UserProof

Session replay sampling calculator for founders on a free plan

Updated November 17, 2026 6:00 AM ET · Allowances read November 17, 2026 · Capture rules read November 16, 2026 · 10 free plans

For founders choosing a free session replay plan by how much of their traffic it records: at 50,000 visits a month, Microsoft Clarity’s free plan records the most (100%). FullStory’s records 30,000 (60%), UserProof’s about 16,250 (33%) and Hotjar’s 2,500 (5%). Sentry’s snippet defaults to 10% of sessions. Enter your own monthly visits below.

By Ryan Doyle, founder of UserProof. Published ; allowances rebuilt every morning from each vendor’s pricing page.

UserProof publishes this calculator and is one of the 10 tools in it. Its row comes from its own plan code; every other row links to the vendor’s page.

Showing 50,000 visits a month. Change the number to recalculate.

Free plans at 50,000 visits a month, most recorded first
Tool and free planVisits recorded a monthShare of your visitsThe rule
Microsoft Clarity Free50,000100%It states no traffic limit and no sampling, and its FAQ caps recording at 100,000 sessions per project per day (we count 30 even days). Microsoft Clarity lists 30 days of data retention.
FullStory Free30,00060%Records up to 30,000 sessions a month. FullStory lists 12 months of data retention.
UserProof Free ours16,25033%Every visit up to 5,000 a month, then about 1 in 4 visits until the month resets (visitors picked by a hash, so real traffic lands near this number). Keeps recordings 7 days.
Amplitude Free10,000 replays20%Records up to 10,000 replays a month. Amplitude lists 30 days of data retention.
PostHog Free5,000 recordings10%Records up to 5,000 recordings a month. PostHog lists 30 days of data retention.
Hotjar (Contentsquare) Free2,5005%Replays 5% of sessions, at most 10,000 a month (the plan tracks up to 200,000 sessions for its other features). Hotjar (Contentsquare) lists 30 days of data retention.
LogRocket Free1,0002%Records up to 1,000 sessions a month.
OpenReplay Free cloud1,0002%Records up to 1,000 sessions a month. Also free to self-host with no stated limit.
Mouseflow Free5001%Records up to 500 sessions a month. Mouseflow lists 30 days of data retention.
Sentry Developer50 replays0.1%Records up to 50 replays a month. New users get 5,000 a month for their first 3 months; this page uses the lasting 50. Sentry lists 30 days of data retention.

Quick facts

Which free session replay plan records the most visits?
Microsoft Clarity at 50,000 visits a month. It states no traffic limit and no sampling, and its FAQ caps recording at 100,000 sessions per project per day (we count 30 even days). Microsoft Clarity lists 30 days of data retention. Among plans with a monthly cap, FullStory Free records the most: 30,000 sessions.
What share of sessions does Hotjar’s free plan record?
5% of sessions, at most 10,000 replays a month. At 50,000 visits that is 2,500; at 250,000 it is 10,000.
How many replays does Hotjar’s Growth plan record?
Every session up to 20,000 a month, then 20,000 replays from 25,000 to 120,000 sessions, 22,000 at 150,000 and 30,000 at 200,000 (Hotjar’s own slider data, read November 16, 2026).
What is Sentry’s default replay sample rate?
Its Loader Script starts with replaysSessionSampleRate 0.1, which records 10% of sessions, and replaysOnErrorSampleRate 1, which keeps a replay for every session with an error. You can change both in Sentry.init.
Can you set a sampling rate in PostHog?
Yes. Each trigger group has its own sample rate, and the choice is made from the session ID, so a session is either recorded throughout or not at all. PostHog recommends starting at 100% and lowering it as needed.
What happens past UserProof’s free limit?
Nothing is cut off. After 5,000 recorded visits in a month, UserProof keeps recording about 1 in 4 visits, picked by visitor, until the month resets. At 50,000 visits that is about 16,250 recorded.

Share recorded at six volumes

The same rules at 1,000 to 250,000 visits a month. A capped plan records everything until its cap, so its share falls as traffic grows; a plan that samples a fixed share never records more than that share.

