10 Fundamentals for a Successful Advanced Analytics Program

For the last six years I have run analytics businesses for many leading global organizations across a variety of industries, and have observed first-hand, the dramatic evolution of the advanced analytics market during this time. From the arrival of Big Data and IoT to the development of AI-driven predictive, prescriptive and cognitive branches, this space can be pretty confusing and daunting, to say the least. There are many ways to succeed, but, it is also very easy to fail, and a robust advanced analytics roadmap can make all the difference in the world.

The advanced analytics space goes well beyond the realm of traditional analytics and can help us look ahead, through predictive analytics for example, instead of only looking through the rearview mirror. Organizations that have successfully implemented an advanced analytics program have seen incredible results for their businesses with significant spikes in growth and profits.

In my experience, while many firms today are attempting to tackle ambitious advanced analytics programs, many also start off on the wrong foot by overlooking some of the key fundamentals. Based on my experience, working with dozens of international clients and across diverse projects, here are 10 simple fundamentals that, in my view, should serve as the foundation for any advanced analytics journey.

(1)  Establish a balanced advanced analytics strategy.

This may seem fairly obvious at first glance, but it is quite astonishing to find many Fortune 500 companies struggling to just get the basics right. Any analytics strategy must incorporate these three essential aspects to lay a solid bedrock:

• Business capabilities or use case roadmap that answers the question—what solutions should we implement for our business, users and business partners to achieve our business goals? This will determine the value side of the equation.
• An information technology roadmap that aligns tightly with the business capabilities roadmap. A strong data foundation does not, by itself, create value (few exceptions aside), it enables the full range of business capabilities and value to be realized over time.
• A governance approach and roadmap that is in tune with the overall business strategy. We have often seen that implementing the technology can be the easy part. It is usually the organizational structure, culture, governance and related company politics that hinder success.

 (2) Establish the program goals of 10x value-to-cost and self-funding.

With the right level of executive sponsorship and correct scope definition, a $10 hard value for every $1 of program cost is an easily achievable target. If prioritization of business capabilities is also in order, self-funding is obtainable within a mere 90 days of program launch. If your analytics strategy is not meeting these metrics, perhaps it is time to pause and rethink your strategy.

(3) Learn to think and act with an ‘in parallel’ mindset.

An analytics program that is self-funded cannot be realized by working serially. Often clients think that they need to get their data fixed first, then do some basic business intelligence and reporting, before finally taking on an advanced analytics program. The key to success in today’s fast-paced world, however, is multi-tasking and working in parallel. Some advanced analytics, for example, can help fund other elements of the program and new business capabilities can fuel data transformation.

(4) Secure the right level of business sponsorship.

From my experience, the most successful analytics programs have senior business sponsorship and clear ownership. Some of my most successful analytics implementation programs were backed by key c-suite members—the global CFO, for example.

(5) Pick the right place to start.

If you have 50 new innovative ideas for advanced analytics use cases, the best place to start is usually on use cases that (a) have strong executive sponsorship, (b) have data readily available to solve problems, even if it comes from multiple sources, and (c) are low cost but high value for Phase 1.

Again, this may seem basic, but we see mistakes on this front frequently. Recently, I met a client that had started their analytics program with a global management dashboard spread across 10 countries. It was a very expensive and very time-consuming affair. Perhaps, they would have been better off starting out with a few predictive analytics “sprints” in parallel to gain faster momentum with business stakeholders.

(6) Scale beyond science projects.

There are plenty of analytics vendors and business partners out there who can offer to build cool tools for you as you launch your journey in the advanced analytics space. However, the bigger question you need to address is which will be the most beneficial a couple of years from now, when we need to scale these solutions. Science projects are easy, it’s the scaling that’s difficult. A few challenges include (a) making sure you have a plan for how the users will actually be able to consume the analytics and take appropriate action with it, and (b) making sure you have a platform as well as data science methods that can keep up when you have dozens of projects going full throttle.

(7) Embrace diversity.

During the first half of my career, I worked extensively in the enterprise resource planning space where there was a fair amount of homogeneity across project resources with similar skill-sets and training background. During my last several years in the advanced analytics domain, I have met hundreds, if not thousands of professionals who are best described as being from completely different planets. While the incredible diversity in this space can make it a bit more difficult to assemble a “winning team”, I personally love the challenge and so should you.

(8) Focus on teamwork and collaboration.

The formula for a successful advanced analytics program is a balanced resource mix of not only skills but also personalities. I would, for example, rather work with an A- data scientist who works well with others rather than an A+ data scientist who is always the smartest person in the room. Collaboration is king.

(9) Inculcate the spirit of life-long learning.

As Thomas Friedman explains in his recent book, Thank You for Being Late, we have reached a point where technology is changing faster than humans are able to adapt. We must keep up with the rapid pace of change or we risk becoming obsolete. This challenge can only be overcome through life-long learning and constant adaptive change.

The recently announced Infosys Education Center in Indiana, U.S., is a great example of investment in the culture of life-long learning and inculcating vital skills needed for a continuously evolving digital economy.

(10) Be hands-on.

We all need to find ways to be hands-on with analytics technology. You may just be a project leader or a business analyst or a practice leader, but you need to find ways to sign-on and learn your trade at a hands-on level. Generalists who are not particularly tech-savvy will struggle in the coming years, but hands-on specialists will thrive.

These 10 simple but fundamental recommendations barely scratch the surface of the complex subject of advanced analytics. However, with these essentials in place, I am confident that you and your organization can and will be very successful.