Generative AI benefits contact centres in several ways, including:
- Reducing costs: Automating tasks such as answering frequently asked questions (FAQs) or routing calls to the appropriate department and freeing agents to focus on more complex tasks.
- Improving efficiency: Providing agents with access to real-time information and suggesting solutions to common problems.
- Improving customer satisfaction: Delivering quick, personalised responses and shorter wait times for customers and improved customer satisfaction.
AI is undoubtedly powerful, but we need to tread carefully.
I’ve conducted a quick experiment to see how helpful current AI (June 2023) could be to a Contact Centre Manager by asking:
“In a contact centre with 20 agents, if I use AI reduce the number of calls by 50% how many agents do I now need?”
ChatGPT replied: “If you reduce the number of calls by 50% using AI, the number of calls handled by agents will be 50% of the original number of calls. Therefore, you will need half the number of agents to handle the same volume of calls.
So, if you started with 20 agents, you now only need 10 agents to handle the reduced volume of calls.”
ChatGPT’s response was quick and apparently definitive, but the given answer is WRONG. The screenshot below applies ‘traditional’ Erlang calculation to a scenario requiring 20 agents:
If nothing changes other than reducing the contacts by 50% then the answer is 11 agents NOT 10.
And, as already stated, one of the key AI benefits is to free agents to focus on more complex tasks. So, in the real world the 50% of calls that reach agents are expected to have a longer average contact length. Let’s estimate a 20% increase in contact length and the number of agents required increases to 13.
In summary AI is powerful but isn’t threatening the Contact Centre Manager’s job (yet!)
P.S. it’s interesting to note that whilst ChatGPT answered the question definitively (but wrongly) Bard was a little more tentative.
