Developing a Data Analytics Culture
We have had the opportunity of implementing a data analytics and reporting engine in organizations that were large global and complex, and also in environments that were smaller and more entrepreneurial. Here are some of the things we learnt that, we believe, may be worth sharing. The reality is that every organization is at some point along on a journey to leveraging the power of analytics for internal and external stakeholders. Just setting up an infrastructure for data lakes/warehouses, engines and reporting interfaces is hardly the toughest part. Transforming the culture of the organization, getting people to make hypotheses, build models, test and make strategic and tactical decisions based on evidence, are what make it all productive in the end. It is also interesting that with machine intelligence, interpreting data correlations and finding relevant results is a new skill which requires a good understanding of the business dynamics. And as more organizations start down the data monetization path and sources of data including sensor and unstructured data proliferate, it is even more imperative to think like the customer and create value that someone would be willing to pay for. Getting there is by no means easy, but here are some things to consider in capturing the value in the troves of data available with increased digitization.
Develop trust/ Find believers
- The essential prerequisite for data culture is the veracity, reliability and relevance of the data being presented. Inaccurate and incomplete data hurts adoption like nothing else. Therefore, the building blocks of accurate information i.e. a good master data management tool and data governance process is very important. Data migration and data mapping errors complicate the data store. Some estimate that 95% of the effort in building a data analytics program is in cleansing and pruning the data to make it accurate and relevant.
- Finding early adopters and change agents that have analytics bent is a good way to provide support in the initial phases of deployment. Satisfying the SME’s goes a long way in winning the masses as they are the role models.
Start small; think like the stakeholder
- Ultimately value lies in the eyes of the beholder. Creating an analytics infrastructure without appreciating the way value can be created is only half the battle won. Credibility builds on success. Finding high potential use cases and executing well is a good way to start.
Attack common business problems
- The best way to get adoption, value realization and to gain momentum is to pick common business problems like pricing optimization, inventory optimization, production scheduling and apply analytics to help improve financial and operational metrics in a quantifiable way.
Make it easy
- The process for running the analytics or manipulating the data needs to be quick and easy. This seems like a no-brainer, but it is amazing how many times we have discovered more ways to make it simpler. When doing the analysis in the tool becomes difficult, users start pulling the data and working on it with their own tools like Excel or statistical packages. So, the interface choice is important. If using a standard data visualization tool like Tableau etc. where UX is a plus, adoption becomes easier. Where it is a custom developed tool, spending time on the UX to make it easy to use but also visually interesting is important.
Resonate with stakeholders
- Identify and align the key stakeholders who have influence in the organization and can support the transition to a data-oriented culture. Having them on your side helps change management. In one of my companies, we had large TV screens mounted on the wall near the executive suite where we had visually arresting displays with key performance metrics using widgets that come with data visualization tools. Those became the talk of the town.
Define value/worth of the data to the user
- The opportunity cost of not relying on analytics support can be identified with examples where performance improvement has clear financial benefit. Nothing like bringing this number to daylight with in-depth discussions with SMEs to a) create interest b) prove ROI and c) develop a template to prioritize and get the best bang for the buck of effort. In my first job, as a 20 something logistics systems manager, I used analytics to create an optimization routine that increased the ordering frequency of high-volume products and reduced that of the low volume products to bring total inventory down by millions, without increasing the number of orders.
- Sometimes tooting your own horn is warranted. One should not underestimate the power of perception. It also helps generate more use cases with interest from other parts of the organization.
Like they say, no argument can conclude sooner than the one that is resolved with data. Being able to access, process, develop insights and conclusions differentiates organizations, and builds value. It not only increases internal efficiency but is increasingly a critical element of developing insightful products that solve real world business problems for customers. It’s a long-term cultural transformation that is essential to compete in the new world of analytics. All success in your journey.