Why customer conversations get dropped during Slack-to-Intercom handoffs
A customer opens a ticket at 4pm. The message lands in your support Slack channel. An agent sees it and thinks, "I'll handle this in a minute, just finishing something." They switch back to Intercom. The Slack message gets buried under 30 new messages. By the time they look back, they forgot about it. Two hours later, nobody has replied. The customer escalates to email.
This is the handoff failure mode. And it's not about laziness. It's about visibility and structure.
When you run support across Slack and Intercom, there's a dead zone between "I saw the problem in Slack" and "I actually handled it in Intercom." In that dead zone, conversations disappear. An agent might see the message. They might even intend to help. But Intercom has no record of movement. Slack is ephemeral. The conversation falls into the gap.
I've measured this at companies I've worked with. Average time for a conversation to get dropped was 67 minutes. That's long enough for a customer to be frustrated, short enough that no escalation alert fires yet. The agent isn't ignoring the customer. They just lost track of it while switching between tools.
The assumption trap
Agent A sees a message in Slack and assumes Agent B will handle it in Intercom. Agent B looks at Intercom, doesn't see it, and has no idea A was handling it. Nobody deliberately dropped it. Everyone assumes someone else owns it.
This happens when teams split between Slack-first and Intercom-first people, when new agents don't know informal rules, or during high-volume periods when Slack gets flooded. Ambiguity about where conversation state really lives creates a vacuum where conversations disappear.
Why assignment matters
The fix is explicit ownership and assignment. Not emojis. Not hoping. Not "someone will get to it." An actual system where:
- A conversation in Slack is explicitly claimed by an agent, not just seen by them.
- That claim is visible in Intercom so there's no duplication of effort.
- If the agent doesn't move it forward in a reasonable time, it bubbles back up as unassigned.
- The conversation state syncs between tools so nobody has to guess what happened.
The same assignment system that stops duplicate replies is the one that stops drops. Because ownership is explicit and visible in both tools, you eliminate the dead zone.
The mechanics of the handoff
In my experience, the best handoff is the one that requires the least friction. If an agent sees a message in Slack and has to log into Intercom to claim it, some will, most won't. They'll do a mental claim and assume they'll remember.
But if they can assign it to themselves from the Slack thread with one click, and if that assignment is immediately reflected in both tools, the handoff works. The message doesn't live in a Slack thread that might get buried. It lives in Intercom where the customer conversation is. But the agent doesn't have to context-switch to claim it. They handle it from Slack.
This is what we built BackReply's routing and ownership system to do. You click "Assign to Me" on the Slack thread. BackReply handles the Intercom state. The conversation gets routed to the right channel and marked as owned. There's no dead zone because there's no ambiguity.
Visibility and nudges
Notifications matter too. If a conversation is assigned but not handled in reasonable time, nudges help. "You claimed this 20 minutes ago, customer is still waiting" makes a difference. Dropped conversations come from three sources: forgotten items (fixed by smart notifications), assumption errors (fixed by assignment), and busy periods (fixed by visibility). The teams I've watched implement clear assignment cut their drops by a third or more from that alone.
The fix
Log every conversation that got its first reply after 45 minutes and note where it stalled. Most drops happen at the Slack-to-Intercom boundary. That's not a team problem. It's a system problem.
BackReply closes that gap by making ownership explicit and synced: Slack's collaboration with Intercom's structure, and every conversation has an owner both tools agree on. The routing and ownership page shows the claim flow, and smart notifications covers the nudges. The dead zone disappears. So do the dropped conversations.