We’ve been revisiting our customer health score—mapping out the moments that matter most in building successful customers and figuring out how to double down on them.

Two early ones stood out as root causes behind repeat support tickets: setup and recovery.

If someone isn’t technically installed correctly, they won’t reach maximum value with Appcues. And if an experience breaks and doesn’t recover, that can be a moment where trust starts to slip. Both are foundational. Both shape what comes next.

We already push these actions in-app—through onboarding checklists and reminders. But some moments need multiple touchpoints. So we wanted to double down and focus on catching the customers who might benefit from additional support and offering it through 1:1 conversations, at the right time.

Once we identified setup and recovery as root-cause moments, I wanted support to step in earlier—before these things turned into tickets.

Both plays are powered by Workflows, sending a short, human message at the right time. We chose email so it could land in the inbox, be forwarded to teammates, and booked when convenient. You could also run these as in-app messages if you’re aiming for a quicker, in-context response.

1. Installation: Catching incomplete setup

Setup is one of those things where customers think they’re done—but from our side, we can tell something’s off. A property isn’t flowing through. An event’s missing. And then nothing downstream works the way it should.

We used a technical health score in Salesforce to flag accounts where setup looked incomplete. If that score dropped below a threshold, the account entered a Segment—and a Workflow sent a short message offering a quick install review.

What happened

• The first install check was booked within hours
• Now we’ve got a steady stream of short conversations that unblock value sooner

Pro Tip: If you don’t have a health score, look for proxy signals—like whether key events have fired, or setup steps were completed. Use that to automate a nudge or offer live help when someone might need a hand.

2. Recovery: Stepping in when issues don’t bounce back

The other moment we focused on was recovery. Alerts get triggered, but sometimes nothing gets fixed. That’s not always a problem… but when it is, it’s often too late to turn things around.

So we built a recovery rate: recovery alerts divided by issue alerts. If it dropped below 50%, the account entered a different Segment. That triggered a Workflow and a message directly from me with a calendar link to dig in together.

What happened

• 30+ calls booked in the first few weeks
• The accounts on those calls represented $500k+ in ARR
• Completion rates improved after the sessions, and teams made better targeting decisions going forward

Pro Tip: Every product has a few break points that chip away at trust. Define what recovery looks like, create a lightweight signal, and reach out before it snowballs. It doesn’t need to be perfect—just enough to catch the right accounts, early.

These plays are simple, but they compound. Setup, recovery, configuration, delivery… each one is a moment where momentum can stall or build.

We’re continuing to map precursor signals across the customer journey—looking for the next quiet moment that matters. (And we’ll share what we find soon.)

One message at the right time can change the tone of a relationship. Sometimes it’s all it takes to keep everything else moving forward.

And because these Workflows now run in the background, they’ll keep catching the next account (and the one after that) without any manual lift.