Cognitive Analytics

6
Cognitive Analytics Reimagining the future of Business in the Digital Era

Transcript of Cognitive Analytics

Page 1: Cognitive Analytics

Cognitive Analytics Reimagining the future of Business in the

Digital Era

Page 2: Cognitive Analytics

1

The status quo Cognition is the mental action or process of acquiring knowledge and understanding through thought,

experience, and the senses. Cognition leads to formation of perceptions and intuitions that leads to knowledge.

The evolution of cognitive computing has passed through the stages of pure academic setup to intelligent

commercial applications like Watson, Numenta at enterprise level and voice assisted applications like Siri and

Google Now.

Such Applications are sensory or mobile based (Apple Siri), contextual and dynamic. Enabled by scalable

processing power and associative memory such event driven applications emulates human behavior. These

applications utilize context based hypothesis to generate associative insights using machine learning and

natural language processing. Cognitive analytics combines the sophistication of natural language processing

and machine learning to provide probabilistic responses to user queries based on contextualized information.

The evolution of analytical data processing follows a path correlated positively to Business Data volume and processing speed of enterprise

business applications

Cognitive Applications are the next step forward towards the future of prescriptive analytics. While in

predictive and prescriptive analytics insights are derived by processing structured and unstructured data

sources available across multiple data sources, cognitive computing seeks to add a level of consciousness to

the process of data crunching. This is driven by massive processing power as it involves processing of

peta/Exabytes of data spread across multiple disparate data sources to generate actionable insights by

eliminating spurious co-relations in a narrower limit of confidence. The massive amount of structured and

unstructured data is characterized by volume, variety and velocity but keeping in mind the usability of analytical

insights veracity is the prime dimension that dictates the applicability and utility of derived business insights.

This is where the effectiveness of cognitive applications come in, extracting the trustworthy insights by

eliminating deviations on either side of the data spectrum.

Page 3: Cognitive Analytics

2

The Industry The industry is rapidly evolving and an empowered consumer is the one deciding the success of organizational

strategy. The enterprise has to rapidly adapt to understand the pulse of the customer by utilizing analytical

models that describe, predicts and prescribes the consumer behavior to ensure proactive responsiveness.

Surveys suggest that organizations that have historically been trying to understand the customer sentiment

and behavior have achieved highest sales growth and profitability in the long run. Amazon is a prime example

of a firm who has traditionally focused on comprehending customer behavior.

But the stakes are increasing, competitive rivalry driven by availability of substitutes and increasing bargaining

power of the consumer has increased price pressures and customer churn rates. Surveys in e commerce

domain suggest that a 1% decrease in CSAT scores roughly corresponds in 50% increase in probability of

defection among the top 5% customer who contribute to an average of 40% of a firm’s revenue.

The challenge lies in assuming a critical market position that enables us to optimize price and demand, enhance

customer loyalty and identify advocates of a brand that would enable us to scale up operations by mitigating

risks arising from customer churn that threatens to bring down the growth potential of a firm.

Retail/CPG, banking, healthcare are some of the sectors that have traditionally leveraged customer experience

management to enhance loyalty and advocacy to drive revenue growth. These sectors have always been the

early adopters of disruptions in the customer experience management.

At present these industries are in a state of flux. Traditional business processes are at a strategic inflexion point,

customer relationship management has evolved from mere solving of queries to emulating and simulating

consumer behavior to achieve near real-time results. The new era of CRM will be predominantly dictated by

intelligent applications like that of IBM Watson which will not only help you identify the best consumer

portfolio for revenue growth but at the same time they will help you gather real time customer behavior and

business insights.

Cognitive Computing Architecture

Page 4: Cognitive Analytics

3

The cognitive model is driven by a highly scalable processing platform that can be cloud based or on premise.

