Skip to main content

Hi Everybody,

 

We're a small SaaS company (https://trysmartplan.com using Intercom for our Customer Service. I'm the co-founder of the company and in the beginning I did all the customer support.

 

Now we have 5 part time employees covering our customer support shifts.

 

After hiring employees for this job, I have found it hard to stay in the loop of what kind of support we have. Yes I talk to my employees and yes they report feature requests etc. But to feed my gut feeling about what is going on I have to click through a lot: closed, snoozed and open.

 

The same goes for Quality Assurance. How do you make sure the quality of the responses are good enough?

 

I would love to hear if this is something you also struggle with or if it's just me :)

 

Hope you are all safe.

 

Best regards,

Mathias

Hey Mathias! Speaking from my own experience of having worked with Intercom's Support team, the best way to stay in the loop about the type of support your team is offering is through Quality Assurance.

 

For a team of your size, my suggestion would be for you to select a random sample of conversations for each of your employees on a regular basis (once a week), review these and then provide positive and constructive feedback in a 1:1 scenario.

 

There are a number of third-party apps that integrate directly with Intercom that will help you achieve this - for example, I know Klaus have been active in our Community! @daniel f11​, what would be your top tips for keeping on top of QA?


Hey @mathias a​ and @eric f11​ 👋

 

In 5 steps, this is how you can get started with conversation reviews:

 

  1. Signup with a dedicated customer support quality assurance software that integrates seamlessly with Intercom (like klausapp.com does)
  2. Then create a simple scorecard with the rating categories you want to pay attention to (tip: keep it simple and unambiguous)
  3. Use a filter to find the most relevant conversations - e.g. random sample from the previous 30 days, conversations with many replies and/or low satisfaction ratings, conversations with churned customers, etc
  4. Start reviewing conversations yourself and/or ask your team members to review each others' conversations (tip: make sure to add comments explaining why something was great or needs improvement)
  5. Review individual and team quality scores so you can see how much you have progressed and what exactly you still need to improve (tip: use the software dashboard for that purpose)

 

Repeat, refine and you will see tangible results in customer satisfaction :)

 

PS: You may find this guide useful to get started!

 

Cheers,

 

Daniel


Thanks for both of your inputs. I just tried the Klausapp and it was a bit too much for my need.

 

I like the QA approach to select a sample and go through it.

 

I also talked to one of my developers and asked to get a list of all new messages everyday and show them in a feed for me. I think that would be a nice way to glance through all new activity "today", just to feed my gut feeling about what our customers reach out about.

 

This way I could fetch that feed every afternoon or night and check out the activity.

 

Thanks again for your input.

 

Mathias


If you're trying to stay up to date with the types of conversations your team is having (common questions, feedback, issues, customer love, etc.) I'd suggest using a more advanced tag manager that would enable you to both get Slack alerts and for you to dive in at any time to see different categories of convos or which topics/requests/questions are most frequent. We do this with Userfeed.io.

 

Landon Bennett

Co-founder of Intercom apps: Userfeed & Bump


On top of all of the above, I would suggest using the bot flow and conversation attributes to allow you to wrap your conversations with a taxonomy that fits your business and then as called out look to sample the conversations.

 

  1. Set up some quality criteria that you can use as your checklist and as a calibration point (if you want others to also check)
  2. Decide on what data you want to sample from (see 3 / 4)
  3. I would start via customer sentiment, and sample negative (if any) and positive (say a 70/30 mix of neg/pos)
  4. Look at your performance metrics / reports (MTTR/FCR/SLA) to identify peaks in volume based on your taxonomy and target those conversations (smoke = 🔥)

 

It reads like you do not want to go to crazy on 3rd party tools or automation, so a simple checklist and data to indicate what to sample should get you going. Once you land the manual processes the above responses are all good shout outs on what else to look at in terms of tech etc.

 

Let us know where you land and how it goes.

 


By the way, here's how Intercom did some of this when they were scaling up their support team: https://www.intercom.com/blog/how-to-maintain-high-quality-conversations-with-your-customers-at-scale//p>

 

Landon Bennett

Co-founder of Intercom apps: Userfeed & Bump


Really similar to Klaus App


Definitely recommend checking out some of the Intercom Support Blog Posts, lot's of handy tips and useful information on delightful customer service:

 

https://www.intercom.com/blog/customer-support//p>


Reply