Hotjar vs FullStory vs LogRocket vs Clarity: Replay Compared
If you want the highest-fidelity replay, FullStory wins because it captures 100% of sessions with no sampling. If you want free and genuinely unlimited, Microsoft Clarity is the pick, as long as you can live with a 30-day retention window. In the Hotjar vs FullStory debate specifically, the split is stark: Hotjar is now the marketer's heatmap tool sitting under Contentsquare, while FullStory is the full-capture premium option. LogRocket is the developer's debugger. Here's the one-liner most buyers miss: Hotjar samples your sessions and Clarity time-limits them, but FullStory doesn't. That single fact reshapes what "session replay" actually means depending on which tool you install.
The eval that started this
I installed all four on a real B2B account pushing real traffic. My own credit card on the paid trials, my own event volume, no vendor sandbox. The plan was simple: run them side by side for a month and see whose replays I'd actually reach for when a signup form broke.
The moment that reframed the whole test came about ten days in. Hotjar's dashboard showed a session count climbing toward six figures, exactly the kind of number that makes you feel covered. Then I went to pull replays for a specific drop-off and the recorded-session count was a small fraction of that. I'd been reading an analytics number as a replay number.
According to FullSession's teardown of the Hotjar free plan, that 200,000-session figure is analytics. Only about 5% of sessions get recorded as replays, capped at 10,000. So the gap I hit wasn't a bug. It's the model. "Session replay" sounds like it means every session, but for half these tools it doesn't, and the difference shows up exactly when you're debugging something rare.
How I scored them
I graded all four on the same six axes so the comparison stays honest. Here's the frame. Which axis you weight most heavily depends entirely on who you are.
- Replay fidelity — do you get every session, or a sample? Weight this if you debug rare, high-value flows like checkout or onboarding.
- Heatmap depth — click, scroll, and zone-level aggregation. Weight this if you're a marketer or CRO running page-level tests.
- Console / error capture — stack traces, exceptions, network payloads. Weight this if engineers are the primary users.
- Privacy masking — what's hidden by default, and what leaves your infrastructure. Weight this if procurement or legal has a say.
- Pricing — not the headline number, the real bill. Everyone weights this, few read it closely.
- Mobile support — native iOS/Android replay versus web-only. Weight this if your product is an app first.
Same axes, four tools, no grading on a curve for the free one. A marketer and a backend engineer will rank these in opposite order, and that's the point. The "best" tool is a function of which two axes you actually care about.
The 4-way matrix
| Axis | Microsoft Clarity | Hotjar | FullStory | LogRocket |
|---|---|---|---|---|
| Replay fidelity | Unlimited recordings, 30-day retention | Sampled, ~5% recorded, capped at 10k replays | 100% capture, no sampling | Full capture with technical depth |
| Heatmap depth | Basic click/scroll | Click, scroll, zone-based (zone now Growth tier) | Included | Secondary to debugging focus |
| Console / error capture | None | None | Available | Deep: console, network payloads, JS exceptions, Redux/Vuex state |
| Privacy masking | Input masking on by default | Configurable | Configurable | Configurable; payload capture widens surface |
| Pricing | Free, unlimited | Module-based, billed via Contentsquare | Quote-based, no public entry price | Paid tiers |
| Mobile support | Web-first | Web-first | Web-first | Web + some mobile |
| Data routing flag | Routes through Microsoft infrastructure | — | — | Captures API payloads + app state |
A quick note on where a fifth column would sit. A tool like Kixo would add a native mobile replay row that all four of these leave mostly empty. More on that later, and treated with the same scrutiny.
Two things to flag inline. Clarity and Hotjar both limit what you keep, Clarity by time and Hotjar by count, while FullStory is the only one in the set that retains everything. On governance, the two tools that draw the most reviewer attention are Clarity (data through Microsoft's infrastructure, per Inspectlet's roundup) and LogRocket (payload and state capture, per Lokker).
Microsoft Clarity, free and unlimited
The pricing gotcha is easy here because there's no bill. So the gotcha moves. It becomes the 30-day retention window and the fact that your session data routes through Microsoft's infrastructure. Microsoft's own docs are direct about retention: Clarity keeps recordings for 30 days from capture, and only recordings you explicitly favorite stick around for up to nine months. If you need to look back at a cohort from last quarter, that data is gone.
On privacy, Clarity earns a genuine point. Per the ClarityInsights guide, it masks all input content by default. Password fields, text inputs, and textareas render as placeholder dots instead of real text, and there's a data-clarity-mask attribute for custom sensitive fields. For a free tool, default-on masking is more than some paid competitors ship.
The feature ceiling is where "free" shows its edges. Inspectlet's independent testing notes Clarity has no form analytics, no A/B testing, no surveys, no error logging, and no advanced search or filtering. Heatmaps are basic. You get click and scroll maps and unlimited recordings, and that's the deal. For a lot of teams, especially early-stage ones just trying to watch how people use a new page, that deal is completely fine.
