How long do SaaS outages actually last?

Every vendor posts when an incident opened and when it closed. Nobody adds them up. We do, every day, across 804 status pages — 14,657 incidents that carry both timestamps.

14,657incidents with a start AND a resolve timethe measurable set

2.3 hmedian incident lengthhalf are fixed sooner

28.1 h90th percentile1 in 10 runs longer than this

77vendors with 20+ resolved in 90 daysranked below

Slowest vendors to resolve, last 90 days — median length of every incident they opened and closed in the window (minimum 20 measured). · Source: each vendor's own status page, read daily by this project · as of 2026-09-16

Median time to resolve, last 90 days

Filter by vendor

Vendor	Median	90th pct	Measured
Exact	27.0 h	11.8 days	68
Qualys	25.6 h	6.0 days	54
Ionos Cloud	23.2 h	6.5 days	33
Akamai Edge DNS	22.2 h	5.1 days	25
Luno	21.5 h	4.7 days	20
ThousandEyes	17.1 h	14.9 days	23
F5	11.4 h	40.7 h	29
Datto	10.1 h	6.3 days	57
IPVanish	6.8 h	21.8 h	37
Twilio	6.2 h	2.5 days	114
Kraken	6.0 h	5.1 days	62
Snyk	5.5 h	47.2 h	20
Ivanti Cloud	5.5 h	3.5 days	43
Imperva	5.0 h	25.4 h	20
MeridianLink	4.6 h	18.3 h	32
QuickNode	4.0 h	23.2 h	62
Voximplant	3.8 h	4.7 days	23
Vonage API	3.5 h	2.1 days	56
Bandwidth	3.4 h	21.4 h	71
Gemini	3.4 h	29.0 h	24
Digitalocean	3.3 h	21.2 h	30
Granicus	3.3 h	44.2 h	49
Plaid	3.3 h	22.0 h	21
Zenoti	3.2 h	20.9 h	29
Linode	3.1 h	24.6 h	26
Webroot	3.1 h	3.8 days	53
Sentry	3.0 h	6.3 h	29
Elastic	2.9 h	4.1 days	22
Harness	2.7 h	12.8 h	31
Flexera	2.7 h	2.1 days	29
The Linux Foundation	2.6 h	5.4 days	26
Circle	2.6 h	3.1 days	46
Hashicorp	2.5 h	22.9 h	20
Genetec	2.4 h	4.9 h	24
Sinch	2.3 h	16.7 h	61
Tailscale	2.3 h	7.7 h	24
Shippo	2.2 h	13.0 h	59
Temporal	2.2 h	10.7 h	31
Supabase	2.1 h	27.7 h	54
Coinbase Prime	2.1 h	16.8 h	41
Dbt Cloud	2.0 h	6.3 h	27
WatchGuard	2.0 h	17.3 h	23
IONOS	118 min	26.0 h	43
OpenAI	117 min	9.6 h	43
Snowflake	116 min	6.8 h	25
Cisco Umbrella	109 min	11.2 h	25
Grafana Labs	107 min	6.5 h	60
Array	101 min	4.8 h	31
Cycode	100 min	7.6 h	27
Cisco Systems	97 min	7.1 h	39
Flyio	92 min	5.8 h	52
BitGo	89 min	3.3 h	22
Scaleway	88 min	2.2 days	67
Matillion	84 min	7.2 h	20
Circle CI	82 min	6.1 h	29
Coinbase	82 min	2.3 days	61
Cloudflare	81 min	8.4 h	79
GitHub	79 min	5.2 h	56
Radware Cloud Waf	78 min	3.6 h	36
Alchemy	77 min	19.9 h	54
Amplitude	76 min	4.1 h	21
Whatnot	73 min	5.7 h	34
Expo	71 min	5.0 h	32
Caseware Cloud	66 min	5.2 h	25
Vercel	66 min	3.3 h	29
Box	63 min	2.5 h	20
Anthropic	61 min	3.4 h	54
Hostinger	57 min	4.1 h	69
Zoom	44 min	11.9 h	58
Fireblocks	41 min	3.1 h	29
GoDaddy	41 min	3.0 h	37
Ledger	23 min	19.8 h	55
Kore.ai	19 min	3.1 h	25
HeroCoders	11 min	12.6 h	28
Visma	11 min	8.7 h	55
Liveramp	10 min	85 min	46
Nitro	2 min	19 min	27

Only vendors with at least 20 incidents opened AND resolved in the last 90 days appear — a median over three incidents reads like a fact and is not one. The full per-incident table is the free outage-duration dataset (CC BY 4.0).

What is excluded, and why

Of every incident on record, these are dropped rather than estimated:

Rows	Excluded because
470	scheduled maintenance windows (a planned window is not an outage)
155	incidents still open, or with no resolve time posted
28	resolutions we inferred because the incident simply vanished
233	incidents the vendor posted as resolved BEFORE they started

Read this before quoting a number. Duration is the gap between the vendor's OWN posted timestamps. It is not measured downtime, not uptime, and not an SLA figure. A vendor that posts every small blip and closes it honestly will look "slower" here than one that posts nothing at all — this measures how vendors communicate as much as how fast they fix. Statuspage-hosted vendors back-fill their history, so they dominate the measurable set; platforms that expose only currently-open incidents contribute almost nothing, which is why we do not publish a per-platform comparison.

Watch your own vendors — free MIT template