CSAT Flow Conundrum
Hoping some other customers can help give us some ideas here.
The Problem:
We like to follow up on bad rated tickets to improve the overall customer experience, but want to ensure that ratings are tied correctly to agents in reporting.
Current Survey Set up:
- Embedded html survey added to response when ticket is solved by agent
- Automation for survey when this does not happen (with several exclusions)
- Trigger to re-open bad rated tickets upon rating received
- Custom Field to designate need for Escalation/Follow Up
- Custom Field for Bad CSAT Reasoning
Version 1: Tickets don't close after the rating is submitted by the requester
Pros:
- Allows potential for bad > good
- allows follow up in the same thread
- allows QA categorization of Bad CSAT reason (as opposed to the customer self-selected option) as a custom field
Cons:
- Reporting does not allow us to link CSAT with the last agent who responded when the rating was offered
- Tricky "house of cards" trigger set up to avoid accidental reassignment of bad CSAT tickets. Does not solve for escalated bad tickets across teams and is subject to user error
- Our volume is too high to force individual ownership of bad ratings
- Rarely, a customer will rate a ticket twice, both bad, and this reflects on the agent unfairly.
Version 2: Close Tickets Immediately after Requester Submits Rating
Pros:
- Preserves CSAT/Agent connection in reporting
- More transparency in agent views that aren't clouded with bad tickets
- Less potential for trigger failure or lost tickets
Cons:
- Follow up process is cumbersome and creates double tickets for each interaction
- No way to mark closed tickets as reviewed
- No way to categorize closed tickets (we do this in the follow up)
- Loss of ability to change bad > good (we purposely exclude ratings on follow up tickets to prevent double bads)
Please offer any and all suggestions!
-
Hello Jeremy,
Certainly can see your point. We currently track CSAT as a team.
How many agents are involved? (affects the type of solutions that are practical)
What about using a custom field that captures the agent at time of rating? If you dont have a huge amount of agents, having a field with Agent1_good, Agnet1_bad, Agent2_good, Agent2_bad etc, might offer you some other possibilities...
Your thoughts?
-
We have ~120 agents currently.
So you're suggesting the following flow?
Trigger 1:
- All Conditions: Satisfaction changed from offered
- Actions: Custom Field Initial CSAT = agent1_good/agent1_bad
Trigger 2:
- Any Conditions: Satisfaction changed from Bad, Bad with comment
- Actions: Custom Field Secondary CSAT = agent2_good
Trigger 3:
- Any Conditions: Satisfaction changed from Good, Good with Comment
- Actions: Custom Field Secondary CSAT = agent2_bad
How do you handle this in reporting?
-
Hello Jeremy - the agent numbers makes that a bit daunting. That is why I asked about that.
What I was thinking would be awkward after 20 agents or so. All of this will be painful I think with your number as everything is so individual - however you may think of a way to make it more generic. My way is going to need a heap of triggers!
I was thinking to report on the tags... however I was also thinking to have a setting for each agent.
This would depend on when the agent was changed, so I guess we'd need to capture the agent before it is rated...
- When CSAT is offered... set field 'CSAT Agent' to name of agent
- When rating is recieved - if CSAT Agent is 'John' and CSAT rating is Good, set field CSAT Rating to CSAT_Good_John' etc etc and etc...!
- Report on Field CSAT Rating...
I think we'd need to look for some way to do this using dynamic content and freetext fields... and some other magic beyond me! We are interested in this ourselves and have about 35 agents, so I might put some time into trying to find a more generic solution. We took one process that used 30+ triggers/automation down to about 4 total... if we can do the same for this it would be workable.
-
Agreed that what you're suggesting with our numbers is not really sustainable. Admittedly I'm annoyed that there isn't a standard solution to CSAT, even though it's a supported feature, but nonetheless, I do need to come up with something.
Anyone else in your contacts that might be able to offer advice?
-
It's a bit of an advance on the current CSAT offering. Would you want to capture the feedback rating and agent at the first instance only? or if the rating changed, would that still be attributed to the original agent?
My thinking is that a final good rating, despite the original bad, may be, at least in part, based on the earlier agent performance - ie. getting good data etc and maybe laid the basis for a good result. I think we've all seen situation where the agent has received a bad rating with little justification and either the same agent or another is able to bring a successful conclusion..
And if we go for this... then we don't have to capture the rating, just the original agent, which would be much simpler. At worst, one trigger per agent. At best, possibly just a single set of trigger and target.
Your thoughts?
-
We had something like in place, but the issue was that it seemed unfair for a follow up agent to not get credit for a good rating if the initial agent generated the bad one.
Either we allow reassignment post CSAT or we prevent it. I don't think there's a way to condition that.
-
ok - got another possibility... what if we capture Original Agent and Original CSAT seperately?
- Trigger/Target to record O-Agent
- A similar to record O-CSAT
Not too complex...
What have I missed - except of course I haven't figured how to make it work yet...?
How are we going to report on that tidily?
I'm sure there's a better way than this.
Vous devez vous connecter pour laisser un commentaire.
7 Commentaires