Datadog synthetics cost calculator — real pricing from Checkly's data

TL;DR. Datadog browser checks cost $12 per 1,000 runs. API checks cost $5 per 10,000 runs. Sounds cheap. Then Checkly walks the math: 16 routes × 4 regions × every 4 minutes = $8,509/month, or $66,000/year, locked in. This post is a calculator-style walkthrough using Checkly's real numbers. We'll size three real teams (small SaaS, mid-market e-commerce, regulated fintech), show the line-item math, and end with the question every team should ask before signing a synthetics contract: does my user actually care if this route is checked from Tokyo every 4 minutes?

The Checkly post is the single best piece of public writing on synthetics pricing. They're a competitor to Datadog synthetics, so read with that lens — but the numbers are the numbers, and they cite Datadog's published pricing. We'll use them straight.

The unit economics

Datadog's synthetics pricing, as of April 2026:

• Browser check: $12 per 1,000 runs.
• API check: $5 per 10,000 runs.
• Multi-step API check: higher tier, varies.

A "run" is one execution from one location. So a single check configured to run from 4 regions every 5 minutes produces 4 runs per 5-minute window = 48 runs/hour = 1,152 runs/day = 34,560 runs/month.

That's one check. At browser-check pricing, that's $415/month for one check.

Now scale.

Checkly's example, walked

The $12-at-a-time post lays it out:

• 16 routes
• 4 regions
• every 4 minutes
• = 16 × 4 × (60/4) × 24 × 30 = 16 × 4 × 15 × 24 × 30 = 691,200 runs/month
• At $12/1,000 browser runs = $8,294/month, plus check overhead
• Checkly rounds to $8,509/month
• Annual: ~$102K, but Checkly cites $66K/yr because their reference team uses a mix of browser and API checks

The lesson: the multiplication is multiplicative, and the regions × frequency dimension is where the money goes. Going from 4 minutes to 5 minutes saves 20% of the bill. Going from 4 regions to 2 saves 50%. Most teams don't do the math — they pick "every 4 minutes" because it sounds reasonable, and "all regions" because it sounds thorough.

Team A — Small SaaS, 10 engineers, 1 product

Real numbers from a team we talked to in late 2025:

• 8 critical routes (homepage, signup, login, dashboard, billing, account, settings, logout)
• 2 regions (eu-west, us-east — they're EU-headquartered with US customers)
• Every 5 minutes
• All browser checks (they want pixel-level confirmation)

Math:

• 8 routes × 2 regions × 12 runs/hour × 24 × 30 = 138,240 runs/month
• $12/1,000 = $1,659/month
• Annual: $19,900

Verdict: tolerable, but it's their fourth-largest line item after APM, logs, and metrics. They could halve it by moving 6 of the 8 routes to API checks (homepage and signup stay browser).

Optimized: 2 browser × 2 × 12 × 720 = 34,560 runs at $12/1k = $415, plus 6 API × 2 × 12 × 720 = 103,680 runs at $5/10k = $52. Total: $467/mo. Saves $1,200/mo.

Team B — Mid-market e-commerce, 50 engineers, 4 brands

• 24 critical routes per brand × 4 brands = 96 routes (deduped to 60 after audit)
• 4 regions (eu-west, us-east, us-west, ap-southeast)
• Every 3 minutes (this is the killer)
• 80% browser, 20% API

Math:

• 48 browser × 4 × 20 runs/hour × 24 × 30 = 2,764,800 runs at $12/1k = $33,177/mo
• 12 API × 4 × 20 × 24 × 30 = 691,200 runs at $5/10k = $345/mo
• Total: $33,522/mo, $402K/year

Verdict: this is a renewal-shock candidate. The team's instinct will be to negotiate the rate. The right move is to audit the configuration first.

Optimized:

• Drop ap-southeast for 80% of routes (only the customer-facing checkout actually needs APAC coverage)
• Move from every 3 minutes to every 5 minutes for non-checkout routes
• Move informational routes (FAQ, about, terms) from browser to API

Result: ~$11k/mo. Saves $22k/mo, $264k/year. The audit takes a week.

Team C — Regulated fintech, 30 engineers, EU

• 40 routes (regulated industries inflate the route count — every legal disclosure page is "critical")
• 6 regions (regulators want EU + US + APAC + UK + LATAM + ME)
• Every 2 minutes (compliance team set this; nobody questioned it)
• 100% browser (they need rendering verification)

Math:

• 40 × 6 × 30 runs/hour × 24 × 30 = 5,184,000 runs at $12/1k = $62,208/mo
• Annual: $747K

Verdict: this is the bill that broke them. We talked to this team in Q3 2025 and they were paying close to this number.

