Being a technology-focused MLS provider for the west coast, TheMLS faced a lot of challenges with their clientele of more than 4,500 brokers/agents (users) being able to use their platform. Although their platform was quite simple and easy to understand - not all users knew how to navigate and use it effectively.
As a result of this, TheMLS’s team of 10 customer support agents faced a never-ending influx of phone & chat support requests, which revolved around answering mundane questions about the platform. Most of these FAQs were already addressed via the documentation section, however, the users found it easier to simply raise a support request instead. This reduced their ability to provide valuable support to users who were facing more pressing issues.
Given the size of their clientele and the aggressive growth that TheMLS was chasing, it was evident that scaling customer support via some level of automation was key.
Aavesh Naykodi, Product Owner at TheMLS took it upon himself to find alternative solutions that helped in preventing the platform users from clogging the support reps' time, and provided a far more robust customer service experience.
Aavesh came across the growing trend of chatbots for answering FAQs, and saw it as a potential solution to their problems. WotNot was approached to devise a solution that could help TheMLS create some breathing room for their support reps, and cut down on time as well as effort spent on commonly asked questions. The solution would also allow their reps to jump in, and work towards high priority issues, as and when required.
WotNot proposed to build an FAQ chatbot + live chat solution which would have a knowledge-base of the most commonly asked questions. Aavesh’s team was using Zendesk at the time to manage chat support. By extracting historical conversations from Zendesk, the WotNot team was able to expand the bot’s knowledge base even further, and make it smarter by training it across different intents and utterances. Having the chat to human handover capabilities built-in, TheMLS could use WotNot as a one-stop solution to managing their chatbot and live chats.
Whenever a user would open the chat interface and ask a question, the chatbot would first understand the intent of their question and see if it matches an existing intent in its knowledge-base. If there was a match, the answer associated with that question/intent would be sent as the response, else, the chatbot would handover the chat seamlessly, to an available support rep based on the chat assignment conditions.
With the FAQ bot, TheMLS was able to provide tier-1 support to its users via the chatbot, and only have their support reps’ jump in on queries that the bot wasn’t trained on. This utilized the support reps’ time effectively and helped streamline their support problems.
The problem did not stop here. Although the bot was trained for hundreds of intents, their clientele, which were in the thousands, had different ways in which they could frame the same question. TheMLS team could now add all those different utterances behind one intent in WotNot’s FAQ builder, and train the bot to correctly answer the previously failed queries.
WotNot provides a better experience for users to learn about how our platform can help their business and, at the same time, makes our support team more efficient in addressing service requests. Our support reps now spend more time dealing with more critical issues, and bring back valuable feedback which helps us in refining the product even further
With the WotNot FAQ chatbot and livechat going live on TheMLS website, in about 12 months, the solution delivered -
1400+ hours of support saved
60% increase in turnaround time to respond
Improved productivity among support reps
Aavesh has now started drawing up plans on other areas of the product, where he envisions the bot would help. Additionally, the FAQ chatbot and the support team will continue to provide proactive responses and improve on the user experience. The goal – a customer support experience that grows and improves with the company.