Analytics offers heaps of promised benefits – from reduced churn to increased sales to deep and actionable insight on customer preferences. But sucessfully deploying analytics in your organization can be challenging. Where do I start? Which projects should be prioritized? How should my organization be structured for success? What technology innovations are key to include? The answer, simply put, is to develop an analytics strategy – or, in layman's terms - a plan. It may sound obvious, but in our experience, the missing step for many companies is spending the time required to create a simple strategy and roadmap for how data, mathematics, algorithms, tools, and people come together to bring about business value.
In this blog post, appropriate for both business and IT professionals, I introduce a proposed approach for developing an analytics strategy. The approach is based upon my own experience from managing analytics strategy processes in the private and public sector, formal and informal talks with SAS clients, and research from SAS Institute, McKinsey Global Institute and Gartner.
The approach consists of the following six steps, each described in greater detail below:
- Understand the business strategy and strategic focal areas
- Develop your analytics vision and set target maturity levels for your core processes
- Develop business ideas for analytics supporting the vision
- Prioritize and develop a strategic roadmap
- Develop a blueprint of the resulting target architecture
- Decide on organization and capability development
Needless to say, some of these steps will need to be revisited more than once as your analytics strategy process moves forward. My experience, however, that they highlight the key aspects to consider when you want to take your organization into the era of data-driven decision making.
Step 1: Understand the business strategy and strategic focal areas.
It might sound obvious that your analytics strategy must be rooted in the business strategy. What we tend to see however, is that analytics initiatives are not given priority simply because they fail to support the top strategic focal areas of the company. We therefore recommend that you start your analytics strategy process by spending time with the Chief Strategy Officer, or even the CEO and his management team, to really understand:
- What is your company trying to achieve the next 1-3 years?
- What core processes of the value chain is the strategy aiming to change?
- What key change programs are taking place?
Once you have the business strategy broken down into manageable pieces, (e.g. by using a simple strategy tree or Porter’s value chain), you are ready to start focusing the analytics strategy on the areas that really matter for business success.
Step 2: Develop your analytics vision and set target analytics maturity levels for your core processes.
Simply put, a business strategy is all about changing one or more of the core processes of your company. In setting up the analytics strategy, your job is to create a vision for how analytics should be adding value to these processes. One way of further concretizing this vision is by using maturity models. The use of a maturity model allows your organization to have its methods and processes assessed according to management best practice, against a clear set of external benchmarks. The maturity model should focus on the core processes of the company at a fairly high level, and using these you need to have two separate discussions:
- What is our current maturity level? Ie to what extent is the organization utilizing analytics in this very process – and are we using it in a consistent manner?
- What should be our target maturity level? What should be our ambition for utilizing analytics and data in this process? Should we aim for automated, real-time analytics where we embed advanced analytical models into business decisions and customer-facing processes? Or should we simply aim at a lower maturity where it is up to each decision maker to utilize own-grown analytical models?
Both questions can be very hard to answer. In my experience, the first question is very much about facilitating discussions to drive a common understanding and consensus among managers and analysts. The second question should be based on facilitated discussions with the same group of people, taking into consideration e.g. strategy guidelines, market best practices, your current maturity level - and what your industry peers and competitors are doing.
Step 3: Develop Business Ideas for Analytics.
This is the creative and fun part of the strategy process. Developing Business Ideas for Analytics is about forming the set of concrete initiatives that will help you reach your strategic ambition, ie the target analytics maturity levels set in the previous step. In developing these initiatives (or possible project charters, if you will) you must be explicit about answering three questions:
- What business challenge is the initiative addressing?
- What are the key elements of the proposed solution?
- What is the business case and the associated risks?
In developing the business case, special considerations include:
- What data is required? Assembling and integrating data is essential. Companies are buried in information that’s frequently siloed horizontally across business units or vertically by function. Critical data may reside in legacy IT systems that have taken hold in areas such as customer service, pricing, and supply chains. Complicating matters is a new twist: critical information often resides outside companies, in unstructured forms such as social-network conversations.