Share of visits each free plan records, at six monthly volumes
Tool and free plan1,0005,00010,00050,000100,000250,000
Microsoft Clarity Free100%100%100%100%100%100%
FullStory Free100%100%100%60%30%12%
UserProof Free ours100%100%63%33%29%27%
Amplitude Free100%100%100%20%10%4%
PostHog Free100%100%50%10%5%2%
Hotjar (Contentsquare) Free5%5%5%5%5%4%
LogRocket Free100%20%10%2%1%0.4%
OpenReplay Free cloud100%20%10%2%1%0.4%
Mouseflow Free50%10%5%1%0.5%0.2%
Sentry Developer5%1%0.5%0.1%0.05%0.02%

Hotjar Growth replays at each price step

Hotjar’s pricing page lists Growth’s capture as “15% sessions captured (min. 20k)”. Its own slider data shows what that means: every session is replayed up to 20,000 a month, then replays stay at 20,000 until the slider reaches 150,000 sessions, where Hotjar lists 22,000.

Hotjar Growth: replays per month at each step of its price slider
Monthly sessionsReplays recordedShare recorded
7,0007,000100%
10,00010,000100%
15,00015,000100%
20,00020,000100%
25,00020,00080%
30,00020,00067%
35,00020,00057%
45,00020,00044%
60,00020,00033%
75,00020,00027%
85,00020,00024%
100,00020,00020%
120,00020,00017%
150,00022,00015%
200,00030,00015%

Read from the plan data on contentsquare.com/pricing on November 16, 2026. Above 200,000 sessions the page says “Let’s talk”. Pro and Enterprise list “Up to 100% sessions captured” at custom prices.

Sampling you can set yourself

How to read the result

“Visits recorded” is how many of your monthly visits the free plan keeps a replay of. Some tools count recordings or replays instead of sessions; each is close to one visit but not always the same, and the table names the unit when it differs. The share is visits recorded divided by the visits you entered.

A high share matters most when traffic is low: at 2,000 visits a month, a 5% sample is 100 recordings. At high traffic Clarity’s free plan records far more than ours does; if you want the most replays for nothing, start there. FullStory’s free plan keeps its data 12 months. Retention differs a lot, so the table lists how long each free plan keeps recordings.

If you build with Claude Code, Codex or Cursor and want the recordings turned into one fix with the file and line, UserProof does that. Free is 3 sites and 5,000 recorded visits a month; Pro is $29 a month for 50,000; Scale is $99 a month for 250,000 (plans). It has no heatmaps, surveys or replay dashboard for analysts. To compare what the paid plans cost at your traffic, use the session replay cost calculator; for the free plans side by side, see the best free session replay tools.

Method

Limits. Real traffic is uneven, so a daily cap can bite on a busy day before a monthly total suggests. Sentry gives new users 5,000 replays a month for their first 3 months; the table uses its lasting 50. Paid plans are not in the main table.

Left out:

Cite this

UserProof, “Session replay sampling calculator,” https://getuserproof.com/session-replay-sampling-calculator, allowances read November 17, 2026, capture rules read November 16, 2026. Source: each vendor’s pricing page and docs.

FAQ

What is a session replay sampling rate?

The share of visits a replay tool records. A tool that samples 10% records about 1 visit in 10 and drops the rest. Free plans often set it for you through a monthly cap or a fixed share; some tools let you set your own rate in code or settings.

Is a sample enough to find what stops visitors?

Often, if the sample is large. A problem that hits 1 visit in 20 shows up about 5 times in 100 recordings and about 50 times in 1,000. A rare problem, or one on a page few people reach, can be missed by a small sample. Check the visits recorded column, not only the share.

How often is this updated?

The allowances are re-read from each vendor’s pricing page every morning and the page is rebuilt from them. The capture rules that pricing pages don’t list as a number (Hotjar’s share and maximum, Clarity’s daily cap, Sentry’s and PostHog’s sampling) were read on November 16, 2026.

Is this calculator independent?

No. UserProof publishes it and sells one of the tools in it. Every other number links to the vendor’s own page so you can check it.

Sources, each with the date we read it

Corrections

If a number here is wrong, the vendor’s page is the source of truth. Report it through our GitHub issues.