Winning with Analytics

//Winning with Analytics

Winning with Analytics

Decision-making is changing in companies today as fiercely competitive environments are forcing businesses to do more with less, respond to client demands faster, and provide greater value to customers for a profitable future.  There aren’t any decisions currently being made in corporate boardrooms that haven’t been staged by the data and analysis.

Today it’s not just e-commerce or online firms such as LinkedIn, Facebook, Amazon or using data to compete, but every business across all industries are adopting data and analytics based strategies to outthink and out-execute competitors. The primary driver of the adoption of analytics is due to the fact that once a business has successfully embedded tools and techniques for performance management, quality improvement or process improvement methodologies into their business operations and the rest of its industry follows suit to adopt it in a similar fashion, there is very limited room for differentiation. It’s then that analytics provides a distinct competitive advantage for companies to win in market differentiation by harnessing the analytical intelligence to accelerate their decision-making speed and sophistication.

As businesses size up the analytics maturity model from being analytically impaired to analytical champions, the winners are successful leaders that have infused analytics throughout their business to drive smarter decisions, enable faster actions and optimize the competitive advantage.

From my point of view, the following are the three key differential attributes of businesses with the anatomy of winning with analytics

(1) A clear analytics strategy

Analytical champions that are at the forefront of analytics always have an evolving but clear strategy that identifies the business case it’s solving and the corresponding business value of the investment.www_harphajan_com_business_strategy.jpg

Whether the analytics strategy is to find the optimized pricing for each customer to maximize revenue, calculate customer lifetime value for optimized spend, detect churners and increase their loyalty, or increase network profitability, there are very explicit and actionable business cases underscoring the business analytics strategy.

Analytics winners typically embark on an analytics business case in two stages:

  • Identification of Business Problem
  • Analytics Problem Framing


The primary stage involves first framing the business case and determining whether the problem can be solved by an analytics solution. This includes analyzing the problem statement and its delineating constraints, defining the business benefits of solving the problem, obtaining stakeholder detailed understanding on the problem statement and subsequently determining whether the problem is a candidate for an analytics solution. Each business case will endeavor to demonstrate moving the business benefits from observable to measurable, quantifiable and financial.

The second stage then advances to reformulating the business case into an analytics problem with a potential analytics solution. At this stage, analytical champions develop a proposed set of drivers and relationships to likely outcomes, state the sets of assumption related to the problem, and finally outline the key critical success factor for measuring the analytics solutions.

Once the analytics business case is established and aligned with corporate business priorities, it’s then approved by senior stakeholders for buy-in by the top C-Suite executives.

Analytics champions would subsequently transcend from strategy to execution with an analytics roadmap which constitutes of clearly defined target operating model, and implementation milestones including planning, execution, and post-implementation framework from the perspectives of people, process, technology, data, culture and governance.

(2) Widespread use of modelling and optimization techniques in business problem-solving

Essentially, there are three types of analytics:

Descriptive Analytics: Primarily reporting of past performance or historical events without providing context as to why a certain event occurred or www_harphajan_com_descriptive_analytics.png
probability of the reoccurrence of the event in the future. For example, monitoring whether sales of children’s wear is greater than men’s wear or which segments of subscribers are driving sales growth.

Predictive Analytics: Uses statistical models based on past data to predict
www_harphajan_com_predictthe future outcome and provides explanations for the occurrence of event. For example, forecasting the number of cars sold in a city considering macro economic factors such as inflation, seasonality etc.

Pre­scriptive Analytics: Uses statistical models to provide evidence of the optimal level of key variables to yield a particular outcome, behavior or action.
www_harphajan_com_flight_scheduleA typical example is the aviation industry, where airlines companies schedule the timing
of the flights in order to optimize the utilization of assets and load factors in real time

Most businesses are already performing some form of descriptive analytics for operational or management reporting, which often summarizes certain groupings or simple counts of events, for example dollar sales across categories, inventory levels, average earnings growth rate etc. However businesses that are analytical champions are using predictive and prescriptive modelling to further advance their competitive advantage by looking beyond the obvious.

To illustrate, in the field of customer analytics, analytics champions are harnessing customer analytics not only for improved customer understanding, but also for establishing profitable insights such as contacting only prospects with a high propensity to respond to specific marketing campaigns or cross-sell and up-sell promotions; tailoring retention offers solely to customers who are mostly likely to churn – and not those who would stay anyway; acquiring and retaining only profitable customers over a lifecycle – and not losing customers, among others. For instance, a leading retailer wanted to segment its customers into five main categories based on their overall value to devise marketing

strategies accordingly. The retailer then used the RFM model for analyzing customer value, where RFM stands for Recency of purchase, Frequency of purchase and Monetary value of the spend. Once each of the attributes had appropriate categories defined, segments were created from the intersection of the values. The resulting matrix had twenty-seven possible combinations and based on distribution, these were collapsed further to come up with five main categories of the customers. The resulting segments were ordered from most valuable (highest recency, frequency, and value) to least valuable (lowest recency, frequency, and value)

Analytics champions use complex statistical algorithms and experiments to ensure they have leading efficiencies in their analytics investments.
www.harphajan.com_maths The goal of the analytics model is to develop the best statistical relationship between a dependent variable and one or more independent variables, and subsequently forecast and predict the coefficients for an optimum model fit.

