Hey Dhawal! Morgan here

We’ve recently started paying more attention to our overall CX score, rather than CSAT - This gives a more holistic view of all meaningful conversations, rather than just those customers rated. I believe it’s in full release, but if you don’t see it in your reports it may be worth speaking to Intercom support or your CSM about. 

 

CX score is great but I don’t find it reliable enough for our usecases. For example, CX score is low if customer is leaves midday despite my support team doing whatever is required both in a process and empathy standpoint. 

Naturally, response rate depends on multiple factors, but considering we are still in the holiday season, lower conversation number due to holidays also may affect CSAT score. Also, percentage points may look more dramatic and fluctuating if the number of responses is low generally speaking. A couple of recommendations without knowing the context:

1. In the “goodbye” canned response, you can add a message that hints about the survey coming in the end 
2. If you use third party CSAT tools, you can trigger them at the same time with goodbye message, so it’s more likely to be seen

Happy to dive deeper in DMs! 

Hey ​@Dhawal Thumar, 

Thanks for sharing this! It's great that your CSAT score remains high, but I understand the concern with the declining response rate. A few things that might help:

• Timing matters – Try sending the survey a few minutes after the conversation ends rather than immediately or much later.

• Keep it super short and clear – Let users know it'll only take a second. Sometimes even changing the wording slightly can boost responses.

• Personalize the message – A friendly, human tone often leads to better engagement.

• Test incentives – If it fits your brand, small rewards or acknowledgments can nudge more people to respond.

• A/B test different formats – Try email vs in-app, or tweak the subject line if using email.

Would love to hear what others have tried as well!

Also you can check this article to understand CSAT properly.

Thanks

Hey Dhawal, totally hear your concern as declining response rates make CSAT feel less reliable, even when the scores look great. One approach you might find useful is looking beyond surveys altogether.

At Isara, we’ve built a predictive CSAT (pCSAT) that uses AI to analyze conversations directly, so you get a satisfaction score without needing customers to fill in surveys. It helps you track sentiment trends and identify frustration points across all conversations, not just the 25–30% who respond.

This way you don’t lose visibility when response rates dip, and you still have actionable insight to guide improvements. Happy to share more if you’re curious!

Sounds amazing Florian B! How is it different from CX score and how did you build pCSAT? 

The main difference is our AI predicts the CSAT score for each closed conversation so you get

• CSAT scores for all your conversations
• Unbiased CSAT

There are a few differences from standard CSAT but it very much depend of the company and how they send and manage CSAT surveys.

We offer a lot more features and offer 30-day free trial. We are available on the intercom appstore if you want to have a look

https://www.intercom.com/app-store/?app_package_code=tappa-insights&category=analytics

Florian