From chocolatiers to jazz club owners, more than seven million customers rely on Intuit QuickBooks to track every last penny that their business makes. An essential tool for bookkeeping, payroll and everything in between, QuickBooks is expected to deliver fast, engaging customer support 24/7.
But as the speed and frequency of support requests increase, the answer is not necessarily more customer support agents. In fact, QuickBooks is filling the gap by working smarter in Slack, the channel-based messaging platform.
‘We wanted to utilise the power of artificial intelligence (AI), Slack and our product experts to drive a more robust support ecosystem,’ says Omer Khan, former Digital Lead for Customer Success at QuickBooks. It was also important to increase the company’s net promoter score (NPS), an indicator of customer satisfaction.
- Increased agent confidence by 20% overnight
- Solved customer support cases 36% faster, saving 9,000 annual agent hours
- Boosted its NPS score by 12% in about six months
We had a chance to chat with Khan during Slack Frontiers, our annual conference, about how this Slack-driven approach opens up new learning opportunities, democratises knowledge within QuickBooks and boosts team morale.
Using Slack and Quincy bot to resolve support cases 36% faster
Support teams had been using Slack for a while, but there was a lack of structure. Agents were working across multiple applications to find the information that they needed to resolve a single issue. Data was easily fragmented, managers lacked insight and agents were dissatisfied. Plus, there was no central place to recognise teamwork.
‘We wanted to make Slack a one-stop shop, where answers to every question would help agents work more effectively and efficiently,’ Khan says.
To do this, they turned to Quincy bot, a custom bot built by QuickBooks that leverages AI to answer agents’ questions. Early this year, Quincy was resolving 10% to 15% of questions. Then they integrated the bot with Slack.
‘At any given time, agents can ask Quincy a question right in Slack,’ Khan says. With a growing knowledge base and machine learning algorithm, Quincy’s value is quickly increasing.
‘Now, Quincy consistently answers 60% of questions,’ Khan says. ‘That’s an incredible amount of savings in terms of repetition and what we had to do to answer the same question again and again.’
Cases used to take two days to solve, but Quincy has helped to bring that down by 36% to 1.2 days. ‘If you put that annually, we saved our agents 9,000 hours,’ says Khan.
‘It used to take us two days to resolve a case. With the help of Quincy bot in Slack, we’re now 36% faster. If you put that annually, we saved our agents 9,000 hours.’
Saving customers 2,000 hours with a searchable knowledge base
When an agent can’t find an answer with Quincy, they create a ticket using the atSpoke app for Slack. This is then assigned to a subject-matter expert, referred to as a product champion at QuickBooks.
‘The product champion takes 60 to 90 seconds to ask questions, clarify the need and provide support in Slack,’ Khan says.
Not only does this solve that agent’s problem, it builds on a searchable knowledge base of step-by-step solutions, speeding up time to resolution. Management can easily identify content gaps for both the customer who originally got in touch and for the agent working on the case.
QuickBooks takes this information and holds regular content hackathons to evaluate the question backlog, hack select cases and provide solutions. This newly expanded content is fed into both Quincy bot (to serve agents) and QB Assistant (the customer-facing bot).
With an evolving library of data-driven knowledge at their fingertips, customers can get back to business faster. ‘We’ve saved customers 2,000 hours by giving them the right information at the right time,’ Khan says.
Increasing agent confidence by 20% and NPS by 12%
When it comes to his team’s learning and development, Khan recognised that ‘10% of people learn through formal training, 20% of people learn from peers and 70% learn through experiential or on-the-job learning.’ Instead of disrupting agents by pulling them out of the queue and into day-long training, he turned to his Slack ecosystem to increase customer-facing teaching moments.
When an agent simply hasn’t asked a customer the right question to solve a case, a product champion sends a direct message in Slack or has a Slack call with the agent for one to two minutes, offering a proactive solution for the next time that a similar case comes up. Both parties have full context, so these sessions are especially efficient.
‘Hundreds and hundreds of questions come in every day and create these micro-coaching opportunities, which drives our skill set to a level we’ve never seen,’ Khan explains. This has a powerful effect on team morale and confidence, which Khan measures on a regular basis.
‘Before, our overall agent confidence helping customers in the product was about 60%,’ Khan says. ‘After the Slack ecosystem, we instantly saw a 20% jump to 80%. This is a huge win because confidence is contagious.’
At the end of the day, increased agent confidence, faster issue resolution and a streamlined workflow add up. ‘After we implemented Slack, we saw a 12% increase in our NPS,’ Khan says. ‘We’re effectively driving results for both agents and customers.’
To learn more about how Slack is enabling better customer support around the globe, take a look at these other stories.