10 Predictions For Analytic Decision Support by 2015

Aug 30, 2010 (10:08 AM EDT)

Read the Original Article at

With many businesses still struggling amid anemic economic conditions, increasing numbers of companies are focusing on analytics-driven decision support across all facets of their businesses. Leaders have defined their competitive edge on the foundations of data-driven decision making while others are struggling to play catch up.

This article presents ten predictions about changes in decision science that will have the greatest impact on business over the next five years. Developed by Mu Sigma, the analytics services firm I founded and head, these predictions touch on trends that will challenge conventional wisdom, drive new business models, and enable organizations to differentiate themselves to compete and drive growth.

Conventional wisdom suggests that organizations need to evolve from Descriptive Analytics to Predictive Analytics, but I encourage the concurrent use of math, business and technology decision sciences across what Mu Sigma calls the DIPP Index:

D for Descriptive Analytics -- What happened in the business (using reports and dashboards)?
I for Inquisitive Analytics -- Why something happened in the business (using analyses)?
P for Predictive Analytics -- What will happen in the business (using predictive modeling)?
P for Prescriptive Analytics -- the So What, Now What?

As companies adopt analytics as the new science of winning, the future of analytics will not just be based on applied math, business and technology, as it is today. In the future, decision sciences will make use of Math + Business + Technology + Behavioral Economics, as outlined below.


  • Business + Technology allows us to simply automate
  • Math + Business allows us to present more cogent arguments in the boardroom
  • Math + Technology allows us to anticipate and operate proactively
  • Math + Business + Technology allows us to execute better
  • Math + Business + Technology + Behavioral Economics let us develop nudges (cognitive repairs) against human biases.

I'll explain in more detail by describing our top-ten predictions for decision sciences. Some of these predictions are obvious and some are relatively obscure. The relative importance of each prediction depends on the unique needs and business drivers in your industry and the analytical maturity of your organization.

Prediction 1: Hyper Competition Will Proliferate

The evolution of decision sciences will lead to hyper competition. Organizations will engage in information arbitrage games to exploit information asymmetry against the competition. In analytically mature industries, information advantage will be a swiftly moving target due to the intense competition. Airlines and financial services have historically pioneered the use of analytics as an integral component of their business models. Today, many airlines are unprofitable because they can't maintain a competitive edge due to the swiftly moving information arbitrage in an intensely competitive marketplace. On the other hand, some financial services firms have gained a formidable edge over competitors in high speed algorithmic trading by exploiting cutting-edge models.

To unleash the true potential of analytics, companies will have to move beyond the traditional frontiers of pricing, marketing and risk, using analytics to compete on innovation and relatively obscure areas of the business. Companies in all industries, be it banking, insurance, retail or airlines, will have to innovate in supply chain management and new-value creation to stay competitive on price and quality, and to generate enough capital to outlast competitors.

Prediction 2: Companies Will Compete on Consumption of Analytics, Not Creation of Analytics

Analytics will become more commoditized, with organizations competing on consumption of analytics rather than creating analytics from scratch. Context and relevance will be the key enablers for effective consumption of analytics. Organizations will have to focus on various competencies to get consumption of analytics right. The consumption cycle (shown below) will necessitate a more balanced combination of right-brained and left-brained thinking.

Analytic Consumption Cycle

The increasing use of intuitive technologies such as visual analytics, interactive dashboards and decision-support simulation tools is indicative of the trend toward consumption of analytical insights. Flex-, Silverlight- and Ajax-based Rich Internet Applications are enriching these user experiences. Companies such as Tableau Software, Tibco Spotfire and Qlik Technologies have pioneered visual analytics as a means to give business users quick insight into data. It is noteworthy that Qlik Technologies recently released an IPO and has a P/E ratio of 80 and a market cap valuation of around $1.5 billion on revenues of only about $150 million.

