When we first started x.ai we had this vision that Amy and Andrew would be “fire and forget it” scheduling assistants that you only had to interact with once. Our design philosophy was to reduce the numbers of emails we sent you. Meaning, we felt that any time we emailed you to ask for clarification that we had somehow failed you.
I’m not so sure that is right. In fact, now I’m confident that it’s wrong.
What we’ve now introduced is this idea of “Decision Points” in certain settings, ones where we might not be able to immediately deliver on what you asked Amy and Andrew.
I want to highlight two Decision Points today:
- Constraint Ignored
- Silent Guest
This is a Decision Point that gets sent out when a customer asks to schedule during a specific time frame, but we can’t find any times within that because of their calendar, preferences, etc. Previously we would start proposing times for the next available time, but now we go back and ask them what they would like us to do rather than assuming.
In terms of Data, we’ve seen about a 10% improvement in the schedule rates of these types of meetings.
This is the exciting one as this is a decision point that gets sent out in multi-participant meetings (meetings with more than 1 guest) when we are waiting on the last guest to accept a time. Before a Silent Guest could stall out a meeting and it would ultimately end up cancelled or reconfirmed at a later time.
Now, we’ll proactively check in with the host to let them know that one of the guests has been silent and we’ll ask if the host if they want to make that guest optional. In this setting, we have seen an impressive 378% improvement on the schedule rates of these meetings. Which is 4x!
We learned that one should not be so married to their original design principals that they are unwilling to change them. When you do change them, go measure them and imagine the optimization you can squeeze out of any system.
Be sure to try out the new features of Amy and Andrew and don’t hesitate to shoot me any questions on Twitter (@DennisMortensen )!
Like What You’re Reading?
Experience autonomous scheduling
with our AI scheduling assistants