Last year we blogged about how Wootric prioritizes people, product and process to build products that our customers love. We also emphasized that focus of next 12 months would be to add more intelligence to our product offerings and briefly talked about the new AI-powered enhancement to our product, Wootric CXInsight™.

Focus on Machine Learning has paid dividends for our customers.

We are very happy to tell you that our product offerings have indeed become more intelligent. Through CXInsight’s proprietary sentiment analysis and feedback classification, our customers have imported over 300 data sets to analyze 2 million+ customer feedback comments spanning both B2B and B2C ecosystems. To our surprise, most of this feedback originated on competitor survey platforms like Qualtrics and review sites like GlassDoor. This further validates our initial hypothesis that CXInsight™ would give our customers cutting edge insight in Voice of Customer data regardless of how that data was collected. While most of this feedback has been collected through surveys, a significant portion came from support chats, social media and phone surveys. More than half of our CXInsight customers are bringing two or more types of feedback (e.g. survey responses and online reviews) into the platform.  We are helping them move toward the CX holy grail: unifying all of the “voices” of the customer in one analytics platform.  

Insights from Wootric have given CX pros the support to gain cross-functional alignment behind improving the customer experience. Product roadmap prioritization, customer journey optimization and improved employee engagement are some of the diverse goals our customers are tackling with the power of machine learning.

We developed an NLP solution that is uniquely powerful

One of the main reasons behind this success is our pragmatic approach to leverage machine learning (ML) and natural language processing (NLP). We realized very early that building a generic platform that provides turn key insights to our customers across different industries would be difficult and not valuable simply because products and services across different products categories have different concerns. Instead, we have created industry specific ML models to surface targeted insights.

For example, if you are an e-commerce company then you are likely interested in understanding comments related to shipping and packaging, product quality, checkout process, returns, etc. These concerns will not be applicable to a developer centric SaaS/PaaS products — where topics are often fall in categories like UX, UI, performance, bugs, feature request, etc. Here is a screenshot of an example customer dashboard. As you can see, we not only classify feedback into different categories but for each category we also surface sentiment.

A last point about our sentiment analysis. Initially we used Google Cloud for Sentiment Analysis and experimented with AWS service as well. We realized that while they are pretty good, their analysis had several false positives. Our customers rely on accuracy of our analysis so we can not afford to have a big margin of errors. No machine learning system will give 100% accuracy but by narrowing our focus to CX, we have created our own sentiment analysis model which beats Google’s by 4.6%.

A more robust, connected survey platform

While our R&D efforts focused on analytics, our core survey platform has been growing — both in terms of volume and in new features. Here are some new features, many built as a result of our own customer’s feedback:

• New integrations with HubSpot, Gainsight and Totango
• Our Salesforce integration addressing how handle Contact and Lead records, as well as account level reporting, for Salesforce deployments with thousands of accounts and contacts.
• Wootric surveys inside the Intercom Messenger app.
• Triggering surveys from Salesforce and Zendesk with richer metadata to easily segment your customers.
• Our Segment.com integration now supports all three surveys — NPS, CSAT and CES — and surveys can now be triggered based on events.
• Segment Source: Wootric data can now show up in your data warehouse such as AWS Red Shift, Google BigQuery, Snowflake via Segment.
• Targeted sampling for in-app surveys. This feature allows you to set up a different survey strategy for various segments of your customer base. No code required.
• Of course, GDPR compliance. 

Our already big data platform continues getting bigger. In the past year, we have surveyed 110%+  more people than previous three years combined, 150%+ growth in number of customers using our bi-directional Salesforce integration, 600%+ growth in number of API calls. This is truly an “up and to the right” kind of chart. 😉

What’s next

In 2019, look for us to provide deeper insights from your data and new ways to facilitate reporting and action.  Stay tuned — we have big plans for boosting customer happiness in 2019!

Prabhat Jha is CTO at Wootric. Follow him @prabhatjha