How Poll Everywhere cut on-call tooling costs and made response faster with Rootly.

Poll Everywhere powers 40M+ presentations, engaging 200M+ participants and 10M+ presenters. The company is fully remote, with most of the team based in the US and additional coverage in Malaysia and Japan—a footprint that supports customers wherever and whenever they’re presenting.

That same mindset carries into engineering. With just six engineers total (including the CTO, staff engineers, and a dedicated SRE), the team runs a lean, high-velocity operation: continuous delivery with multiple deployments per day, backed by a practical reliability target of ~99.9%. For Travis Haas, Poll Everywhere’s SRE, the mission is simple: “Making presentations accessible”—and operationally, that means balancing speed with stability so the platform stays dependable while customers are actively using it.

When alerts wake the wrong person

Poll Everywhere already had strong detection: Datadog handled monitoring and alerting. The gap was what happened after an alert fired—how quickly the right people could coordinate, and how consistently the rest of the company could stay informed.

Paging didn’t match how the team worked

Before Rootly, alerts flowed from Datadog to PagerDuty, which largely meant SMS and phone calls to the on-call engineer. That created friction in their distributed team.

Poll Everywhere has meaningful “awake coverage” across time zones—especially Sunday nights through Friday afternoons in North America. But the paging workflow couldn’t take advantage of it. If a monitor triggered, the US on-call would still get woken up, even if someone in Malaysia or Japan was already working and could have handled the initial response.

Incident context lived outside Slack

Poll Everywhere runs as a fully remote company. Slack is how the team communicates and coordinates day to day. The problem was straightforward: the on-call engineer received a page, while the rest of the organization lived in Slack.

As Travis put it: “Slack is a huge… no integration was unacceptable.” Without Slack-native incident response, the wider team lacked shared awareness. Support and customer-facing teams could learn about incidents late, and incident coordination started from a smaller, more isolated context than it needed to.

Escalation behavior was harder to tune

With a six-person team, incident response is shared ownership by necessity. Everyone can triage; specialists jump in as needed (infrastructure, LMS integrations, and other domain areas). The on-call engineer’s job is to stabilize the situation and bring in the right expertise, if needed—not to fix everything alone.

But their previous tooling had less flexibility around escalation behavior—especially secondary notifications and rollover timing. Travis described the practical impact at 3 a.m.: if the primary has acknowledged, they don’t want the system escalating immediately to secondary while they’re still orienting themselves.

Cost didn’t align to value

At renewal time, cost became the forcing function. For the limited set of features they were using, PagerDuty felt “exorbitant.” Poll Everywhere didn’t need more tooling—they needed incident response that fit a remote-first workflow without paying for complexity they weren’t using.

Put together, the pattern was clear: detection wasn’t the bottleneck. Coordination was. Poll Everywhere needed paging, scheduling, and incident visibility where the team already worked—Slack.

“Poll Everywhere runs live—when customers are presenting or teaching, reliability isn’t optional. Rootly helps us keep that experience dependable without slowing us down.” -Travis Haas, Principal SRE

From paging to coordination.

Poll Everywhere adopted Rootly for On-call and Incident Response, intentionally keeping scope focused on the parts of the workflow that were hurting most: alert routing, escalation controls, and shared visibility during incidents.

Time-to-value in under an hour

Setup was fast. Travis configured schedules, invited the team, and got to first value in less than an hour.

The main setup challenge was exactly the kind of detail that matters to small teams: Poll Everywhere runs primary and secondary rotations as separate schedules, and they needed to ensure no overlap—nobody should be both primary and secondary at the same time. Rootly solved that and made onboarding straightforward: invite users by email, and each responder chooses their preferred notification method.

What changed operationally

Rootly didn’t change what Poll Everywhere monitors or where they write retros. It changed the coordination layer:

• Alerts and incident context now land in Slack, where the team already works
  
• The team can leverage time zones—someone online can triage before paging wakes the US on-call
  
• Escalations are intentional (acknowledge → assess → page secondary when needed)
  
• Incident visibility is broader by default, making internal updates more consistent

“Rootly was fast to roll out—within an hour we had schedules, escalations, and Slack-based alerting working, with primary and secondary on-call cleanly separated.” -Travis Haas, Principal SRE

Less pager fatigue, more shared visibility.

Poll Everywhere is explicit about what Rootly changed and what it didn’t. They often go weeks without an incident, and when incidents do happen, the difference is response quality: who gets pulled in, how fast coordination happens, and how well the rest of the business stays aligned.

Lower pager fatigue and more sustainable on-call

For a six-person team, sustainability matters. Less noise means less cognitive load, fewer context switches, and more energy for reliability work that prevents incidents in the first place.

Clearer handoffs and shared ownership in practice

On-call handoffs became clearer and cleaner, directly crediting the Slack-native workflows. That’s a meaningful operational shift: when incident coordination is visible in Slack, escalation becomes faster and more confident—especially in a team where the on-call engineer triages and then pulls in specialists as needed.

More consistent stakeholder communications

The biggest qualitative improvement, beyond cost, was internal incident visibility. Travis described the advantage of letting teammates “join a Slack channel and [know] there was an incident… at this date and time.” That matters because it helps customer-facing teams respond with clarity when customers ask questions, and also get ahead and update status pages in a timely manner.

“Rootly made on-call sustainable for a small team: less pager fatigue, clearer handoffs, and better visibility for the whole company when something’s going on—plus we got there without paying PagerDuty prices.” -Travis Haas, Principal SRE

Proof points (Poll Everywhere’s internal read-out)

• 5/5: Pager fatigue lower
• 5/5: On-call handoffs are clearer and cleaner
• 4/5: Time to first status update improved
  
• 4/5: Stakeholder communications more consistent 
• 5/5: Engineering morale improved

The most directly measurable business impact was cost. PagerDuty pricing at renewal didn’t match the value Poll Everywhere was getting, and Rootly offered a better fit for the workflow they actually needed—scheduling, escalation, and Slack-native incident response.

As Travis summed it up:

“Slack is huge for our dispersed, remote team to communicate, collaborate, and ultimately get things done efficiently… no integration was unacceptable.”

“For the features we were using…the price point we were paying was…exorbitant.”

For Poll Everywhere, the story is an SRE classic: monitoring wasn’t the hard part. Coordinating humans under pressure was. Rootly helped them align incident response with how a remote team works—keeping reliability high, on-call sustainable, and the rest of the company in the loop when it matters most.