Simple pricing for small SaaS teams
Start tiny. Cover your most important journeys. Expand when it's paying for itself.
Starter
Best for solo founders and early MVPs
3 workflow checks
Runs every 15 minutes
Slack or email alerts
Screenshots on failure
7-day data retention
Team
Best for small teams shipping weekly
15 workflow checks
Runs every 5 minutes
Slack + email + webhooks
Screenshots on every run
30-day data retention
Multiple environments
Growth
Serious monitoring across environments
50 workflow checks
Runs every 1 minute
Advanced alert routing
Multi-region checks
90-day data retention
Priority support
Team roles & permissions
What counts as a “check”?
A check is one workflow journey (e.g. your login path). Each check runs on a schedule and counts as one check regardless of how many steps it has.
You can pause checks anytime. Paused checks don't count toward your limit.
Frequently asked questions
Yes, upgrade or downgrade anytime. Changes take effect immediately and billing is prorated.
Not yet, but it's on our roadmap. Monthly billing means you can try it risk-free and cancel anytime.
Yes. On Team and Growth plans, you can run checks against multiple environments (production, staging, etc.).
We'll let you know and give you the option to upgrade. We won't suddenly stop your existing checks.
A check is the workflow you define (e.g., “Login → Create invoice → Download PDF”). A run is each execution of that workflow on a schedule or trigger. Most pricing models in the market charge per check (how many workflows) and/or per run (how often you execute them), sometimes with add-ons for locations, retries, and higher-frequency schedules.
Uptime tools typically price per monitored URL/host with simple HTTP/TCP checks. Workflow/transaction monitoring often prices by runs because browser/API steps consume more compute and generate more telemetry (screenshots, traces, logs). For small SaaS teams, the key cost driver is usually frequency (e.g., every 1 minute vs every 5–10 minutes) and number of locations, not the number of teammates.
Start by mapping workflows to business risk. Revenue-critical flows (signup, login, checkout, billing) often justify 1–5 minute cadence; internal/admin flows can run every 15–60 minutes. A practical approach is tiered frequency: run a fast API smoke check every 1–2 minutes and a full browser workflow every 5–15 minutes. This catches outages quickly while keeping run volume predictable.
In many tools, yes. Retries can multiply run consumption, multi-region execution can multiply it again, and rich artifacts (screenshots/video/har files) may be included or may increase storage/retention costs. When comparing plans, look for how “run units” are counted (scheduled run + retries, per-location multipliers) and what’s included for artifact retention.
Per-check pricing is predictable when you add workflows slowly but can discourage monitoring “nice-to-have” flows. Per-run pricing is predictable when you keep frequency and locations stable, but costs can jump if you move from 5-minute to 1-minute checks or add regions. The most small-team-friendly setup is usually: a small number of high-value workflows, clear caps on run volume, and simple add-ons when you need more frequency or locations.
Yes, but do it carefully. Use dedicated test accounts, sandbox environments (e.g., Stripe test mode), and stable test data. Avoid high-frequency actions that look like abuse (e.g., repeated password resets, repeated card authorizations). Prefer read-only or idempotent steps where possible, and rate-limit workflows. For auth providers, use test tenants and avoid MFA flows unless you can automate them safely.
Look for pricing that’s transparent about overages or auto-upgrades. In practice, small teams often increase frequency during incidents, which can cause unexpected run consumption. A good pricing model either (1) allows burst capacity with clear overage rates, (2) provides hard caps with graceful throttling, or (3) supports incident-mode schedules that are time-boxed so you can safely increase coverage without a surprise bill.
By the numbers
Organizations using SRE practices commonly target 99.9%+ availability for user-facing services, which still allows ~43 minutes of downtime per month at 99.9%.
Google SRE Book (Site Reliability Engineering), Google (2016)The average cost of downtime is commonly estimated at $5,600 per minute (often-cited benchmark across industries).
Gartner (widely cited downtime cost estimate) (2014)A large share of outages are caused by changes (deployments/configuration), making post-deploy workflow checks especially valuable.
Google Cloud, "The State of DevOps" / DORA research (change-related failure insights) (2019)Synthetic monitoring adoption has increased as teams shift left on reliability; modern observability reports consistently highlight proactive testing (synthetics) as a complement to real-user monitoring.
