NelsonHall recently visited WNS’ Analytics CoE in Bangalore. Research & analytics (R&A) services are key activities for WNS, as a first point of entry into new clients, and also as a higher margin activity in existing accounts. Readers of NelsonHall’s Quarterly Updates on WNS will be aware that WNS is looking to increase the level of annuity revenues from analytics contracts. Our visit gave us an opportunity to catch up with the Analytics Practice and learn about future focus areas.

Analytics domain experts with strong sector focus

WNS’ analytics practice has over 2.5k resources dedicated to the company’s six core verticals, developing specialist analytics offerings in marketing, CX, campaign and loyalty management, and risk & compliance. WNS offers both Analytics-as-a-Service and embedded analytics within its BPS services; in both cases, the focus is on co-creating solutions which weave domain expertise with the tools and models to drive business impact for the client’s specific requirements.

WNS’ proprietary framework WADE covers all analytics activities, from identification and management of data sources to data ingestion & storage, data preparation & processing, data consumption, model building, to the business use. Key WNS proprietary IP/products include:

• Brandttitude: marketing ‘storytelling through data’ platform
• SocioSeer: social media insights, sentiment, and benchmarking platform
• InTouch: social engagement platform with customized responses and prioritization
• OutReach: campaign management platform
• Unison: omni-channel integration platform with category-level insights, sentiment analytics
• Aptymize: WNS’ omni-channel personalization engine
• Commercial Planning Suite: integrated revenue analytics platform for airlines.

Many of these tools have been developed to address a specific client need, bypassing ISV solutions that do not offer the required domain depth and customization flexibility. WNS then expands the use cases to additional clients or new verticals. For example, the Brandttitude platform was initially implemented for a CPG sector client, mapping marketing and brand performance for product managers, but now also has a version for the entertainment industry, comparing artists’ true value and competitive analysis, and is currently being customized for an insurance client.

Infra to insights

In a world where BPS providers have lots of experience in building data lakes and accumulating data across structured and unstructured data sources, a key differentiation is in translating big data into meaningful insights and then operationalizing these insights on behalf of the client, or drawing them to the attention of the relevant stakeholders. For example, in order to maximize ancillary revenue performance for airlines, WNS has developed a library of propensity models that run on airline ticketing data and are able to predict the likelihood of passengers to buy a specific ancillary product.

These models can be further applied on new ancillaries and bundled offerings, creating an ancillary revenue enhancement plan. The results, when applied to the loyalty database, can also be used to identify members who might be interested in redeeming points for ancillary services. This can help in improving the redemption rates of loyalty members with low engagement levels. WNS claims clients have been able to predict high propensity passenger segments with ~3 times the current conversion rates.

From personalization to hyperpersonalization

WNS’ approach with CoE analytics enables it to both address industry-wide challenges and also to go deep into client-specific issues. For example, it developed a personalization engine to help a global high-end hotel chain in their marketing campaigns. The engine segregates customers based on their lifetime value, identifies target segments, creates offers (drawn from 65 different types) for each customer within a segment, and finally estimates the impact of marketing incrementality. This is often in parallel to the models that the client has developed.

WNS runs all five stages of campaign planning for the client, from data gathering, to model building, through to the campaign hyperpersonalization. The campaign analytics services of WNS has been validated as ‘best in class’ by Wharton’s Customer Analytics Initiative. WNS helps with the campaign deployment and monitors performance. The engagement started in 2009 and has delivered nearly half a million dollars in cumulative incremental revenue for the hotel chain.

Proactive analytics to drive CX interventions

Many clients’ requirements with analytics services are evolving from issue identification to issue prevention. Companies are looking for insights on their customers’ future behavior: for example, to be able to map customers’ propensity to post feedback on social media, or predict delinquency or fraud risk.

To support an Australian insurance company wanting to both proactively identify large loss claims and reduce processing time, WNS deployed insurance SMEs specialized in catastrophe modeling, supported by an experienced assessor. Using dimension reduction techniques, WNS selected 71 features to identify ~25% more large claims within 30 days of registration, reducing the average age of large loss claims by ~30%.

What next for WNS in analytics?

A key priority this year and next for WNS is accelerating the use of AI, and this applies to R&A services as well as to BPS service delivery. WNS has trained more than 250 members through its Mastermind training program on AI/ ML. The team is looking at areas such as deep learning and image recognition. For the latter, WNS has a pilot with a P&C insurer using drone images to perform analytics on damage estimation, also automating the identification of risk factors.

WNS’ analytics practice is beginning to reap the benefits of investments in skills development. In 2016 it established a two-year Business Analytics MBA program with NIIT University; the first group of grads from the program are now integrated with the practice.

So what else should we expect? WNS is expanding its partnerships with AI platforms with NLG/NLP capabilities in multiple languages and is eyeing further M&A opportunities, specifically around cognitive analytics, AI, ML, and smart data discovery. This M&A activity will follow a ‘string of pearls’ approach to cater to specific sub-segments and in areas such as marketing and risk analytics.