- What analytical models will be required? Integrating data alone does not generate value. Advanced analytic models are needed to enable data-driven optimization (for example, of employee schedules or shipping networks) or predictions. A plan must identify where models will create additional business value, who will need to use them.
- How will analytics be integrated into work processes? The output of modeling may be strikingly rich, but it’s valuable only if managers and, in many cases, frontline employees can access, understand, and use it. Output that’s too complex can be overwhelming or even mistrusted. Often what’s needed are intuitive tools that integrate data into day-to-day processes and translate modeling outputs into actions.
Towards the end of step 3, you will have a large number of possible initiatives that, if all implemented, will take your company's processes directly to the target maturity levels agreed in Step 2 described above. There is one thing I can guarantee you, however: You will not have the resources to implement them all in a "big bang" fashion. You are now entering the most critical part of the process: Project prioritization and roadmap development.
Step 4: Prioritize and develop the roadmap.
The essence of a good analytics strategy is that it highlights the critical decisions, or trade-offs, your company must make and defines the initiatives it must prioritize. In this part of the strategy process, you need to select, from a long list of the Business Ideas for Analytics developed in Step 3, the ones that will best support business goals. Successfully grappling with these planning trade-offs requires a cross-cutting strategic dialogue at the top of a company to establish investment priorities; to balance speed, cost, and acceptance; and to create the conditions for frontline engagement. The output of this step should be a roadmap highlighting which initiatives that will be undertaken, in which order and who are responsible for seeing them through.
Step 5: Create a blueprint of the resulting target architecture.
For business users, technology itself is almost irrelevant – it’s almost a red herring. Business users couldn’t care less about in-memory analytics. What they care about is being able to rapidly and intuitively analyze large amounts of data. As IT professionals we know, however, that building a robust analytical architecture will be key to realizing the business outcomes set forth in the roadmap. Looking across the set of initiatives, your architects must assess the required changes for data, applications/tools and technical architecture and lay a corresponding transition plan. Typically, Legacy IT structures hinder new types of data sourcing, storage, and analysis. Existing IT architectures may also prevent the integration of siloed information, and managing unstructured data often remains beyond traditional IT capabilities. Fully resolving these issues often takes years. Quickly identifying and connecting the most important data for use in analytics and then mounting a cleanup operation to synchronize and merge overlapping data and to work around missing information can be noe way of getting started.
Step 6: Decide on organization and capability development.
Much as some strategic plans fail to deliver because organizations lack the skills to implement them, so too your analytics strategy will disappoint if your organization lack the right people and capabilities. You need a plan for assembling and organizing a talent pool of the right size and mix to execute on the roadmap and implement the target architecture. And the best plans will go further, outlining how your organization can nurture data scientists, analytic modelers, and frontline staff who will thrive in the coming era of analytics and data-driven decision making.
Once finished, the power of your analytics strategy is that it provides a common language allowing executives, IT professionals, data scientists, and managers to discuss where the greatest returns will come from and, more important, to select the two or three places to get started. When your plan is in place, execution becomes easier: integrating data, initiating pilot projects, and creating new tools and training efforts occur in the context of a clear vision for driving business value—a vision that is less likely to run into funding problems or organizational opposition. Over time, of course, the initial plan will get adjusted. Indeed, one key benefit of analytics is learning things about your business that you simply could not see before.
McKinsey (2013): Big Data - What's your plan?
Gartner (2013): Top CIO Trends
Interviews with SAS clients
Great post, Christian!
The biggest caveat in my mind is, however, how you create a coalition behind the idea of embarking on the journey and commit the significant resources need for creating an analytics strategy. And how do you subsequently ensure that there is a company culture, which will nurture the strategy through the execution phases rather than resisting an era of data-driven decision making? I would venture that carefully planned experimentation could be one way of getting decision makers to warm up to the idea of a full-blown strategy.
Pingback: The Role of Predictive Analytics in the Strategy of an Insurance Company – Data Driven
Are there some examples and success stories that can be share on this topic?