Channel Resolver Feedback

We recently launched Moveworks to the business which included the implementation of Channel Resolver. 

We could definitely see the value in this feature however have some feedback on how the experience could be improved for both the user and agent:

How it currently works:

1. Employee asks question in channel eg. peopleinfo-triage

2. Bot identifies whether it has access to knowledge that could assist the employee

3. If it determines with a high confidence that it can assist, it will DM the employee with a response

4. At this point an emoji will appear below the original employee query in the channel

5. Employee receives bot DM and is asked to rate the response as ‘helpful’ or ‘unhelpful’

   1. If ‘helpful’ a tick emoji is added to the original message in the channel

   2. If ‘unhelpful’ a greenflag emoji is added to the original message in channel and GF agent reaches out to assist

   3. GF agent will then add a tick emoji once issue is resolved

Challenges and feedback with current solution

Bot DM - By design, the bot interacts with the user in DM. This presents a couple of challenges for agents:

• If employee rates response as unhelpful, agent has no sight on why that is the case and so has no context when the employee returns for more help

• There is a desire from teams outside of IT, HR etc to adopt Channel Resolver. However, in these instances, there is a lot more nuance to the issues they resolve. So, not having any visibility over the interaction between the bot and the employee is a blocker to adopting Channel Resolver.

When testing the inclusion of ticketing in Channel Resolver we found that the flow was not as intuitive as it could be:

• User comes to channel

• Bot then redirects user to a DM to assist

• User rates response as unhelpful which then redirects them back to the channel again

• At which point they need to manually add a ticket emoji in order to raise a ticket

Ideally the flow would be:

• User comes to channel and raises issue

• In the thread the bot attempts to assist

• If user rates response as ‘unhelpful’ they are immediately offered the option to raise a ticket

Additional Feedback

• It would be helpful if an agent could be defined per channel. As it stands today, setting an employee as an agent is a universal setting. So, for example when a Finance agent, reaches out for help in the HR support channel, the bot will not respond.

• We’d like the ability to introduce other skills to Channel Resolver (eg. Creator Studio use cases)