Datadog, State of Observability / Monitoring reports (2023)Real-world examples
Pricing sanity check: 5 core workflows at 5-minute cadence
Scenario: A 6-person SaaS monitors: (1) login, (2) signup, (3) create subscription, (4) generate invoice, (5) export CSV. They run each workflow every 5 minutes from 1 region, with 1 retry on failure.
Outcome: Approx. 5 workflows × 12 runs/hour × 24 hours = 1,440 scheduled runs/day. With occasional retries, they budget ~1,600 runs/day. This keeps costs predictable and focuses spend on revenue-critical paths.
Incident-mode burst without surprise costs
Scenario: During a payment incident, the team temporarily increases the “Checkout → Stripe payment → Receipt email” workflow from every 10 minutes to every 1 minute for 2 hours to validate mitigation and detect regressions.
Outcome: They add 120 extra runs (1 run/min × 120 min) for a single workflow—easy to forecast. With clear per-run pricing or a burst allowance, this is a controlled, low-risk way to gain fast feedback during incidents.
Multi-location coverage for a global customer base
Scenario: A B2B SaaS with EU and US customers runs the login + dashboard load workflow from 2 regions (US-East and EU-West) every 10 minutes to catch regional CDN/DNS issues and auth routing problems.
Outcome: Run volume roughly doubles for that workflow (per-location multiplier). The team catches a region-specific OAuth callback failure within 10 minutes instead of waiting for customer tickets, reducing time-to-detect and support load.
API-first smoke + browser workflow for cost control
Scenario: A team splits monitoring into: an API token-auth smoke check every 1 minute, and a full Playwright browser workflow (login → create record → verify UI) every 15 minutes.
Outcome: They get near-real-time outage detection from the cheap API check while keeping heavier browser runs 15× less frequent. This pattern typically reduces synthetic run consumption materially while preserving confidence in end-to-end UX.
Key insights
1.
For small SaaS teams, the biggest synthetic monitoring cost lever is frequency (1-minute vs 5/10/15-minute schedules) rather than number of workflows.
2.
Workflow/transaction monitoring delivers value when it mirrors real customer paths; prioritize revenue and access flows first (auth, billing, core CRUD, exports).
3.
Retries and multi-region execution can silently multiply run usage; pricing that clearly defines how runs are counted prevents surprise bills.
4.
A hybrid strategy (fast API smoke + slower full browser workflow) is a common best practice to balance detection speed and cost.
5.
Change-driven incidents are common; post-deploy synthetic runs (or temporary increased cadence) are one of the simplest ways to reduce time-to-detect regressions.
6.
Artifact retention (screenshots, logs, traces) matters during incidents; teams should check whether retention is included or metered separately.
7.
Simple, team-friendly pricing often means: transparent run accounting, easy add-ons for more frequency/locations, and alert routing included (Slack/email) without per-seat fees.
Pro tips
💡
Start with 3–5 workflows that map directly to revenue or access (login, signup, checkout/billing, core action, export). Run them every 5–10 minutes from one region, then expand cadence/regions only after you’ve stabilized.
💡
Control run volume with a two-layer approach: a cheap API smoke check every 1–2 minutes for fast detection, plus a full browser workflow every 10–30 minutes for true end-to-end confidence.
💡
Create dedicated test tenants and idempotent test data (e.g., prefix records with “synthetic-”), and add cleanup steps. This prevents synthetic checks from polluting production analytics and reduces flaky failures from data collisions.
How CheckyWorky compares
vs Datadog Synthetics
Powerful enterprise platform with deep observability integration; pricing can become complex as you add browser test runs, multiple locations, and higher frequency. CheckyWorky positions for small teams with simpler workflow-first packaging and straightforward Slack/email alerting.
vs Checkly
Developer-centric synthetics with strong Playwright support and usage-based pricing. CheckyWorky differentiates by focusing on “pretend customer” SaaS workflows and keeping pricing easy to reason about for small teams adding checks gradually.
vs UptimeRobot
Great low-cost uptime monitoring for simple endpoints (HTTP/ping/keyword). It’s not designed for multi-step SaaS transactions (auth, billing, CRUD). CheckyWorky is aimed at end-to-end workflow checks where a 200 OK isn’t enough.