Subscribe to DSC Newsletter

5 Steps to Building a Big Data Business Strategy

“The problem is that, in many cases, big data is not used well. Companies are better at collecting data – about their customers, about their products, about competitors – than analyzing that data and designing strategy around it.”  

How can this still be the case? I mean after 5+ years of experience with Big Data, have we not learned a darn thing? We get the following observation from no less than the Harvard Business School:

The new attention being given to data today is because suddenly, everywhere, it’s become much cheaper to measure,” says John A. Deighton, the Baker Foundation Professor of Business Administration at Harvard Business School. “Used well, it changes the basis of competition in industry after industry.”

Used well, it [big data] changes the basis of competition in industry after industry!” What more does one need to say? In spite of its game-changing opportunity, organizations still have not gotten Big Data right.

This is consistent with what I observe on my many travels and conversations across a multitude of different organizations. I would estimate that less than 2% of these organizations really know what they are doing to exploit the potential value of big data to power their business models.

I think the biggest problem is the focus on creating a big data strategy is as soon it’s completed, it’s outdated. New unknown-potential data sources emerge, new hardware innovations drive new capabilities, new data management tools and techniques evolve, new open source advanced analytic tools pop out of universities, new edge analytic architectures become scalable, etc.

The growing wealth of “monetizable” data (social media, mobile, IOT, wearables, images, photos, video) and the absolutely astounding availability of advanced analytic tools (many of them open source such as MADlib, Mahout, H2O, OpenAI and Google’s powerful Tensorflow) are obliterating your big data strategy before your expensive consultants can even get the leather binding in place.

Too many organizations are making Big Data an IT project instead of making big data a strategic business initiative that exploits the power of data and analytics to power the organization’s business models.

Figure 1: Change Your Big Data Focus to Change Your Big Data Results

 

Build a business strategy that incorporates big data. Build a business strategy that uncovers detailed customer, product, service and operational insights that can be the foundation for optimizing key operational processes, mitigating compliance and cyber-security risks, uncover new revenue opportunities and create a more compelling, more differentiated customer or partner experience.

Build a business strategy that exploits the power of data and analytics to exploit changes in market demands, customer expectations, competitive moves, commodity prices, student debt, stagnating salaries, the underemployed, political trends, decline of the manufacturing middle class, fashion trends, Chicago Cubs winning the World Series after 108 years…yes, and many of these changes are sudden and unpredictable.

So how does one build this business strategy that exploits the power of big data?

Here is our 5-step approach:

Step 1: Start with the Business Initiatives

How can you transform the business if you don’t understand what’s important to the business? How can you transform the business if you don’t intimately understand what the business is trying to accomplish, why, and the desired business outcomes? Understanding the organization’s key business initiatives is the key to identifying the supporting decisions (use cases), analytics, data, and underlying big data architecture and technology requirements.

Invest the time upfront to envision how the growing bounty of internal and external data coupled with advanced analytics might impact the organization’s most important business initiatives. Brainstorm with the key business stakeholders the decisions that they are trying to make and envision how predictive analytics, prescriptive analytics and ultimately cognitive analytics can help the organization to accelerate, optimize and continuously learn from those decisions.

Step 1 requires intimate engagement between the Business and IT stakeholders. This is not something that IT does alone and then “presents” the results to the business stakeholders in some monthly “alignment” meeting. If the business stakeholders are not leading this effort, then the effort is doomed. Welcome to the 98% who just don’t get it.

Step 2: Identify and Validate Supporting Use Cases

Step 2 involves taking the decisions captured in Step 1 and clustering or grouping the decisions around common subject areas. These clusters become the key business “use cases” that support the organization’s key business initiative (see Figure 2).

Figure 2: Capture and Validate Top Priority Use Cases

 

The use case documentation should capture both the sources of business value as well as the potential implementation risks (“eyes wide open”). Tie the use case back to the organization’s key financial goals and assess the impact of the use case on each of financial goal. Estimate the financial impact and Return on Investment (ROI) if the use case is successfully executed over the next 12 months. Focus on the “4 M’s of Big Data”: 

Step 3: Prioritize Use Cases

Step 3 may be the most difficult step because it requires organizations to do two things that they don’t like to do: prioritize and focus. Prioritizing and focusing are not popular concepts for many organizations because of political and organization pressures. But “peanut buttering” key resources and organizational commitments across a multitude of use cases is the best way to guarantee that no use case gets successfully executed.

If (and that’s a big IF) you can convince the organization to build out their big data business strategy one use case at a time, then that enables the organization to become expert at harvesting the organization’s data and analytic digital assets (and customer, product, service, operational and market insights) and applying those digital assets to subsequent use cases.

The Prioritization Matrix in Figure 3 is an excellent management tool for driving organizational alignment AND commitment around the organization’s top priority use cases.

Figure 3: Prioritization Matrix

 

Step 4: Brainstorm and Prioritize Data Sources

Step 4 focuses on brainstorming and prioritizing the different data sources that support the top priority use cases. Since Data Science is about “identifying the variables and metrics that might be better predictors of business or operational performance”, it is important to have a process where the business stakeholders can collaborate with the data science team to identify and test different data sources to identify those that might yield the best predictive models (see Figure 4).

Figure 4: Mapping Data Sources to Use Cases

 

Step 5: Determine Economic Value of Your Data

Step 5 focuses on linking the financial value of the use cases to the data sources (variables and metrics) that support the predictive capabilities necessary to successfully execute that use case. That is, the financial value of the use cases becomes the financial value that is then allocated or attributed to the support data sources (see Figure 5).

Figure 5: Determining the Economic Value of Your Data

 

“Used well, big data changes the basis of competition in industry after industry.”

“Used well” means leveraging advanced analytics to uncover insights about your customers, products, services, operations and markets that the organization can use to optimize key operational processes, reduce compliance and security risks, uncover new revenue opportunities and create a more compelling, differentiated customer experience.

Remember: organizations do not need a big data strategy; they need a business strategy that incorporates big data.

Views: 1934

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Follow Us

Videos

  • Add Videos
  • View All

Resources

© 2017   Data Science Central   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service