What This Update Actually Is
HubSpot added a fallback option to its conversation routing engine. When a conversation arrives and no human agents are free, the system can now automatically assign it to your Customer Agent instead of leaving it unassigned.
Before this, if your team was at capacity or off the clock, that conversation sat in limbo. It went to a waitlist or stayed unassigned. The customer waited. The ticket aged. The SLA clock ran.
Now there's a defined path. Admins configure a Customer Agent fallback inside each channel's routing settings. The routing engine checks for available humans first. If none exist, and a fallback is set, Customer Agent picks up the conversation automatically.
The technical implementation is clean. HubSpot added an optional customerAgentFallbackId field to the RoutingRuleDefinition interface. It's backwards compatible. No existing routing types or resolvers break. If you don't configure a fallback, nothing changes.
Why HubSpot Shipped This
The problem was real and common. Humans on support teams don't work around the clock. Lunch breaks happen. High-volume spikes happen. When conversations arrived outside those windows, they piled up unassigned.
That created two pressures. First, SLA breaches. Second, a hesitation to turn on Customer Agent at all. Admins worried: what if Customer Agent is on but humans are also available? Who gets priority? Is the AI going to step on a live rep?
This update resolves both. It gives Customer Agent a clear, defined role: step in only when no human is there to step in first. That hierarchy makes it safer for admins to deploy Customer Agent with confidence.
It also fits a broader pattern HubSpot is building toward: AI that augments humans rather than competing with them. If you've been following the Agent Builder rollout for custom AI agents, this is the same philosophy applied to the inbox.
How to Use It Step by Step
This is an admin-level configuration. It lives inside each channel's routing settings, so you can set different fallbacks for different channels if your support structure calls for it.
- Go to Settings, then Inbox, and select the channel you want to configure.
- Open the Routing settings for that channel.
- Find the new fallback option below the existing user or team assignment selector.
- Select your configured Customer Agent as the fallback.
- Save. The routing engine will now use this fallback automatically when no human agents are available or at capacity.
A few things to confirm before you configure this. Make sure your Customer Agent is fully set up and tested in a live environment. A fallback that routes to a misconfigured agent creates a worse experience than no fallback at all.
Also confirm your human agent availability settings are accurate. The routing engine's fallback decision depends on those signals. Stale availability data will cause premature fallbacks.
What It Touches in Your HubSpot Strategy
This update ripples further than the inbox. Here's what it actually affects across your portal.
Service Hub and SLA Reporting
Unassigned conversations are invisible to SLA tracking in a meaningful way: they still breach, but no one owns the miss. With fallback routing active, Customer Agent picks up the conversation immediately. The clock still runs, but now something is happening. That changes your SLA reports from showing missed breaches to showing active handling.
Key Takeaway
Customer Agent fallback doesn't pause the SLA clock. It starts active handling immediately, which protects your first-response metrics even when humans are unavailable.
Conversation Handoff and Workflow Triggers
When Customer Agent handles a fallback conversation, it can still triage, gather information, and set properties on the contact or ticket record. That means any workflow you've built on those properties can still fire. Escalation workflows, ticket prioritization, CRM updates: all of them can still run even when a human didn't start the conversation.
If you've recently configured user availability controls in lead workflow rotation, this fallback follows a similar availability-aware logic. Both updates reinforce the same principle: routing should respect who's actually there.
Channel-Level Configuration Flexibility
Because the fallback is configured per channel, you can be strategic. Set Customer Agent as the fallback for your live chat channel but leave your email inbox running to the waitlist. That level of control matters if you have different SLA expectations or staffing patterns across channels.
Key Takeaway
Configure the fallback channel by channel. Don't apply it universally until you've confirmed Customer Agent is trained and tested for the volume and topic mix of each channel.
AI Deployment Confidence
This is really a deployment-safety update as much as a routing update. The humans at your company who've been nervous about turning Customer Agent on now have a clearer mental model: humans first, Customer Agent only when humans aren't there. That framing reduces internal resistance.
Who Should Care Most
This update matters most to specific roles and company profiles. Here's who should act first.
- Service Hub admins managing live chat or messaging channels with limited after-hours coverage. Configure the fallback now and stop SLA bleeding overnight.
- RevOps leaders tracking first-response SLA performance. This directly improves the metric without adding headcount.
- Small and mid-size teams where one or two humans cover the inbox. Capacity spikes hit these teams hardest, and Customer Agent fallback absorbs those spikes.
- Companies that have Customer Agent configured but not yet deployed because of uncertainty about human-versus-AI handoff. This update removes the main blocker.
- Global teams covering multiple time zones where gaps in human coverage are structural, not occasional.
George's Take
I've seen this exact pain in portal after portal. A company gets excited about Customer Agent, sets it up, tests it, and then hesitates at the finish line because they can't clearly answer "but what happens when a real person is available?" That ambiguity was the actual adoption blocker. This update removes it. The routing order is now explicit: humans first, Customer Agent as the safety net. That's the right mental model, and it's the mental model most service leaders already had in their heads. HubSpot just made the software match the intent.
“The routing order is now explicit: humans first, Customer Agent as the safety net. That's the right mental model, and HubSpot just made the software match the intent.”
If you haven't looked at how custom agents and Customer Agent fit together yet, check out our breakdown of HubSpot Custom Agents moving to public beta. The two features are becoming a real ecosystem, and this fallback routing update is a key piece of that picture.
If your team is ready to configure this but wants a second set of eyes on your Customer Agent setup, your routing rules, and your SLA strategy, let's talk. The Sidekick team reviews portals every week and we know exactly what a well-structured fallback configuration looks like in practice. Book a strategy call and we'll help you get it right the first time.
Frequently Asked Questions
What is the Customer Agent fallback in HubSpot conversation routing?
It's a routing option that automatically assigns an incoming conversation to your configured Customer Agent when no human agents are available or at capacity. Admins set it per channel in routing settings. If no fallback is configured, conversations continue to route as they did before: to a waitlist or left unassigned.
Which HubSpot hubs and tiers include conversation routing fallback to Customer Agent?
This update is available on all HubSpot hubs and all tiers as of July 27, 2026. You don't need a specific hub or paid tier to access the fallback configuration. You do need a Customer Agent already set up in your portal before you can assign it as a fallback.
Does Customer Agent fallback affect my SLA reporting?
Yes. When Customer Agent picks up a conversation via fallback, active handling begins immediately. This protects your first-response metrics. The SLA clock doesn't pause, but a response is being generated rather than a ticket sitting unassigned. Your SLA reports will reflect active handling rather than an open breach.
Can I configure different Customer Agent fallbacks for different channels?
Yes. The fallback is configured at the individual channel level inside routing settings. You can set Customer Agent as the fallback for live chat but leave email routed to a waitlist. This lets you match fallback behavior to the SLA expectations and staffing patterns of each specific channel.
What happens if I don't configure a Customer Agent fallback?
Nothing changes. If no fallback is set, conversations that arrive when no human agents are available will continue to go to the waitlist or remain unassigned, exactly as they did before this update. The fallback is fully optional and backwards compatible with existing routing configurations.
Does Customer Agent fallback override human agent assignments?
No. The routing engine always checks for available human agents first. Customer Agent only receives the conversation if no human agents are available or all are at capacity. It's a true fallback, not a parallel assignment. Humans remain the first priority in the routing sequence.
