Humm Group boosts agent effectiveness and automates QA in under six months

Buy now, pay later company Humm Group has managed to boost agent effectiveness, automate quality assurance and focus on self-service opportunities in under six months.

The payments company said its used AI-backed conversation intelligence, Operative Intelligence to help with its call centre process.  

Humm Group first signed on in mid-2023 with all models ready for go-live by mid-December the same year.

Brendon Clark, head of workforce and customer improvement, at Humm Group told a Genesys Xperience Sydney crowd there are roughly over 200 agents between its customer service, collections, lending and fraud teams, based across Adelaide, Auckland and Manila.

“We have high volumes, very high contact rates, little visibility on why customers were contacting us and what their pain points were.

“Also, we had a lot of repeat calls as well. That could have been due to not being able to do first-call resolution type things.”

He said once the team saw the insights produced from the speech and text analytics platform “everyone was on board the team and wanted it now.”

Before the work, the team could only see quick call summaries or wrap codes, which were “limited in what they show”.

“Once we could see the solution, we realised we have hundreds of different contact types across our contact centre.

“It was good to see a trend. Previously, when we’d pull a report and see what our wrap records were, we needed to see that over a six-month period to truly understand.

“Being able to see a trend was is really beneficial. Also being able to measure any changes that we did do in the business.”

Clark explained that has the company evolves “we need to make sure we can measure it and make sure that it's a good return on investment.”

Humm Group now has around 1000 unique reasons customers are calling they are able to select rather than the agent themselves.

“You can't challenge it because this is from the customer's mouth. There was a bit of an adjustment for us, but as soon as you go through it, you look at the transcript along with this and it's really clear.”

He added the team pulled raw data from the platform and made sure “they were accurate” and it was categorising them correctly.

“That's how we did our validation. There was maybe a one or two that wasn't quite right and we just worked with [its partners] to get that sorted.”

According to Clark, there’s wide use of the platform, including product, digital and marketing teams as well as workforce management, customer improvements and team leaders.

Its quality assurance analysts also used to overview five calls per agent per month which is roughly about 9 percent of total contacts.

“Obviously, there's a huge amount there that's not being looked at, which could be a compliance risk or not.

“Being able to now look at 100 percent of our volumes and …. it reduces the risk on the business,” Clark said.