Intel Corporation considers how best to evaluate analytics projects to meet enterprise goals and grow business trust.
Wherever you are on the analytics journey, shifting up a gear can be a challenge. Many organizations we work with find that success comes incrementally, using a series of projects to develop capabilities, experience, and trust with lines of business.
Increasing business interest can create its own challenges as expectations increase and more project requests come your way, not all of which will be suitable. So, how do you avoid being swamped with requests and decide which projects should be prioritized? Which lines of business would benefit the most, or is most able to adapt to the insights you provide?
Since we set up Intel’s Advanced Analytics group in 2011, we have learned to prioritize projects based on criteria that maximize our chances of success. Alongside the level of potential impact, we have found that criteria such as sponsorship and resource availability are equally critical. Here are the criteria we follow to assess potential projects:
- Executive sponsorship. We have learned that the primary criterion for success, above and beyond any potential benefit coming from the data, is pre-emptive buy-in, from both senior management and teams who are going to benefit from the results. If this does not exist up front, it only becomes harder to obtain later in a project.
- The right problem. Data sets are agnostic about the problems they can solve, or the opportunities they bring. It is therefore incumbent on the project to focus on a problem or opportunity that maximizes value to the business. It should be possible to clearly state this in terms all parties understand.
- Data. Once sponsorship and value have been considered, questions turn to achievability – and therefore, the data itself: wherever it is sourced, is it of sufficient quantity, quality, and type to make the project feasible and worthwhile? It is usually better to test this in advance, for example confirming API availability, rather than finding out later that it is not good enough.
- Resources. The availability of suitable and sufficient skills, tools, and processing power will all dictate how well the project is delivered, and whether it achieves the anticipated level of business value. If you are unclear on the resources available in advance, which may link to a lack of executive commitment, you are adding unnecessary risk to the project.
- Time. The time taken to deliver a project can also have an impact on the value it can bring, which can decline over time as requirements and priorities change, which is why we advocate agile, early value approaches to delivery. For example, you can consider how quickly minimal viable insights can be delivered, as well as longer-term insights from a broader data set.
- Projected benefits. Business value, in terms of solving a specific problem, generating or saving money, is the ultimate gauge of success for an analytics project. For organizations at an earlier stage on their analytics journey, it needs to be possible to evaluate benefits to a specific part of the business as mediocre results could undermine any future efforts.
Over time, analytics projects will get longer and the infrastructure needs greater. As your experience grows and your relationship with the business continues to develop, you will be looking to establish a more solid basis for the future, potentially adopting more advanced analytics solutions while adding higher-risk, bigger reward analytics projects into the portfolio.
As you advance, you should keep measuring and testing projects against the criteria you have set. As your business changes, so will requirements, constraints and indeed the value you can bring. Your analytics program can practice what it preaches, using project data to ensure your activities are aligned with a transforming enterprise.
Learn more about how advanced analytics can help you transform your business, and what you can do to make it happen, by downloading the new Five Steps To Delivering The Data-Driven Business Whitepaper from Intel.