The Business World is Transforming
  • By 2025 the worth of the Internet of Things will be $6.2 trillion.
  • The sharing economy will reach $330 billion by 2025.
  • For people starting their education, 65% will enter the workforce into jobs that don’t exist today.
  • The average tenure on the S&P 500 is dropping. Only 25% of the companies in 2012 will remain by 2023.
  • Automation and robotic usage will grow 2,000% from 2015-2030 amounting to $190 Billion market.
  • 86% of global CEO’s are championing digital transformation of their companies.
  • By 2025, half of world’s companies with revenues exceeding $1 billion will be headquartered in today’s emerging markets.
  • By 2018, the data created by the Internet of Things will reach 403 zettabytes a year.
  • By 2030 the population will be over 8 billion people and 50% of Global GDP growth will come 440 cities in emerging markets.
  • By 2030 more than 30% of workforce will be older than 55 in developed countries.

The Biggest Myth Preventing You from Deriving Value out of Your Data

The Biggest Myth Preventing You from Deriving Value out of Your Data
04/16/2018, Lu Hao, PhD , in Change Management

In the previous article in the analytics series, we introduced the next disruptive innovation in data and analytics and the importance of getting ready for it. But how? In this article, I will share what I believe to be the biggest misunderstanding about launching your data and analytics effort. I will also introduce the “Question-to-Value” approach to help set you on the right track for deriving value from your data faster.


No trend in the past 10 to 15 years has created as much buzz as big data and analytics; these initiatives have climbed to the top of the C-suite agenda for many organizations. According to the Harvard Business Review (https://hbr.org/2012/10/big-data-the-management-revolution), companies that inject big data and analytics into their operations show productivity rates and profitability that are 5% to 6% higher than those of their peers. And Forbes found that analytics have been successfully used to improve pricing, and promotions to optimize 6% or more increase in revenue. (https://www.forbes.com/sites/adamtanner/2014/03/26/different-customers-different-prices-thanks-to-big-data/#731bfeba5730).

However, skepticism and hesitation abound. Leaders often convince themselves that their organizations simply aren’t ready for the change. The most common remarks I hear from senior leaders include: “We are not there yet”, or “A better timing for analytics will be after we accomplished our data warehouse project”, or “We need to build our BI team before getting into advanced analytics.”

All these comments are founded on one single myth: the idea that a data and analytics-enabled business value must be accompanied or even preceded by a fully upgraded and implemented data infrastructure (data warehouse, data lake, database…, you name it).

This is the biggest misunderstanding surrounding the big data question. And whoever realizes it and adjusts his or her mindset accordingly, will win big.

True, data infrastructure is important for the long-term health of your analytics program, but infrastructure is not an indispensable ingredient for you to generate business value out of your data. If you are waiting for a perfect data warehouse to start on analytics, you’ve probably already deployed a process that is driven by data availability, which often means using whatever data that’s available to cook a data soup and see what problems bubble up. This is a lengthy approach, and it often leads to random correlations rather than meaningful insights. Therefore, a very likely scenario is after spending piles of money on data-warehousing programs, you then invest on powerful analytics programs analyzing all your data but are still unable to yield any insights that can be put to use.

This is why data should serve the purpose, not drive the process. In other words, data analytics is not a crystal ball into which you dump all the data and ask, “what does my data tell me to do?” (see previous article Predictive Analytics Ain’t No Crystal Ball). Instead, the right question to start with should be “what business issues I need to solve that could potentially be solved using data and analytics?”. With this change of mindset, you are shifting from a data availability-driven process to a business-driven process, where you start by thinking about the desired business results. This mindset better fits the agile company needs to create value from data and to create it fast. You should tie analytics tightly to your biggest value drivers and largest pain points, and focus on how to use data to make better decisions.

To help implement this business-driven process, the “Question-to-Value” approach (below) provides a pragmatic solution. This approach involves six key stages to help set you on the right track to start using your data and gaining business value in no time:

  1. Question: Too often we find that companies launch analytics or big data efforts without a clear view of what exactly they want to accomplish, which results in a solution that is not tied to a business problem. Asking the right question is half of the answer. Therefore, the best way to start the process is to first identify the business question, or issue that you wish to answer/solve through data and analytics. The question often does not emerge naturally. It takes deliberate effort of identifying, understanding, and focusing on the main business drivers. It is also a collaborative process where analytics project leaders need to communicate with key business leads and stakeholders to gain a holistic view before deciding on the right question.
  2. Data: With the right question in mind, you should then identify the necessary data to solve the puzzle. This process includes examining the data collected in your organization and select only the relevant segments of data: a specific time range, a handful of factors that have a direct impact on bottom line and a set of business levers you could manage to pull. In addition to the internal data source, it is also worthwhile to explore external data sources available that might contain key influencers, such as customer demographics data.
  3. Analytics: The next step is to choose the right analytics models that could best deliver practical insights. The model designer should have mixed background in IT and business. This hybrid role needs to understand the basics behind predictive modeling, as well as the types of business judgments made in the day-to-day operations. Conversations with frontline managers will ensure that analytics and tools complement existing decision-making processes, so that the ultimate goals can be met.
  4. Insights: Analytics modeling in step 3 reveals hidden patterns in data, which may or may not be useful to the business. The next step is to interpret these patterns into business insights. Findings should be shared with business lead(s) who understand the day-to-day operations to be framed into a business context. And based on their feedback, the analytics models might need to be fine-tuned and re-ran to derive the most impactful business insights from the data.
  5. Action Items: The business insights discovered in step 4 need to be developed into customized action items. The business lead(s) should identify the feasible action items to address the insights, and these insight-driven action items should be aimed at the goal/expectation identified in step 1. Analysts are also responsible for developing hypothetical scenarios to quantify the expected improvement from different action items to help with the final decision making.
  6. Value: The final step is the execution of the action plan. This means getting the insights into the hands of your frontline employees who will ultimately realize the value from data through day-to-day operations. For this, the business needs an adoption strategy, both for short-term implementation and long-term organizational culture transformation.

Analytics will take root faster only when it is tied directly to business outcomes. Therefore, taking the “Question-to-Value” approach is critical because it puts the focus where it should be: on tying analytics directly to outcomes, taking action and delivering value, in an agile way. It provides a value realization mechanism that helps move organizations from a data-based mindset to an outcomes-based mindset and minimize needless or unproductive analytics efforts. It also helps break down the organizational barriers that impede information sharing by setting clear goals and expectations up front.

When you leave this page, start writing down the single, most pressing question that you have with your organization. Then, go through the “Question-to-Value” chain. Sooner than you expect, you will see opportunities to derive true value from your data rather than see a misguided use of R&D funds on the latest business fad.

Already thinking about what’s next: acquiring talent, deploying tools, or building your analytics team? In the following articles we will touch upon all these aspects and help you better prepare your analytics strategy. Stay tuned.

Lu Hao, PhD

Data Scientist and Management Consultant, MSS Business Transformation Advisory

Dr. Hao is a data scientist with comprehensive experience in data analytics, and complex data visualization. She has experience conducting data-driven analyses to find and develop solutions for clients in both the federal and public sectors. Dr. Hao earned both her Ph.D. and M.S. degrees in Transportation Engineering from University of California at Berkeley, Berkeley, California. MSSBTA

Leave a reply translated

Your email address will not be published. Required fields are marked *