Prediction 3: Incremental Innovation Will Be Routine, Disruptive Innovation, Rare

The premium for truly disruptive innovation will increase as we see more continuous, incremental innovation. The speed of innovation is typically inversely proportional to the kind of innovation. Disruptive innovations tend to take longer to release and have a longer shelf life. Incremental innovations can be developed more quickly and are easier to replicate.

Many companies will adopt incremental innovation strategies taking advantage of investments made in data assets. The truly disruptive innovators will focus on creating the right mix of qualitative and quantitative analytics by delving into new areas, such as behavioral economics and social anthropology, to create and execute on innovative ideas.

Historically, we have seen the value of disruptive business models increase exponentially from Hotmail (online email) to Yahoo (online portals) to Google (search-based online advertising) to Facebook (social networking). Decision sciences will make it easier for companies to deliver continuous innovation, thus placing a premium on creativity and truly disruptive innovation.

Prediction 4: New Data Sources Will Emerge

New data providers will emerge focusing on intelligently interpreted data -- especially for social media analysis, location data, etc. Multiple Scores (such as FICO for credit, HICO for health, and so on) will be created to better understand human behavior. Web 2.0 business models exploiting network effects will lead to an exponential increase in user-generated data (with FourSquare and Twitter being current examples). Smartphones will monitor our vital signs and chronic conditions, yielding a wealth of data for disease prediction and medical research.

User-generated content will empower levels of prediction that were never before possible. "If we look at enough of your messaging and your location and use artificial intelligence, we can predict where you are going to go," recently commented Eric Schmidt, CEO of Google.

An increasing proportion of analytic analyses will rely on unstructured data such as text and video. Social media data (Radian 6, Viral Heat and so on) and RFID data will be merged with location information for supply chain analytics. Similarly, text mining will be used for social media analytics, customer voice data for customer-service and call-center analytics, and video data of customers’ emotional responses to products/store layouts, traffic monitoring and so on for retail analytics.

Prediction 5: The Role of 'Chief Analytics Officer' Will Emerge

The CAO or their effective equivalent will look override organizational and departmental boundaries to build data-driven competencies wherever information-driven decisions are made. The expertise of professionals from varied disciplines such as technology, applied math, anthropology, behavioral economics and so on would be applied to decision making. To tackle supply-side issues, for example, globally distributed, location-agnostic teams and partnerships with the required talent would be formed.

Companies such as Emsense (neuroscience-based marketing), Q Interactive (online ad network), Adenyo (mobile marketing) and Equities First Holdings (securities lender) have already appointed Chief Analytics officers from backgrounds as diverse as human behavior, economics and computer science. Going forward, even business analysts would need to morph into content engineers exploiting the synergies of left-brain and right-brain thinking to generate more holistic insights.

Prediction 6: Analytics Education Will Be Formalized

University of Ottawa, North Carolina State University and DePaul University among others have started offering formal degree programs in analytics. As companies use analytics more and more, they will recognize the need to create and develop talent and will provide support for universities to build formal degree programs in analytics. Mature organizations will augment formal education with their own corporate analytics programs that will bring in domain knowledge and business context.

Prediction 7: Process Automation Will Take the Forefront

Certain areas of analytics, such as marketing mix modeling and trade promotion analytics, will move from the research/discovery phase to become mature, repeatable methodologies. These methodologies will be “operationalized” with automation technologies, enabling analytics at a greater level of detail and scale than previously possible. For example, we'll see marketing effectiveness analyzed by product, SKU, channel, country etc. However, each company will need to assess the applicability of the automation paradigm to its analytic challenges; there is a potential danger that pure automation will supersede necessary human intervention and oversight.

