From Insight to Payoff: 4 Ways to Speed Data’s Time to Impact

by   |   November 20, 2014 5:30 am   |   0 Comments

Today, the promise of big data programs is out of sync with the speed of driving real dollars from big data. According to a Gartner report published in September 2013, 64 percent of enterprise organizations have made or are planning to make investments in big data, but only 8 percent have started using big data to make decisions.

Business leaders are frustrated with longer-than-expected delays. Analytics practitioners struggle with driving, evangelizing, and consuming insights. How does one compress the time from insight creation to generating dollars? How can one do analytics at the speed of real business needs?

The following four catalysts can help businesses accelerate time to impact.

  1. Be Greedy About Consumption

In the last few years, organizations have competed on the creation of analytics – that is, the findings and insights that can be gleaned from a robust analysis of data. As we look forward to increasing analytical sophistication, creation of analytics will be commoditized. Organizations either will use external partners to resolve their problems, or machines and tools will become sophisticated enough to extract insight from data. In this future world, consumption of analytics – that is, the tangible actions companies take based on insight – will become the competitive differentiator.

Related Stories

Big Data ROI: Does the Payoff Match the Potential?
Read the story »

5 Strategies for Recruiting and Training Decision Science Talent.
Read the story »

Webinar: Maximize ROI and Accelerate Adoption of a Successful Cloud Strategy with Workload Analysis.
Read the story »

Webinar: Improve Real-time Processing and Insight with a Hadoop Operational Data Lake.
Read the story »

Lack of focus on consumption is one of the primary reasons that the potential of decision sciences has not been fully realized. Organizations need to be greedy about consumption and ask themselves a few questions that will heighten the importance of consumption early in the process:

    • How will the analysis planned be consumed? (This should be asked early in the process.)


    • Which stakeholders should be involved in the creation process? (Which will help with buy-in?)


    • How should stakeholders be involved in the creation process? (It’s important that they don’t slow the process.)


    • Are the end-users ready and trained to consume the analysis?


  • What learnings from this problem can be applied to other problems?


Consider a typical marketing problem: calculating the lifetime value of a customer. In many organizations, the tendency is to jump into the data and run a variety of models to get a number for each customer. At this point, step back and ask the business questions around how the results are going to be consumed: Are you planning to change the messaging for different customer segments? Are you using Customer Lifetime Value (CLTV) to understand the effectiveness of past campaigns? Are you using it to forecast the impact of future campaigns?

The actual analysis required for each question could be very different. Creating a model based on assumptions can lead to a lot of wasted time and effort, and result in inappropriate decisions that could lead to the opposite impact that the business wanted to create.

When the analysis is complete, measure the results of the implementation and derive key learnings that can be applied to a similar problem in the future.

  1. Capitalize on the Entire Decision Supply Chain

As global businesses shift their focus from the physical to digital, the manufacturing supply chains of yesteryear are being replaced by decision supply chains. Issues in the manufacturing supply chain find parallels in the decision supply chains. For example, instead of trying to avoid a product stock-out in a physical supply chain, you want to avoid a decision stock-out in the digital supply chain.

The manufacturing and decision supply chains. Click to enlarge.

The manufacturing and decision supply chains. Click to enlarge.

To translate big data into big decisions, businesses must be able to appreciate and capitalize on the different aspects of the decision supply chain. Organizational muscles must be built across the continuum. For example, does your decision sciences team have a consistent and clear method to define problems? Is there a systematic process to generate insights and communicate them in the organization? How do you measure the success of analytical initiatives with different ROI windows?

  1. Appreciate the Interconnected Nature of Problems

In today’s networked problem space, very rarely does a solution to a single problem drive impact. Solutions to connected problems have a greater chance of driving impact and being consumed.

To transform big data into big decisions, map out the interconnected problem network at the beginning of the problem-solving process. Building this inter-connected map up front helps uncover connections previously not seen in the problem universe, leading to a more holistic solution. For example, the insights from an analysis on setting the price of a new online product based on feature usage also may be used to determine which features are effective and identify gaps that need to be addressed.

Organizations that spend time and effort connecting problems will have a higher probability of breakthrough insights. Connected insights like these play in a huge role in developing a coherent consumption strategy. Each of these insights plays a part in successful implementation.

  1. Develop an Integrated Mindset, Skillset, and Toolset View

Most organizations focus on the analytical skillset (What do employees know?) and toolset (What tools can employees use?), with relatively little emphasis on mindset (What do employees believe?). In many cases, the gap between the promise and the results of big data lies in the mindset of the organization.

Organizations must look for certain traits during hiring and then promote a culture to further foster these traits:

    • An appreciation for learning over knowing


    • A belief that questions are more important than answers


    • A desire to constantly experiment and learn from the experiments (both successful and unsuccessful)


  • An appreciation for interdisciplinary perspectives


The winners and losers of the big data world are yet to be determined. The results will hinge on companies that are able to rapidly translate big data into big decisions. Organizations that build a strong consumption bias, appreciate the decision supply chain paradigm, explore the interconnected problem space of analytics, and integrate a healthy mindset, toolset, and skillset will emerge as winners.

Rajat Mishra is a seasoned business leader with 15+ years of experience in management consulting (McKinsey), technology (Microsoft), and business intelligence (Google). At Mu Sigma, Rajat is leading transformative change in decision-making processes across various clients in the San Francisco Bay Area. Rajat holds an MBA from the Wharton School of Business at University of Pennsylvania, where he was on the Director’s List, and a Bachelors in Technology from IIT Delhi, where he was on the Dean’s list.

Neethi Mary Thomas, Account Manager, Mu Sigma – Neethi is an experienced analytics consultant with more than five years of demonstrated accomplishments in the decision sciences industry. She has been recognized for her strong analytical and management skills, serving as a trusted thought partner in data driven decision-making and mentoring teams to drive performance.

Gaurang Patel is Engagement Manager at Mu Sigma.

Ankur Uttam is Senior Engagement Manager and Client Partner at Mu Sigma.

Subscribe to Data Informed
for the latest information and news on big data and analytics for the enterprise.

Tags: ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>