Large Account Receivable Company was spending most of their time chasing up overdue payments from their clients. Most clients had no direct debit in place, other clients with direct debits had insufficient funds in their bank accounts, causing the automatic payment to fail.
It was taking at least 6 hours per day, for two employees, to gather the list of customers that had defaulted or were late in paying their bills.
After compiling the list of clients who had overdue accounts, each individual was contacted either via a phone call or by letter. This was very time consuming, costly and inefficient.
In a few occasions, the company incorrectly contacted the wrong customer and that caused lawyers to be involved to fix the mistake.
We adopted the crawl-walk-jog approach.
Firstly, we automated the process to retrieve the list of clients who were behind in their payments. This preparation process cut down the time from 12 hours effort to a few seconds, which meant that instead of running campaign only twice per week, the company was able to retrieve a list of clients to chase up, every single day.
When clients missed a payment, a SMS message was promptly sent to their phone.
We also put in place another SMS functionality, where the clients receive a text, as a reminder, before their payment is due. This helped to reduce late payments and direct debit failure.
The next step was to create an automated voice tool for customers to be able to enquire about their accounts at any time. This helped to automate approximately 85% of all inbound handling that was previously done via the company’s employees. We used voice sentiment to route the calls appropriately.
The data collected was also used to profile customers in different groups and their preferred way to be contacted, either SMS or phone call.
We were demonstrating this technology capability at one of the conferences that we regularly run and this company’s management representative asked us to help them with their problem.
We built on the same technology of machine learning, firstly dealing with cost savings and then moving to product revenue. This was a low-cost solution to start up and only costed the price of the phone call or SMS.
The solution was created using Google API (Machine Learning) voice to text, and Upwire, which plugs into Google API Machine learning.
We did some research and understood that emails were not the solution, and that text message was a better solution.
These are the capabilities that we were able to implement:
- Integrated the solution and expanded its use to other areas of the business
- Created automated voice tools that proactively call and collect payments
- Created automated negotiation tools via voice and SMS
- Conversational SMS workflows to ensure customers can ask about account details anytime of the time
- Automated 85% of all inbound call handling
- Through AI and machine learning we created a profiling solution
- Recognise when and by which medium to interact with the customer
- Use voice sentiment to route calls
- All interactions are recorded, transcribed and scored allowing for better training and customer handling
- All data is fed back into the engine to learn and evolve
From start to finish, we worked on this project for about six months.
We were able to roll out a basic functionality first. One week to get the first one up and running, we created demonstration workflow and we thoroughly tested to make it completely reliable and be able to determine the sentiment or frustration of the customer.
First functionality SMS, then expending the solution to other areas of the business and adding voice tools, and adding extra uses throughout the company, customer can enquire for their account details, last one is using machine learning to profile the customer and identify who will use which channel.
The results achieved with our solution drastically improved their daily operations, empowered the business by simplifying complex technologies, reduced default payments and increased this company’s revenue from more product sales.
We were able to save our client 4 million dollars in six months, and these were the main benefits produced by our solution:
• Automated list cleanser – reducing human error and time
• Customers defaults campaigns could be run every day and not only twice a week
• Massively reduced late payments and debit failures and debts were recovered quicker
• Reduced cost from having to contact each customer either by phone call or letter to remind them about the late payments and direct debit failures
• The solution automated 85% call handling
• Customers engagement became more successful for both parties and customer satisfaction increased
• Revenue growth from selling new products
• The business soon became more innovative in other parts of their operations and products – transforming their thinking
• Competition advantage
• Empowered a business by simplifying complex technologies
• Drastically improved their operations
• What started small has grown HUGE We then expended the flow to other users in the business