Some of the more popularly deployed analytical models include Regression Model, K-Mean Clustering Model, Markov Chain Model, Apriori Model, Causal Model, RFM Model and Monte Carlo.  www_harphajan_com_einstein6

(3) Embedded analytics in all areas of business processes

Analytical champions, who are at the forefront of harnessing analytics, have successfully embedded analytics into operational and transactional processes to provide their businesses with intelligence in the moment. While companies usually embark on their analytics journey by establishing a dedicated analytics team to analyze historical events in separate applications for insights, analytical champions have fully automated analytics systems, including scoring algorithms www.harphajan.com_stats_1
and analytic-based rules embedded in their core everyday business transaction systems such as ERP, CRM, or other sales, financial and marketing setups.

Embedding analytics in their core business process has not only helped businesses with better operational decisions, but also improves on existingbusinessman point on business process

business processes as they can now react in real time, for instance to changes in demand and supply, thus adjusting capacity to reduce costs and increasing efficiency all at the moment of decision-making.

To set the stage, let’s understand the example of a world-class international package shipment company that has successfully embedded analytics in its business operations from a people, service and profitability perspective. First, from a people perspective; it schedules hundreds of thousands of workers to match the estimated work package

Workers are seen in the warehouse in Milton Keynes, north of Londonloads to specific workers’ skill sets on a daily basis, which fluctuates depending on the amount and specific location of the packages, whether it’s at stations, gateways ramps, hubs, contact centers or sorting centers. Here a multi-stage mixed integer analytics program is used to assign all work tasks by the time of the day, then work is distributed into full-time and part-time shifts and subsequently specific employees are assigned in accordance to their skills and shift schedule.

Second, from a service perspective; on a day-to-day basis, the international package shipment company needs to handle fluctuating volumes of shipments (numbering in the millions) and ensure its delivery routes not only meet demand, but also expand and reduce coverage based on volume changes. Recently the company also allowed flexibility to its customers to change pick-up, drop-off location and delivery timings on its website on the same day to accommodate for the boom in e-commerce. At the heart of balancing the workload and optimizing the routes for delivery couriers is a heuristic-based
DHL expressvehicle routing analytics systems which optimizes delivery routes based many variables in play, including promised delivery times, different customer types, different packaging types, the number of stops for each delivery route, traffic and time constraints, and customs clearance all in real-time.

Third, from a profitability perspective;

www_harphajan_com_profitabilitywith close to 100,00 package cars, vans, tractors, motorcycles and over 500 aircraft fleet, transporting multi-million packages across the world on a daily basis undoubtedly poses an astronomical challenge in optimizing profitability for the international shipment package shipment company. Indeed, it must not only manage its aircrafts and fleets of vehicles, but also balance the costs of fuel, facilities (including expensive airport facilities, distribution warehouses), equipment and operations support. Capitalizing on every cost-saving opportunities whilst managing the inherent complexities are the advanced techniques of operations research, which deploys myriads of optimizing techniques and machine algorithms
dhl2to provide operations insights into all aspects of operations performance (service quality, operational efficiency, operation compliance) and support operational business planning (working capital, resource and fleet planning etc.).

Analytical champions who have successfully embedded analytics seamlessly now have significant competitive advantage over their competition. This is because they are assured of optimized decisions at all levels of the organization in all aspects of business functions including operation excellence, business planning, process and systems engineering, quality & performance management, and customer service, among many others.

Championing Analytics

Today we live in an era of extraordinary opportunities amidst fiercely competitive environments. Those businesses that desire to rise up and stay ahead of the competition in the years to come will be required to embrace new ways of thinking and innovative technological advancements. One thing is clear; analytics is the new way of thinking and the extraordinary technological differentiator to win,oscar
…and I believe we’re just getting started.

By | 2017-09-08T22:06:27+08:00 November 3rd, 2014|Categories: AI Blog|Tags: |2 Comments

About the Author:

Being an Innovation Evangelist, Harphajan is a well-established expert in leveraging Artificial Intelligence and Data Science


  1. schreef November 11, 2014 at 6:48 pm - Reply

    I like the helpful information you provide in your articles. Ill bookmark your weblog and check again here regularly. I’m quite certain I will learn plenty of new stuff right here! Good luck for the next!

    • Smithc330 November 11, 2014 at 6:48 pm - Reply

      Indeed. Very insightful sharing

Leave A Comment

Leading the charge towards AI

Leading the charge towards General Purpose AI with Reinforcement Learning as a first step that can survive in a variety of environments(instead of being tied to certain rules or models)

Innovating with AI

Building new equation for future AI innovation, one in which business disrupts and spurs a great leap forward