Aimie vs Humans
New data report: how Aimie AI measures up to human cash collectors
At Sidetrade, our AI functionality was designed with cash collection efficiency in mind. Customers can choose to treat our AI assistant, Aimie, as a team coach balancing workloads across their collectors and dispensing to-do lists that focus their efforts on the most lucrative cash opportunities; or like another head in the collections team, receiving a share of the workload to maximize team capacity; or as a behind-the-scenes operative carrying out routine dunning tasks. Either way, there’s no doubt that Aimie makes collectors’ lives easier.
This is a bold claim, but well founded. Recently, our data science team have been delving deep into the numbers behind Aimie’s success. Since April 2019, and particularly during lockdown, we’ve compared Aimie’s actions to those of her human counterparts across all of our customer sites – and the data shows that even in times of chaos, Aimie consistently outperforms manual workflows in terms of efficiency, to the point where one of our clients switched all collections activity over to Aimie while their collectors were furloughed – and saw an almost instant spike in overall efficiency. Read on to find out how.
How we define ‘efficiency’
Before we dive into the numbers, we had better explain how ‘efficiency’ is defined according to our metrics. We measure efficiency by counting days it takes for payment to arrive following an action being completed. For the purposes of this dataset, if payment happens within 7 days following either an executed human or Aimie action, the action is considered ‘efficient’.
Aimie’s performance
With this definition of efficiency in mind, let’s look at an overview of the whole dataset. Looking at actions across all of our clients, since April 2019 Aimie has outperformed the actions of human collectors in terms of percentage efficiency. In December 2019, for example, Aimie’s actions were 9% more effective than those of her human counterparts. During the first COVID-19 lockdown period, from March to May 2020, it’s interesting to note that as invoices generally became less likely to be paid, both Aimie and human workflows suffered a drop in efficiency – proving that even our non-sentient digital assistant was not immune to the challenges of the global pandemic. Aimie’s efficiency percentage even dipped below 50% for the first time since we started reporting, to 45% in April 2020. However, it’s worth noting that Aimie was still more efficient than human workflows, which dipped to a low of just 37%). However, from June 2020 onwards, we started to see both Aimie and human workflows return to their usual level of efficiency.
How lockdown impacted collections
The COVID lockdown dataset is also particularly interesting as it applies data to what we’d been seeing and hearing anecdotally: with less available resource for collection due to the pandemic, our clients reduced the amount of manual collection actions executed and increased AI-driven automation.
During lockdown, Aimie’s efficiency dropped by just 4%, compared to a 12% drop in efficiency for the actions of human collectors
In most of the customer sites we surveyed, Aimie continued to perform well during lockdown; one of our customers saw a small drop in Aimie efficiency in April 2020 by just 4 percentage points, while in the same month the drop in regular manual workflow efficiency was a significant 12 percentage points. At the same time the client had chosen to switch many of their customer accounts from workflow to Aimie, making the continued strong performance of the AI even more impressive. This rings true across our entire client base – the volume of accounts managed through workflow dropped significantly from March to June 2020, while accounts managed by Aimie steadily increased.
One client even stopped all manual collection as their collections team was furloughed. Instead they switched Aimie on for all accounts. Following a very short-term drop in efficiency as Aimie settled in, the efficiency of AI grew steadily and outperformed the previous situation with human collectors. This tells us that for companies experiencing overstretched resource in the collections team, using Aimie to automate dunning actions can help preserve efficiency even in times of stress.
Key takeaways
So, what’s next – and how can you use this insight? As we adjust to life post-pandemic, whether your key pain points are a need to reduce collection resources, a desire to increase the coverage of accounts with your existing team or a long tail of unpaid invoices, Aimie can help your collection activity run smoothly. If you assign Aimie to collect cash from your smaller or lower-value accounts, she can work on reducing the backlog while allowing your human collectors to focus on your VIP accounts – without experiencing a drop in efficiency.
In fact, when Aimie gets going you should even see the efficiency improve!
Check out the next blog in the series.
