Bill Schmarzo

A recent article on Fortune.com titled “Why Big Data Isn’t Paying Off for Companies (Yet)” highlights that the vast majority of companies are still struggling with big data. The article quotes a study from the American Institute of CPAs (AICPA) and its London counterpart, the Chartered Institute of Management Accountants:

Just 27% of C-level executives interviewed in the study think their company makes “highly effective” use of data, while about a third (32%) say access to mountains of information has actually “made things worse.”

The research highlighted three hurdles that are inhibiting big data success:

• First, organizations are in too big a rush to do something. Consequently, they skip critical business and organizational alignment steps, and dive right into the technology decisions.
• Second, organizations start too big. Organizations start by developing a “big data strategy” when they still do not know what they are trying to accomplish from a business perspective.
• Third, organizations do not address the political issues upfront, hoping that these problems magically take care of themselves later

This dumbfounds me, because the key to business success with big data is the upfront work – the stuff that you must do before you dive into the technology.
To address these hurdles, we use a very simple approach called the Big Data Vision Workshop. The Big Data Vision Workshop focuses on driving the business success of big data by identifying and prioritizing the business, data, analytic, and organizational requirements necessary to support an organization’s targeted business initiatives.
Note: many organizations are confused about the difference between a “business initiative” and a “use case.” Here are my definitions:

• A business initiative tends to be a cross-functional business priority that the organization is trying to accomplish over the next 9 to 12 months. A business initiative is “what” the organization is trying to accomplish over the next 9 to 12 months from a running the business perspective.
•  A use case identifies “how” the organization will achieve the business initiative. Think of the use case as a common grouping of decisions or outcomes that need to be accomplished for the organization to successfully achieve the business initiative.

For example, if the business initiative is to “improve customer acquisition”, then the use cases could include:

• Improve customer profiling effectiveness
• Improve customer segmentation and clustering effectiveness
• Improve prospect targeting
• Improve messaging effectiveness
• Improve promotional effectiveness
• Improve trial effectiveness
• Optimize channel effectiveness
• Improve re-targeting effectiveness
• Etc.

But before describing the Big Data Vision Workshop process, I want to highlight some key principles behind it.

Big Data Vision Workshop Underlying Principles

• Organizations do NOT need a big data strategy as much as they need a business strategy that incorporates big data.
• Organizations don’t fail at big data because of a lack of business opportunities; they fail because they have too many.
• Business and IT leaders must co-lead the big data journey starting with identifying, brainstorming, and prioritizing the big data business opportunities or analytics use cases.
• Business leaders need to be committed to treating analytics as a business discipline, in the same way they treat accounting, finance, marketing and organizational sciences.
• Organizations culturally need to embrace data and analytics in order to optimize their key business processes, uncover new monetization opportunities and deliver a more compelling customer experience.

If these principles conflict with your organization’s business and IT objectives, then the Big Data Vision Workshop is not for your organization.

Big Data Vision Workshop Description

The Big Data Vision Workshop is a 3 to 4 week engagement which is 100% focused on identifying and prioritizing the customer’s technology, data, organizational and business needs . The Big Data Vision Workshop is designed to show business leaders how they can start their big data journey today via an engagement that is:

• Business-centric: focuses on identifying and prioritizing the data sources and analytics necessary to advance the client’s targeted business initiatives
• Non-technology: creates 10 to 12 illustrative analytics use cases using data science on a small sample of the client’s data to demonstrate the “realm of what’s possible”
• Organizational-alignment: drives IT and Business collaboration around high-value, high-feasibility use cases where the business owns business value determination and IT owns implementation feasibility determination

The Big Data Vision Workshop commits the first two days understanding the client’s strategic goals over the next 9-12 months, and then targets a business initiative on which to focus that supports those goals. Interviews (by business function) and group brainstorming sessions (across business functions) are conducted in order to capture the decisions, questions and data necessary to support the targeted business initiative (see Figure 1).

