Cisco Systems (Ind) Pvt Ltd - Innovative Use of Technology in Customer Service

Company: Cisco Systems (Ind) Pvt Ltd, Bengaluru, Karnataka, India

Category: Award for the Innovative Use of Technology in Customer Service > Computer Industries

Nomination Title: SentiMeter – Real Time Customer Sentiment

Essay:

Executive Summary:

Knowing our Customer's real time pulse has been an age-old problem. Feedback survey like NPS and customer escalations are all lagging indicators of customer experience and we come to know of a problem, once the damage has been done. So, the question is: Can there be a better way to gauge sentiment? We found our answer in Machine Learning.

To address this, Cisco CX researched in Artificial Intelligence, machine learning and natural language processing. We built a model that has learnt from over a 140000 Service Requests having feedback survey and can now predict sentiment real time with an accuracy of ~82%. The tool provides a number of views with actionable insights useful for support engineers, managers, accounts teams and sales stakeholders to improve customer loyalty and retention.

Challenge & Opportunity:

Today's customer experience with Cisco is solely transactional and does not cater to historical/past experience. The reason for it is divergent data sources and lack of converged interface. Add to that, at the moment, Cisco only relies on surveys like NPS to judge customer sentiment which are lagging indicators. This means that damage has been already done and there is very little scope left to recover. This tends to limit our scope of understanding our customer profile and the problems that they face as a whole. This incurs costs and huge manpower efforts later. It also impacts any potential Sales and Business opportunities. We did competitive analysis only to find that there is no other company with this type of approach/model. All the products available in market are reactive in analysing customer experience Refer Sentimeter1 picture

Action Taken: There is no existing solution that can proactively predict customer experience real time and also give a 360-degree holistic view of the customer's network.

'Sentimeter' is our answer to this problem.

Using a Neural Network algorithm, the model learnt from over a 140000 Service requests and derives the key influencers of experience. It generates a point in time health score which gives specific user actionable to be addressed. It also gives a technology /product level score card for the products deployed on the network thus helping identify the key areas to be focused in product quality. This equips our sales team to be better prepared to answer some unforeseen customer questions arising during a sales opportunity. It also generates engineer profile which, along with customer profile, can help in routing the case to right engineer next time. Refer Sentimeter2 picture

Resulting Impact:

Cisco's mantra has always been customer loyalty and customer retention. In this competitive world, where customer experience is the forefront of every business, having a holistic customer profiling and sentiment analysis model will go a long way in ensuring that Cisco's customer experience always stays ahead in the game. This will elevate the existing service support levels from a reactive to a proactive model. This will help Cisco build a greater customer affinity. It will enrich customer experience which will lead to customer loyalty and retention. This will also lead in reduction in escalations. More importantly, the focus will now be on customer ownership rather than transactional interactions. A 10-point increase in customer loyalty could create ~$1.2 billion incremental revenue for Cisco."