Chat and MarTech Integrations Drive Better Customer Experience, and Increase Sales
As many readers know, Udacity is a leader in the online education industry. Udacity has trained over 10 million students worldwide in over 160 countries. They also partner with over 50 enterprise clients and 200 industry partners to encourage high-quality employee training on the latest technologies, processes, and concepts.
When Scott K. Wilder joined Udacity last year as their Head of Lifecycle, Growth and Performance Marketing, he was tasked with global customer growth, acquisition, and user engagement across Marketing channels and the customer journey. On the Marketing systems side, Udacity was in a slightly “Wild West” state — Scott saw many opportunities to improve business processes, create company wide-dashboards, and improve revenue stack integrations.
Scott initially focused on installing effective MarTech systems, integrating across systems, and putting necessary processes in place. On the MarTech side, Udacity invested in several vendors, including Intercom, Blueshift, Salesforce, Amplitude, Calendly, and Segment:
MarTech System Evolution
Rather than try and boil the whole technology stack “ocean” at once, however, Udacity focused on one or two key MarTech areas every month. Trying to boil the whole tech stack ocean at one time can overwhelm effective implementation. Instead, Scott leveraged a Minimal Viable Product (MVP) approach, learned, and then scaled after 15 – 30 days of testing. Too many companies, he told me, try to launch with every use case addressed. As he learned from his 10 years at Intuit, focus on the critical few use cases first.
On the integration side, the Udacity team ensured that specific “hub” systems had timely access to all relevant data. For example, the Marketing team gained access to a customer’s identity and Website activity prior to an Intercom chat, so the conversation could be as personalized and relevant as possible. Imagine someone greeting you and saying “I noticed you were looking at our AI courses. Are you an AI Developer or AI Product Manager?” The Salesforce system gained access to this Web data as well as chat records, so other customer interactions such as phone calls, triggered emails, etc., could be highly customized.
To stitch all these systems together, the team leveraged not only Segment.io, but also Fivetran to ensure that Marketing data is uploaded into AWS Redshift. There, analysts and executives can query their data, and develop dashboards in Chartio to deeply understand Udacity’s business and customers.
Udacity MarTech Integrations
On the process side, the Udacity team first addressed the low-hanging fruit, and got the most out of their Intercom instance by mapping automated bots and live chats to the acquisition, engagement and retention parts of the funnel. Scott’s team also instituted stringent SLAs for response times, so every customer would receive a chat response within 2 minutes. Finally, they then took the big step of centralizing customer conversation feedback, and integrating all customer conversation channels, including Intercom and Zendesk, into one place. Udacity created a SWAT team that would regularly search through all customer conversations, read them verbatim, and conduct analyses. This team could then monitor feedback changes over time, receive alerts when anomalies appear, and monitor the overall pulse of customer interactions.
Intercom Chat Inbox, with Qualification Details
With these modifications, chat has been a big success at Udacity. Chat is used throughout each stage of the customer lifecycle now, from the initial meet and greet, to classroom discussions, to career services. In addition, each customer segment has different chat messages during their respective customer journey. Chat has been so successful at Udacity in large part because younger generations are accustomed to communicating this way. It is also relatively easy to scale, and enable Sales, Marketing and Customer Success to reach more people on a real-time personalized basis in less time. As a result, Chat is driving a significant part of the company’s revenue.
Customer Chat About Potential Course
It hasn’t all been easy, though. There are challenges to successfully implementing chat systems as well. For example, it’s highly important to set customer expectations appropriately. Udacity cannot always respond to a customer’s chat question in real-time, and without proper expectation setting, that could result in a poor customer experience. Misset expectations around real-time responses can also result in customers posting the same chat multiple time. Geographic sensitivity is also important – the tone used, British English vs. American English spelling, etc.
Udacity has realized tremendous gains with its chat and MarTech systems efforts. Scott would emphasize the following lessons learned for other MarTech practitioners who want to introduce chat:
- Focus on profitable parts of your Website first. If you can quickly improve customer conversions, revenue, customer satisfaction, and/or other measurable metrics, your initiative will gain broad support.
- Centralize feedback from customer conversation channels, like Intercom and Zendesk. This will help you consistently measure the “pulse” of your customer interactions.
- Conduct Deep Dives into issues by searching through all customer conversations and reading them.
- Roll out Integrations Slowly: The Segment.io/ Website, Salesforce, and Fivetran/ analytics integrations have added a lot of value. But don’t boil the ocean, or you will get overwhelmed. Focus on smaller integrations initially that you can quickly complete.
- Leverage Automation: Finally, create custom bots so that customers can still interact with chat after hours.
Finally, Wilder believes improving chat should be thought of as a learning journey. Each conversation offers an opportunity to understand your customers better.
