According to a survey by NVP Big Data and AI Executive Survey 2021, 92.2% of companies “identify culture — people, process, organization, change management — as the biggest impediment to becoming data-driven organizations.”
These results are not anomalous. Year after year, culture continues to stall aspirations of becoming analytic driven. However, some companies have found the magic elixir for transforming their culture from one based on intuition and outdated processes to a digital-first environment fueled by analytics, data science, and ML. The magic happens through five strategies that leaders use to future-proof their investments.
For the companies that can transform, the results can be impressive. According to the International Institute for Analytics (IIA), when comparing companies with different levels of analytics maturity (i.e., localized analytics vs. analytic companies), they found the following differences in financial growth:
As a data and analytics leader, you should consider the following recommendations to future-proof your analytics and data science strategy.
1. Align Analytics & Data Science with Business Strategy
Yes, this sounds a bit cliché, but one of the biggest causes of failed projects stems from a misalignment between analytics and data science professionals, different functional areas within the company, and the overall business strategy.
Organizations need to have a well-defined business strategy with business goals and OKRs (Objectives and Key Results). Then, you need to define a set of initiatives (projects) that map to those OKRs. Next, you need to create and map KPIs (Key Performance Indicators) to those initiatives. Then, and only then, should organizations establish analytics projects to achieve those key initiatives.
2. Start with the Digital Decision and Work Backwards
After the key initiatives and project teams are formed, organizations need to work backwards from the key business decisions that need to be made. Leaders need to pay close attention to the associated business processes that are impacted. The entire goal of data and analytic projects is to create value for the company. If the organization can’t change its behavior as a result of infusing analytics into a digital decision, what’s the point? In far too many cases, this is overlooked and is detrimental to projects and demoralizing to those working on the project.
Assuming the organization can take prescriptive action based on analytics, businesses need to embark on infusing analytic workflows into business systems. To effectively do this, they need to have a robust organization change management process in place as well as an ML Ops strategy.
3. Don’t Forget the People
People are an organization’s most important asset, and there are many dimensions to discuss. First, there continues to be shortage of qualified data scientists for businesses to hire and retain. Thankfully, this gap is quickly closing thanks to upskilling opportunities like the ADAPT program.
Many large organizations are looking to continue to leverage their existing workforce by making them more data literate and providing them with the training opportunities they need to become citizen data scientists. Whether you like the term citizen data scientist or not, there are many professionals that are using data and analytics today that are stuck in spreadsheet land — doing the same thing over and over.
Also, an analytic-driven organization is only as strong as the community that supports it. When looking at analytics technology, pay close attention to the robustness of the community. Does it provide a collaborative experience for employees to learn best practices and solve problems without the help of IT or expensive training programs?
4. Don’t Do it Again, Automate.
Related to #3, automation is critically important for organizations. In the history of the world, as systems and processes become more complex, they need to be automated. Let’s face it, knowledge workers today are fed up with outdated work processes. They abhor doing the same thing over and over. They yearn to use their skills to do innovative things like moving beyond generating descriptive “what happened” reports every week to creating drag-and-drop data science models and exploring “what could happen” and “what should we do next.”
Many organizations see automation as a critical competency. From robotic process automation, to chat bots, to low/no-code user interfaces, to AI and ML, if you have to do something more than once, there’s an opportunity to automate the process. Analytic Process Automation is certainly one piece of the digital transformation puzzle that can help businesses let their employees create breakthrough moments and solve problems in new and innovative ways.
5. Technology Considerations
As previously stated, technology is certainly an enabler to improve employee wellbeing and business performance, but it should be considered after the previous recommendations are understood. First and foremost, technology needs to be intuitive, approachable, and easy-to-use. Yes, I know, every technology vendor states that their tech is easy to use, which is why I encourage people to try before they buy.
Additionally, the technology should have a robust set of building blocks to build data and analytics pipelines that solve real business problems and lead to top- and bottom-line growth.
The technology should have automation embedded with it and a robust partner network who can help with your digital transformation. Lastly, it should have some AutoML capabilities that are backed by open-source projects like EvalML (this YouTube video is for the Pythonistas who may be reading this blog).