We are a small team looking to maximize our efficiency with Intercom. We want to streamline our customer support and communication processes, but we're cautious about over complicating things with too many features. Find Out What are the best ways to set up workflows in Intercom that are simple yet effective for a small team? Are there any must-use features or strategies we should implement to keep things running smoothly? Additionally, how can we ensure our team remains organized and responsive while minimizing manual tasks? Any tips or information from others in similar situations would be greatly appreciated!
Answered
How to Optimize Intercom Workflows for a Small Team?
Best answer by Katee Saffer
We’re also a relatively small support team (6 people), and we’ve focused on using workflows to automate triage and follow-ups while keeping the system simple enough to manage.
Over the past 6 months, together with Penny (our chatbot that’s powered by Fin) we handled 36,113 new conversations, so automation and AI responses have been essential to keep things manageable with a small team.
Here are a few workflows that have helped us the most:
1. Routing conversations by plan and urgency
We use workflows to route conversations to different inboxes depending on the customer’s plan level and the urgency of the message. This ensures higher-tier customers or time-sensitive issues are prioritized automatically instead of requiring manual triage.
2. Automatic follow-ups for snoozed conversations
If a conversation is snoozed and the customer doesn’t respond, a workflow sends a follow-up. This has helped prevent conversations from slipping through the cracks and keeps us from having to manually check on these every day.
3. Automatic CSAT requests after closing conversations
We automatically send CSAT after closing a conversation. This gives us consistent feedback without agents needing to remember to send surveys.
4. Working with Fin
We also use Penny (powered by Fin) to answer common questions and route more complex ones. Over time this has increased our resolution rate to about 70%, which significantly reduces the volume agents need to handle directly.
5. Strong team buy-in for improving AI responses
One of the most important things for us has been getting the entire team involved in improving our AI assistant. Although agents don’t directly train the AI, we built a simple feedback loop into our process:
-
Every conversation that reaches an agent requires a custom attribute to be completed before closing.
-
Agents are asked whether the AI interacted in the conversation (Yes / No).
-
If Yes, they rate whether the reply was accurate.
-
If the reply was anything other than correct, they provide a quick note explaining what went wrong.
That feedback is automatically collected into a report that I (the AI manager) reviews regularly to identify knowledge gaps, incorrect answers, or missing documentation.
This has made improving our AI assistant a team effort instead of a single person’s responsibility, and it’s been one of the biggest drivers of improvement over time.
---
I hope this is helpful! We’ve been using Intercom for about 9 months now, and one thing I’d definitely emphasize is that it takes some time, testing, and iteration to really figure out what works best for your team.
When we first started, we tried to keep things very simple and gradually added workflows as we learned more about our support patterns. Over time, we were able to see which conversations could be automated, which needed routing, and which still benefit most from human support.
We’re actually in the process of making another round of improvements to our workflows now that we have a better understanding of what Fin can reliably handle vs. what should be routed to our team. That learning period has been really important.
If you approach it as an ongoing process rather than a one-time setup, you’ll end up with workflows that truly support your team instead of adding complexity.
Good luck getting everything set up!
- Intercom Team
Hey
Thanks so much for reaching out and for sharing a bit about your setup, it makes total sense that you want to keep things lean and effective rather than overwhelm a small team with complexity.
Here are some practical ways to set up Intercom so it stays simple but powerful for you:
Start with a few high‑impact Workflows, instead of trying to automate everything, I’d suggest beginning with 3–4 core workflows:
-
New conversation triage
- Auto‑tag conversations based on things like:
- Where they came from (website page / product area)
- Keywords (e.g. “login”, “payment”, “scholarship”, etc.)
- Route to a single “Support” inbox for now, rather than lots of different inboxes.
- This gives you structure without needing a complex queue system.
- Auto‑tag conversations based on things like:
-
Out‑of‑hours + busy times auto‑reply
- Send an automatic response that:
- Acknowledges the message.
- Sets expectations on when they’ll hear back.