It supports the language processing, data modelling and analysis functions that can be accessed by enterprise

or business users via mobile, web platforms and enterprise interfaces. The core of the analytics engine is driven

by an ETL engine that integrates and transforms disparate data sources with an additional block for analysis,

hypothesis and qualification of integrated data. The user interacts with the analytical engine via a UI that can

be driven by structured queries or ad-hoc queries driven by form based interface applications

Challenges to adoption There are various challenges that may possibly be constraining adoption of cognitive technologies at the

present context

Technological Readiness

An enterprises technological readiness will define the smooth adoption and contextual fitment of

cognitive technologies in the business processes. Sophistication of underlying platforms will play the

major role in this context. There is a need for interconnectivity and interoperability among existing

business processes to make seamless integration of data sources a reality. While a well-defined data

integration strategy will actually facilitate adoption but at the same time, having the flexibility to re-

organize and re-orient business processes to suit business transformation needs is equally important.

Leadership

Survey data suggests that most of the enterprise leadership is not ready to take the leap of faith due

to lack of awareness and rigidity to break the status quo. Lack of active projects and apprehension

about capital intensive projects have hindered budget allocation and enterprise buy in at the

leadership level

Strategic Fitment

Cognitive technologies are not a generic fit for all industries, there is a need to tailor solutions as per

industry requirements and business strategy. Healthcare, insurance, retail and banking are the primary

sectors that can possibly benefit this disruption in the current state. These industries have traditionally

been investing billions of dollars in predictive analytics to reap long term business benefits and have

been achieving ROIs across multiple high investment projects.

Collaboration

The best of the world class research in cognitive analytics are being carried out in conjunction with the

world’s top technology institutes. IBMs Watson has been a product of collaborative effort with MIT,

CMU and other US universities. Enterprises need to have a productive partnerships with such

institutions to foster creative thinking and co-innovate.

Page 5: Cognitive Analytics

4

Industries and Use Cases

Healthcare

This sector is characterized by widespread use of analytics, research and data crunching. The

healthcare industry generates thousands of GBs of data every moment. More than 70% of this data is

unstructured and disaggregated. Healthcare compliance practices dictates the record of data in a

particular standard but the reporting practices vary from place to place. It is not an industry that is

defined by static business rules and regulations. The healthcare industry has traditionally been volatile

and there are multiple instances where Enterprise data mining projects have proved futile because of

the rapidly evolving business rules. This sector is perhaps the perfect fit for adoption of cognitive

technologies keeping in mind the fact this would allow the combination of structured and unstructured

records to gain insights ranging from patient healthcare, drug research and even healthcare insurance

claim processing. Cognitive analytics is already being used for cancer research to identify patients at

risk and carry out effective treatment based on associative insights.

Retail

Retail sector has been effectively leveraging analytics to improve customer loyalty and promote

advocacy to enhance revenue growth. The online as well as bring and mortar retail has utilized

advanced analytics to merge structured consumer data from POS and unstructured social data to

generate consumer behavior insights that can be used for targeted promotions. This has helped retail

majors both in online and offline space to reduce churn, increase footfall, improve shopping

experience and increase Customer life time value in the long term. The solutions range from self-help

kiosks, virtual in-store purchase advisors, real time behavior based offers. Lost sales opportunities

contribute to revenue losses close to $600 Billion worldwide. Intuitive cognitive solutions can help

solve the problem of customer defection and improve sales. In retail such applications can be

leveraged to improve customer loyalty by adding a touch of personalization and contextualization in

case of real time offer management.

Banking and Financial Services

Recent trends in adoption of analytics in banking suggests a tendency to adopt predictive technologies

to identify high quality investment opportunities, quantify implicit risks and fine tune investment

portfolios. Cognitive digital advisors enable safer investment and reap maximum benefits by providing

specific investment advice in near real time scenario. Development bank of Singapore has turned to

IBM’s Watson to revamp its analytics capabilities in the Consumer Banking and Wealth Management

division. The aim is to integrate corporate customer data and unstructured data available in social

domain to enhance its CRM effectiveness and optimize investment returns. The technology aims to

focus on improving the more creative and human interaction elements of business processes.

-------------------------------------------------------------------------------------------------------------------------------------------------

-----------------------------------------------------------------------------------------------------

----------------------------------------------------------

Page 6: Cognitive Analytics

5

----------------------------------

Abhishek Nandan

Consultant

Business Tech Consulting, Brillio

[email protected]

Mobile 9900912161