Home » Nova »

Big Data Value Creation Drivers

 

One of the key challenges IT organizations face in building support for a big data initiative is to
ensure that the big data initiative is valued by, or of value to, the business stakeholders. The key to the big data envisioning and value creation process is to understand the “big data business drivers.” There are four big data business drivers  that can be applied to an organization’s key business initiatives or business processes to provide new insights about the
business (across customers, products, operations, markets, etc.) and improve decision-making. Let’s consider each of the four big data business drivers.

Driver #1: Access to More Detailed Transactional Data
Access to more detailed, more granular, structured (transactional) data enables a higher degree of fidelity to the questions that the business users are trying to answer and decisions that the business users are trying to make. For example, what types of questions could I answer and what decisions could I make if I had the ability to access and analyze more detailed transactional data, such as point- of-sale (POS) transactions, call detail records, radio frequency identification (RFID), credit card transactions, stock transactions, insurance claims, and medical health readings?
Access to more detailed transactional data is probably the “lowest hanging fruit” for most organizations—to take advantage of the transactional data, sometimes called “dark” data, they already collect. Due to today’s technology limitations and data warehouse cost factors, most business users only have access to a limited amount of data that is supporting their operational and management reporting. However, big data technologies provide the ability to access and analyze all the detailed and granular transactional data. Access to all the detailed transactional data can fuel the creative juices of the business user to ask more insightful “what if” questions, such as:

  •     What’s the potential business value of supporting more localization by developing forecasts and product plans at the product category, store, and department levels?
  •     What’s the potential business value of supporting more seasonal decision-making by developing marketing plans at the customer segment, product, ZIP code, and season levels (such as Christmas, 4th of July, Valentine’s Day)?
  •     What’s the potential business value of triaging claims patterns at the policy type by rider, claim, week, or day levels?
  •     What’s the potential business value of making network capacity and utilization forecasts based at the ZIP+4, local events, and season levels?

As you can see, the potential to analyze your existing transactional data across multiple dimensions of the business—dimensions like location (specific outlets, branches, or stores), product, day of week, time of day, holiday, customer behavioral category, customer demographic category, and others—and at a lower level of granularity can dramatically improve your organization’s ability to uncover actionable and material business opportunities.

Driver #2: Access to Unstructured Data
The ability to integrate the growing volumes of unstructured data with your existing detailed structured transactional data has the potential to radically transform the types of insights that can be teased out of the data. The unstructured data can provide new metrics and dimensions that can be used by the business stakeholders to uncover new insights about your customers, products, operations, and markets. For example, what is the potential business impact of having access to internal unstructured data (such as consumer comments, e-mails, physician notes, claims explanations) as well as external unstructured data (such as, social media, mobile, machine, or sensor generated)? The business users could leverage the new metrics, dimensions, and dimensional attributes gleaned from unstructured data sources, coupled with the detailed transactional data, for finer-fidelity, more complete analysis and decisions to answer questions such as:

  •     What’s the business potential of leveraging new insights about my customers’ interests, passions, associations, and affiliations (gleaned from social media activities) into my customer acquisition, maturation, and retention business processes?
  •     What’s the business potential of adding sensor-generated performance data into your manufacturing, supply chain, and product predictive maintenance models?
  •     What’s the business potential of integrating third-party unstructured data sources (local weather, economic news, local events) into hospital or doctor office staffing and patient care decisions?
  •     What’s the business potential of integrating social media data into your claims fraud analysis to identify unusual claim submissions among communities of linked individuals?

Driver #3: Access to Low-latency (Real-time) Data
The ability to provide real-time (or low-latency) access to data is a game changer that can enable new monetization opportunities. The biggest problem with today’s batch-centric data platforms is that many customer and market opportunities are fleeting—they appear and disappear before one has a chance to identify and act on them. Think, for example, of the business potential of location-based services to communicate to your customers in real-time as they are making their buying decisions. In other words, what is the business potential of having real-time or low-latency access to key data sources and business metrics in order to shrink the time between when a customer, product, or operational event occurs and when that data is available for analysis and decision-making? What is the business potential of being able to update your customer, product, risk, and operational analytic models on-demand, based on immediate business, market other external events events (such as your favorite baseball team winning the World Series, financial markets jumping 4 percent in a single day, or a destructive hurricane predicted within the next two days)?
Access to low-latency, real-time data can fuel the creative thought processes in the timeliness of the types of questions you could answer and decisions you could make, such as:

  •     What’s the business potential of making customer acquisition, predictive maintenance, or network optimization decisions on a minute, hourly, or “on-demand” basis?
  •     What’s the business potential of updating analytic models on-demand based on current market, economic, or local events (e.g., weather, traffic, concerts, professional football game)?
  •     What’s the business potential of continuously updating fraud detection models based upon unusual activities amongst a social group of users?
  •     What’s the business potential of updating hospital staffing and inventory requirements based on local health incidents or disease outbreaks?
  •     What’s the business potential of updating your distribution schedules and planned deliveries based on current weather, traffic conditions, and local entertainment or sporting events?

Driver #4: Integration of Predictive Analytics
The integration of predictive or advanced analytics into key business processes holds the potential to transform every question that business users are trying to answer, and every decision they are trying to make. This is really about introducing a whole new set of verbs to the business stakeholders—verbs such as predict, forecast, score, recommend, and optimize. These new verbs can help business users envision a whole new set of questions to ask around the potential business impact of predicting what could happen, recommending a specific course of action, or forecasting the impact of different decision scenarios.

