big Data technology landscape is the biggest challenge companies are facing in adopting Big Data.

 however is not impossible. Value presents itself in many forms, different for each company, and therefore all possible ways in which Big Data can bring success need to be considered:

  • Define the scope. The first rule of measuring is to know what to measure. What need or problem does the project address? What domains, processes, departments will it affect? Knowing the exact scope of your Big Data initiative is essential to start defining the expected results and benefits.

  • Find the value of intangibles. Don’t forget to measure the soft benefits, like the skills your team gains from early projects. Having this knowledge and skills, your team become experienced enough to manage Big Data initiatives more effectively, which will in turn bring the hard ROI leadership is looking for.

  • Define the total costs. The level of investment in any project is largely defined by the expected total cost. It is therefore important, in order to calculate a project’s ROI, to know exactly what this investment (and cost) is going to be. Besides the initial, obvious costs like licensing, hardware, implementation, and maintenance costs, a Big Data project might also require some investment towards data governance, storage, hiring and/or training. There are two hidden costs however to which most project failures within the Big Data domain are attributed -these are the cost of potential delays and the cost of changing technology requirements. In putting together the budget for the project, it is critical to provision for these two costs.

  • Be realistic about the time frame for achieving ROI. Unfortunately, most Big Data projects take some time (often over a year) before they can showcase hard benefits. In drafting the project plan and defining the scope, the expected benefits and the costs, it is important to be realistic about the time it will take after the implementation to start showing measurable results.

  • Know when to let go. In this fast-evolving landscape most companies are in a perpetual experimentation stage with Big Data. This can actually be an advantage, however it is important to know when to step away from a project that is not proving fit-for-purpose, or the cost is starting to outweigh the benefit. Netflix is a classic example. In 2012, the company spent $1 million on ‘content recommendation engine improvements’ that it never used. This was, for the most part, due to the magnitude of the engineering efforts needed to complete the project. However, the reason why Netflix has ultimately succeeded in building this highly profitable content recommendation system was in part because they opted out of implementing this $1 million worth of code. The decision makers had a goal, which was to reduce the number of customers who unsubscribed, and knowledge of the cost vs. the potential benefits in that particular context, a 10% improvement in recommendation accuracy was, at that time, simply not worth the estimated engineering investment.

If you want to find out more about how much progress has been made with Big Data initiatives in the UK over the past year and whether they are achieving the expected return on their investment, download our latest report The State of Big Data in the UK 2019.

Let’s engage