- Shares 1–2 key Help Center links so they can self‑serve while they wait.
- This keeps your team responsive without anyone needing to be online 24/7.
- Send an automatic response that:
-
Simple follow‑up / “no reply yet” nudges
- If a customer hasn’t replied after X days, send a gentle check‑in (“Just checking if you still need help with this.”).
- If you haven’t replied yet (for example, after Y minutes/hours), trigger an internal alert or escalate the conversation so nothing gets forgotten.
-
Basic lead capture on key pages
- On high‑intent pages (e.g. pricing, applying, FAQs around funding), use a short bot flow to:
- Ask 1–2 qualification questions.
- Collect an email if they’re not logged in.
- Hand over to your team only when there’s something meaningful to handle.
- That keeps your support inbox focused on the most important conversations.
- On high‑intent pages (e.g. pricing, applying, FAQs around funding), use a short bot flow to:
Review and adjust, don’t over‑build.
Since you’re cautious about over‑complicating things, a good rhythm is:
- Start with the small set of workflows above.
- After a couple of weeks, review:
- Which workflows actually saved time?
- Where did customers still get stuck?
- Only then, add or tweak flows. If something isn’t helping, simplify or remove it.
+1We’re also a relatively small support team (6 people), and we’ve focused on using workflows to automate triage and follow-ups while keeping the system simple enough to manage.
Over the past 6 months, together with Penny (our chatbot that’s powered by Fin) we handled 36,113 new conversations, so automation and AI responses have been essential to keep things manageable with a small team.
Here are a few workflows that have helped us the most:
1. Routing conversations by plan and urgency
We use workflows to route conversations to different inboxes depending on the customer’s plan level and the urgency of the message. This ensures higher-tier customers or time-sensitive issues are prioritized automatically instead of requiring manual triage.
2. Automatic follow-ups for snoozed conversations
If a conversation is snoozed and the customer doesn’t respond, a workflow sends a follow-up. This has helped prevent conversations from slipping through the cracks and keeps us from having to manually check on these every day.
3. Automatic CSAT requests after closing conversations
We automatically send CSAT after closing a conversation. This gives us consistent feedback without agents needing to remember to send surveys.
4. Working with Fin
We also use Penny (powered by Fin) to answer common questions and route more complex ones. Over time this has increased our resolution rate to about 70%, which significantly reduces the volume agents need to handle directly.
5. Strong team buy-in for improving AI responses
One of the most important things for us has been getting the entire team involved in improving our AI assistant. Although agents don’t directly train the AI, we built a simple feedback loop into our process:
-
Every conversation that reaches an agent requires a custom attribute to be completed before closing.
-
Agents are asked whether the AI interacted in the conversation (Yes / No).
-
If Yes, they rate whether the reply was accurate.
-
If the reply was anything other than correct, they provide a quick note explaining what went wrong.
That feedback is automatically collected into a report that I (the AI manager) reviews regularly to identify knowledge gaps, incorrect answers, or missing documentation.
This has made improving our AI assistant a team effort instead of a single person’s responsibility, and it’s been one of the biggest drivers of improvement over time.
---
I hope this is helpful! We’ve been using Intercom for about 9 months now, and one thing I’d definitely emphasize is that it takes some time, testing, and iteration to really figure out what works best for your team.
When we first started, we tried to keep things very simple and gradually added workflows as we learned more about our support patterns. Over time, we were able to see which conversations could be automated, which needed routing, and which still benefit most from human support.
We’re actually in the process of making another round of improvements to our workflows now that we have a better understanding of what Fin can reliably handle vs. what should be routed to our team. That learning period has been really important.
If you approach it as an ongoing process rather than a one-time setup, you’ll end up with workflows that truly support your team instead of adding complexity.
Good luck getting everything set up!
Join the Intercom Community 🎉
Already have an account? Login
Login to the community
No account yet? Create an account
Intercom Customers and Employees
Log in with SSOor
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.