Optimized: the right answer is to push back on the compliance team. "Every 2 minutes from 6 regions for a legal disclosure page" is a number nobody wrote down for a regulatory reason — it's accumulated configuration. The team got it down to 4 regions and 5 minutes for non-transaction routes. Saved $40k/mo.

The three knobs that matter

Synthetics cost is dominated by three multipliers:

1. Routes. Pick the routes that map to user-visible failures, not the routes that map to your sitemap. Most teams over-monitor.
2. Regions. Most users care about latency from one or two regions. The rest is theater.
3. Frequency. Every 5 minutes catches a 5-minute outage. Every 1 minute catches a 1-minute outage. Most outages are >5 minutes; the marginal value of high frequency is low.

If you cut each of these by 30%, you cut the bill by ~65% (because they multiply). Most teams have 50%+ slack on each.

When high frequency actually matters

Be honest about this. There are workloads where every-minute synthetics genuinely earn their cost:

• Payment processing (transaction failure costs > synthetics cost in <2 minutes)
• Real-time financial APIs (ms latency matters to users)
• Auth flows during peak load (1-minute outage = thousands of failed logins)

For everything else — homepage, dashboard, marketing pages, settings — every 5 minutes is fine. The synthetic is a smoke alarm, not a heart monitor.

What the alternatives look like

Checkly themselves have a different model — Playwright-based, included regions, more predictable monthly. UptimeRobot, Pingdom, and Better Stack all have different shapes; some are flat-rate-per-check, some are run-based like Datadog.

The Sutrace synthetic offering is included in the plan tier, with a regional fan-out cap. We don't bill per-run for the in-plan regions. The reason: the actual cost to run a check is dominated by the egress and the headless-browser compute, both of which are knowable and bounded. The Datadog model is priced like AWS Lambda but the underlying cost shape isn't that.

If you want the side-by-side, the Datadog alternatives page has the breakdown.

The audit script

Run this before your next renewal:

# Pseudocode
for check in datadog_synthetic_checks:
    runs_per_month = check.regions * (60 / check.interval_min) * 24 * 30
    cost = runs_per_month * (12/1000 if check.type == "browser" else 5/10000)
    if check.last_query_in_dashboard > 30:
        flag("unused", check, cost)
    if check.last_alert_fired > 90:
        flag("never-fires", check, cost)
    if check.regions > 2 and check.route not in CRITICAL_ROUTES:
        flag("over-regioned", check, cost)

Most audits surface 30–50% of synthetics as candidates for deletion or downgrade. The "never-fires" bucket alone is usually 20% of the bill — these are checks added during a launch in 2022 and never reviewed.

Negotiation lever

If you're staying on Datadog, the audit is your renewal-negotiation ammunition. Walk in with: "We have 60 checks. After audit, 22 are eligible for deletion or downgrade. The remaining 38 use 4 regions × 4-minute frequency. Our usage will be 40% lower next year. Adjust the contract."

This works. Datadog AEs would rather discount the deal than lose the customer.

The deeper point

Synthetics is the cleanest illustration of the larger Datadog pricing pattern: the unit price looks reasonable, the multipliers are invisible to the team, and the bill is set by configuration choices nobody documented. The fix isn't a different vendor — the fix is making the multiplier visible at configuration time.

When you add a check, the UI should show you the projected monthly cost. When you change the frequency from 5 to 4 minutes, the UI should show you that you just added $1,200/year. Datadog could ship this in a sprint. They don't, because per-run revenue is a feature.

What we ship

Sutrace synthetics show the projected monthly cost at configuration time. Region selection has a recommended-default (your nearest two) and an explicit cost label per additional region. Frequency has a recommended-default (5 minutes) and a cost-per-step label.

We're not radically cheaper at the unit level. We're cheaper because the UI prevents the configurations that produce $8,509/month bills.

Closing

If you remember one number from this post: 16 routes × 4 regions × every 4 minutes = $8,509/month. That's Checkly's example, and it's the canonical synthetics renewal-shock equation.

Run the audit this week. The savings are real, they're predictable, and they don't require a vendor migration. If you want to compare against Sutrace pricing, the public tiers include synthetics with regional caps, and we'll size your specific workload before any contract.