What your on-call engineers see that your dashboards don’t.
Someone on your team thought this was worth your time.
In short
- High uptime and frustrated customers can be true at the same time. Most monitoring measures systems, not whether customers finish what they came to do.
- How your people understand and value reliability decides whether SLOs, alerts, and reviews actually change behaviour. That perception differs by team, and it can be measured.
- An independent pulse survey measures it across four questions in about five weeks, from $4,500, and repeating it quarterly shows whether your reliability work is landing.
1. The same product, measured two ways
Engineering dashboards report uptime, latency, and error rates. Customers report something else: searches that fail one try in ten, pages that work after a refresh, exports done by hand because the feature is flaky. Both views are accurate. When only the first one reaches leadership, decisions get made on half the picture.
| What the dashboards say | What customers say |
|---|---|
| Uptime 99.9% | “Search fails one try in ten.” |
| Error rate 0.1% | “It worked when I tried again.” |
| 40 deploys a week | “It’s been slow all week.” |
Illustrative example.
2. Why it matters commercially
Small, repeated breaks in the customer experience wear down trust long before they show up as an outage. Customers rarely complain about each one. They quietly decide your product is unreliable, and a competitor with fewer features starts to look safer.
Depending on the industry and the study, winning a new customer costs 5 to 25 times more than keeping an existing one1. Google’s own SRE guidance notes that service availability doesn’t always result in user happiness2.
3. Why people, not tooling, decide the outcome
SLOs, runbooks, and error budgets only work when the people around them understand why they exist and believe they are worth the effort. Without that, alerts get muted, postmortems get skipped, and the error budget becomes a number nobody checks.
Leadership, on-call engineers, product, and support can see reliability very differently, and those differences tend to stay hidden when each group assumes the others share its view.
4. Four questions worth asking every team
| Understanding | How well are reliability principles understood across the organisation? |
|---|---|
| Value | How important do staff feel reliability is to customers and the business? |
| Practice | How well are those principles actually used in day-to-day decisions? |
| Outlook | How positive do people feel about where reliability is heading? |
If leadership and engineering answer the third question differently, that difference is the conversation to have.
5. A low-risk way to find out
YourSRE runs an independent Reliability Pulse Survey. Staff answer a short survey run by us, not by you, which gets more candid answers than an internal form. Results are reported by team, role, and level, never by individual.
| Timeline | About five weeks from scoping to readout |
|---|---|
| One-off survey | $4,500 USD, up to 150 respondents, written readout and debrief |
| Quarterly programme | $14,400 USD a year: four surveys on the same baseline, trend reports, and a leadership briefing each quarter |
Suggested next step
A free 30-minute call to scope it, with no commitment. Or, to see how your organisation scores on its own first, try the free readiness report.
Daniel Holmes, founder of YourSRE · hello@yoursre.com · yoursre.com