Figure 1: Business Initiative Requirements Decomposition

The Big Data Vision Workshop then acquires a small sample (5 to 10GB) of the client’s data. This data is used to build 10 to 12 illustrative analytics use cases to help the client understand the “realm of what’s possible” with respect to applying data science against their targeted business initiative. Simple mockups are also developed to show the business users how the resulting analytics could be deployed within their daily operations. The data science work typically takes 2 to 3 weeks of work.

The Big Data Vision Workshop engagement is capped with a half-day in-person “envisioning workshop” where we leverage facilitation techniques and group dynamics to identify, brainstorm, envision, value, and prioritize the different use cases that support the targeted business initiative (see Figure 2).

Figure 2: Big Data Vision Workshop Timeline and Key Steps

Bottom line: The Big Data Vision Workshop is 100% about driving the customer’s business success with big data.

The value of the Big Data Vision Workshop process may be best understood by viewing the deliverables that come out of the engagement.

Big Data Vision Workshop Deliverables

• Interview Themes. We catalog the business opportunities that come out of the interviews and brainstorming sessions. We then rank the business opportunities on a scale from Tactical to Ambitious (see Figure 3).

Figure 3: Catalog of Business Opportunities or Interview Themes

• Use Cases Detailed Descriptions. We aggregate the interview themes into common use cases (since the interviews can uncover many of the same business opportunities). For each use cases, we capture a description, and then document the business impact (based upon the organization’s business value drivers) and implementation feasibility issues (see Figure 4).

Figure 4: Use Case Description

• Data Assessments. The data sources that come out of the interview and brainstorming processes then need to be assessed from a 1) business value and 2) implementation feasibility (over next 9 to 12 months) perspective. The business value assessment, which is driven by the business users, determines a rough estimate of the value for each data source in support of each use case. The implementation feasibility assessment, which is driven by IT, determines implementation feasibility across a number of data, technology and organization variables such as data accessibility, data cleanliness, data accuracy, cost, granularity, and IT and data science skills/capabilities (see Figure 5).

Figure 5: Data Assessment Matrices

• Data Science Illustrative Analytics. Our data science team spends two to three weeks working with the client’s data to create illustrative analytics in support of the identified use cases. About 10 to 12 different illustrative analytics are created in order to help the business users to envision the realm of what is possible using predictive and prescriptive analytics. This converts a conceptual use case into its possible analytic forms. By far, this is my favorite part of the engagement (see Figure 6).

Figure 6: Data Science Illustrative Analytics

• Mockups. We also create some simple mockups in order to help the business users to envision how the results of the predictive and prescriptive analytics might impact their daily operations. Note: these do not need to be fancy mockups. Heck, most of mine yea look like they were drawn with a crayon. But do not underestimate the importance of helping the business users to visualize how the analytic results might be delivered (see Figure 7).

Figure 7: Illustrative Mockups

• Prioritization Matrix. The capstone deliverable is the Prioritization Matrix. I have written about the Prioritization Matrix before (Guiding the Envisioning Process: Prioritization Matrix Worksheets), but I will reiterate that the Prioritization Matrix is the single most important tool in my kit bag for driving organizational alignment. It is a very simple KISS (keep it simple stupid) tool which brings together the Business and IT leaders to discuss, debate, cajole, scream, and wrestle with where on the Prioritization Matrix to position each use case from a business value and implementation feasibility. The conversations that ensue during the business value and implementation feasibility debate can be the most important discussions to determining the success of the big data project (see Figure 8).

Figure 8: Prioritization Matrix

Big Data KISS Summary

To some, this will seem like a lot of unnecessary work before one starts to deploy the technology. Can’t one simply install the big data technology, load up some data, hire a data scientist or two, and then wait for magic to happen? But as the Forbes article highlights, that approach just does not work. Tick. Tock.

Want to achieve business success with Big Data? Hint: Start with the business! What is the business trying to accomplish over the next 9 to 12 months? Who are the key business stakeholders who either impact or are impacted by that key business initiative? What decisions do these stakeholders need to make in support of that key business initiative? What questions are the stakeholders trying to answer and what data do they need to answer those questions?

It is really that simple. Seriously. KISS

Keys to Big Data Success? KISS!
Bill Schmarzo