If you remember last week’s notes, we’d done some tinkering with the system to see if we could engineer improvements to our stats. So without further ado, this week’s summary is below.
- 1460 calls were offered (that is, received by the system)
- 1075 calls were not deflected (that is, went straight through to an agent)
- 207 calls were abandoned (that is, the caller hung up before speaking to an agent or using the bot)
- 168 calls were partially deflected (that is, the caller used the bot but still spoke to an agent)
- 10 calls were fully deflected (that is, the caller only used the bot)
However, some caveats. The report that this is based on no longer tracks abandoned calls, and we know there is a reporting error for 24 June (corrected in the background). This now includes calls where the caller chose option 1 for the payment line as “partial deflections”, and calls where the caller used the bot but then went to the payment line are “full deflections”.
The deflection rate based on the above is now 12.2%, a massive improvement, but I’m taking that with a pinch of salt as the report did change midway through the week.
What the new, more detailed, report has shown us, however, is that a lot of callers are failing identification & verification (ID&V), where they enter their payment details. Callers to the rent line are more likely to fail than leasehold callers. We suspect that this is because people don’t have their payment reference to hand. We’ve already changed the welcome message to highlight this, and we’ve agreed to switch on the loop counter, so you get two attempts to pass ID&V.
Things are improving on the user feedback front. Three weeks in, we’re getting some useful feedback from callers in response to a short series of questions from agents; people who like the automated options really like it. When asked to give a score out of ten, we’re averaging six – and we’ve already fixed issues that have led to early low scores. We’ve also found a new approach to more in-depth feedback, working with Richard, our lead User Researcher, and that’s already paying off.
The service assessment is next week and I’m looking forward to it. I feel prepared for the questions that will come my way. Wish me luck!