Integrating predictive analytics into key business processes and getting business users to embrace these new verbs can produce more predictive answers to key questions, such as:

  •     What’s the business potential of leveraging predictive analytics to optimize network operations, marketing spending, and staffing decisions?
  •     What’s the business potential of leveraging predictive analytics to predict the financial impact of pricing, route, or supplier changes?
  •     What’s the business potential of leveraging predictive analytics to score customers for fraud, retention, up-sell, and likelihood to recommend (LTR)?
  •     What’s the impact of scoring patients for treatment response and readmission likelihood?
  •     What’s the business potential to score partners for quality, delivery, and service reliability?
  •     What’s the business potential of leveraging predictive analytics to forecast network loads (based on economic conditions and local events) or forecast the performance of new product introductions (based on consumer sentiment and product category dynamics)?
  •     What’s the business potential of leveraging predictive analytics to recommend next best offers to improve customer satisfaction, customer retention, or preventive patient treatment?

Big Data Business Drivers: Customer Micro-segmentation Example
An example relevant for B2C companies that are interested in increasing the effectiveness of their customer engagement and marketing initiatives. For example, organizations can move from just a few customer segments to thousands of customer micro-segments by leveraging the customer and product insights that are buried inside the multitude of unstructured customer interactions. From sources such as consumer comments, call center notes, e-mail threads, and social media postings, organizations can gain powerful insights into customers’ interests, passions, associations, and affiliations that can dramatically improve the relevance and performance of each of the customer micro-segments. This will enable more targeted customer interactions via more focused marketing campaigns campaigns against these more granular customer segments.

For this example, the targeted business initiative would then be:
Customer Micro-segmentation: Increase the number of customer segments in order to improve customer profiling, segmentation, targeting, acquisition, maturation (cross-sell and up-sell), retention, and advocacy processes.
Driver #1
What is the potential impact on the targeted business initiative from having access to more detailed and granular transactional data? This could include the following:

  •     Integrating detailed POS transactions with market basket, customer demographic, and behavioral data to create customer micro-segments based on demographics (age, gender), behavioral categories, geography, product categories, and seasonality
  •     Augmenting customer micro-segments with third-party customer data (from the Acxioms and

Experians of the world, plus digital management platform data from providers such as BlueKai and nPario) to include income levels, wealth levels, education levels, household size, psycho- demographic data and online behaviors
Driver #2
What is the potential impact on the targeted business initiative from having access to new sources of internal and external unstructured data? This could include:

  •     Mining social media data to create richer micro-segmentation models based on customers’ social insights including interests, passions, associations, and affiliations
  •     Leveraging mobile data (from smartphone apps) to create geography- or store-specific micro- segments

Driver #3
What is the potential impact on the targeted business initiative from having real-time, low-latency data access? This could include the following:

  •     Recalculating customer micro-segmentation models immediately after “significant” events such as the Oscars, the Olympics, or severe storms
  •     Updating customer acquisition up-sell and cross-sell (next best offer) scores and propensities daily while customer marketing campaigns are still active

Driver #4
What is the potential impact of predictive analytics—predict, forecast, score, recommend, and optimize—on the targeted business initiative? This could include:

  •     Using predictive analytics to score and predict the performance of the highest-potential customer micro-segments integrating POS transactions, market basket, customer loyalty, social media, and mobile data
  •     Using cross-media attribution modeling to optimize media spending across the highest potential customer segments
  •     Recommending best micro-segments to target given a particular campaign’s audience, product awareness, and sales goals

Business users need to answer these types of questions in order to:

  •     Uncover new revenue opportunities that impact their marketing and sales organizations
  •     Reduce costs in their procurement, manufacturing, inventory, supply chain, distribution, marketing, sales and service, and support functions
  •     Mitigate risks across all operational and financial aspects of the organization’s value chain
  •     Garner new customer, product, and operational insights that they can use to gain competitive advantage over their competitors and extract more profit from the industry

What has changed with big data is how you can leverage new sources of data and new analytic capabilities to answer these key business questions to uncover new insights about your customers, products, and markets. For example, the question about our “most important customers” previously was determined by identifying the customers who bought the most products (take your product sales data, sort in descending order, and the customers at the top are your most important). Then the “most important customers” question was answered with our most profitable customers (integrate sales, returns, payments, product margin, call center, and commissions data to calculate your most profitable customers). Nowadays, the “most important customers” question is answered with our most influential customers (take data from a multitude of social media sites to determine the range of influence and level of advocacy for each customer, and aggregate the profitability of each of their friends to calculate the “range of influence” profitability). As you can see, as new data has been made available, the level granularity at which you can answer the most important, valuable, and successful types of questions has been taken to the next level of fidelity, but has also increased the complexity of those answers almost exponentially.
Big data enables you to answer those questions and make those decisions at the next level of detail in order to uncover new insights about your customers, products, and operations, and apply those new insights to answer those key business questions such as the most valuable, most important, and most successful at a higher-fidelity and in a timelier manner. Big data fine-tunes and accelerates your ability to identify the specific areas of the business and specific business processes where big data can deliver immediate business value.

 

Related Posts

  • No Related Posts