Prediction 8: Open Source Analytics and Analytics-as-a-Service Will Gain Adoption

Open Source analytics platforms will emerge to increase adoption and better mine the wisdom of crowds. Open source R has already emerged as a leading platform for statistical innovation and collaboration both in academic and industry circles. Adoption is evidenced by commercial vendors of R, such as Revolution Analytics, focusing on scaling the R computing language. There is also a trend of collaboration between proprietary and open source platforms. For example, Tibco Spotfire and SAS now offer the option to call R functions or scripts within their environments. This signifies a shift towards loosely coupled, open computing platforms.

Analytics-as-a-Service will be commonplace and will take different forms, including analytics services providers, analytics focused SaaS companies, existing IT services, system integration and data providers moving into value added analytics services. Outsourcing and global delivery will yield significant supply side benefits to the analytics industry.

Big players such as IBM and Accenture have already turned their attention to analytics. IBM, for example, has acquired SPSS and it opening analytics centers in India, China and elsewhere. There is an increasing trend of companies in the Fortune 500 issuing analytics RFPs. Analytic Software-as-a-Service models are being deployed in many areas, including Web analytics (Google Analytics and Adobe Omnitur), marketing analytics (M-factor), and hosted/on-demand business intelligence platforms (Panorama, SAP BusinessObjects On Demand, etc.).

Prediction 9: Behavioral Economics Will Gain Traction

Behavioral economics is increasingly being applied in the corporate world. While analytics helps us gain insights and make decisions, those decisions can create a positive impact only if human biases are taken into account during implementation. Analytics insights will challenge many traditional ways of working. Executives championing data-driven decision making will need to leverage an interdisciplinary approach using business + technology + mathematics + behavioral economics + social anthropology. As this becomes a formal practice in corporate operating procedures and corporate strategies, better understanding of human biases would help develop cognitive repairs or nudges to ensure better application of decisions.

Behavioral economics embodies principles that explain the workings of the brain. It helps in developing ways to influence behaviors. For example, CVS Caremark recently changed the choice architecture on an online renewal form to increase the rate at which patients renew their medicines. Behavioral economics concepts, such as choice architecture, can be combined with consumer insights gained through analytics and applied to targeted marketing campaigns, conversion rate optimization, portfolio optimization and so on.

Today, more and more organizations are investing in dedicated consumer insights teams. They focus on developing segmentation strategies, understanding customer lifetime value, market sizing and so on as independent initiatives. Over the next few years, companies will move toward a more holistic approach to understand their customers.

Prediction 10: Convergence and Collaboration Will Be the Rule

As new business models emerge and the boundaries of the value chain are blurred, a new era of convergence in the use of analytical techniques and frameworks will come into play. Cross-industry and cross-domain learning will lead to significant breakthroughs in the development and deployment of analytics solutions.

New business models are accelerating the need for convergence and collaboration. As an example, we've seen Microsoft enter retail, cellular phone networks enter the Netbook category, Dell moving from custom configurations to prebuilt offerings. Scottrade and Yahoo are collaborating on data and analytics to optimize lead generation for Scottrade.

Certain analytical techniques have historically been developed and used mostly in specific domains, with examples including yield optimization used by airlines, survival models used in Life Sciences, Lean principles used in manufacturing, and diversification used in finance. However, these techniques have a strong potential to be used across industries. For example, you might use yield optimization methodologies for online advertizing, survival modeling concepts for financial risk analysis and diversification for marketing and supply chain optimization.

Summing It Up

Analytics is going to play a key role for business in the coming years. In the wake of the economic crisis, the world is shifting to a "new normal," and we are seeing the advent of a new customer psyche. Institutionalizing analytics is not a destination or a goal for your company; rather it is a continuous process of internalizing and integrating analytics into the decision-making process. This will be a key differentiator for successful businesses and the route to both incremental and disruptive innovation.

Dhiraj Rajaram is the founder and CEO of Mu Sigma, an analytics services firm that helps companies institutionalize data-driven decision making. Before founding Mu Sigma, Rajaram was a strategy and operations consultant at Booz Allen Hamilton and PricewaterhouseCoopers. For more information, visit or write him at