What the Contentsquare merger actually changed for Hotjar
Money first, because at Hotjar the money story is the story now. As of July 2025 the Contentsquare merger closed, and per SaaS Price Pulse, Hotjar's pricing page redirects to contentsquare.com. Features are split across separately-billed modules, Observe, Ask, and Engage, so the single clean plan you remember is gone. If you're anchoring a budget on one line item, you'll be reconciling several.
The second gotcha is the one that started this whole article. That "200,000 free sessions" number is analytics, not replays. FullSession's breakdown puts it plainly: only about 5% of sessions get recorded, capped at 10,000 replays. So a busy month gives you a session dashboard that looks generous and a replay library that runs dry before you've reviewed a fraction of your traffic. Set against FullStory's 100% capture, that's the whole Hotjar vs FullStory decision in miniature.
To be fair about the direction of travel: in December 2025, per PricingSaaS, Hotjar raised the free-plan session limit tenfold, from 20,000 to 200,000, and doubled replay capture from 5,000 to 10,000. Real improvements. The same update moved zone-based heatmaps and Journey Analysis from Pro up to the Growth tier, so some capabilities got more expensive to reach even as the free ceiling rose.
Underneath the module reshuffle, Hotjar still does what made it popular: heatmaps, recordings, surveys, feedback widgets. If you're a marketer who wants to watch how people scroll a landing page and ask them why they bounced, it's a strong fit. Just go in knowing you're buying a heatmap-and-feedback suite that samples replays, not a full-capture replay tool.
FullStory, the no-sampling premium option
The pricing gotcha at FullStory is the absence of one to anchor against. It's quote-based, with no public entry price, so you can't ballpark the bill from a pricing page. You'll be in a sales conversation before you see a number, and where you land depends on your session volume and how you negotiate.
What you're paying for is the thing this article keeps circling back to. Fidelity. FullStory's own writing on session replay makes the distinction explicit: some tools capture a sample, and FullStory captures them all. That's 100% capture, no sampling, which means when a rare bug surfaces, the session is already there. You're not hoping it fell inside a 5% window.
The subtler benefit is durability. FullStory caches page assets (images, CSS, HTML) so a replay recorded before a redesign still renders correctly months later. Anyone who has tried to review an old recording only to see a broken layout because the CSS changed underneath it will understand why cached assets matter. It's the difference between a replay you can trust and a ghost of one.
If replay is your primary tool rather than a nice-to-have, FullStory is the strongest technical answer in this set. You'll just pay for it, and you won't know exactly how much until you ask.
LogRocket, session replay for people who read stack traces
On pricing, LogRocket runs paid tiers, and the real trade-off isn't cost. It's what "capture" means here. Per Lokker's writeup, LogRocket records sessions, captures network requests including API payloads, logs Redux or Vuex application state, and replays errors alongside console output and JavaScript exceptions. For an engineer chasing a reproduction, this is the richest picture in the group. You see the click, the failed request, the payload, and the state the app was in when it broke.
That same depth is the privacy conversation. Capturing API payloads and application state means sensitive data can flow into replay unless you configure masking carefully, and it's exactly the surface reviewers flag. Not a flaw so much as a consequence of the design goal. A debugger, by nature, wants to see everything.
So LogRocket is a specialist. If your primary users are developers who read stack traces for a living, it's arguably the best-fit tool here. If your primary users are marketers, most of that technical depth is noise, and the governance overhead isn't worth it.
Worked example: what sampling costs you in a real month
Let's make sampling concrete. Take one site doing 200,000 sessions in a month (round numbers, easy math) and ask what each tool actually gives you back.
- Clarity: records all 200,000, but only the last 30 days are retained (favorites survive up to nine months). Coverage is total; the constraint is the clock.
- Hotjar (free): the dashboard shows 200,000 analytics sessions, but only ~5% get recorded as replays, capped at 10,000. So 10,000 replays out of 200,000, one in twenty. The other 190,000 sessions were counted, not captured.
- FullStory: records all 200,000, no sampling. The rare session is in there.
Now the scenario from my eval. A signup bug hits roughly 1% of sessions, so 2,000 sessions in this month. With FullStory, every one is recorded. Pull the filter, watch the failures. With Hotjar's free tier, your 10,000 replays are effectively a random 5% draw, so on average you'd capture about 100 of those 2,000 broken sessions. Sometimes fewer. With Clarity you'd have all of them, as long as the bug happened inside the last 30 days.
That's the whole point in one paragraph. "Sampling" sounds academic until a customer reports something you can't reproduce and the replay you need was one of the nineteen in twenty that never got saved.
If you're weighing this against product-analytics platforms too, the same coverage question shows up in that category. I get into it in our Mixpanel, Amplitude, PostHog and Heap head-to-head.
Privacy and data governance, the two talking points
Two flags come up in every procurement review of this set, and they belong to two different tools.
The first is Clarity's data routing. It's free because your session data flows through Microsoft's infrastructure, and Inspectlet's roundup names this as the trade-off behind the no-cost model. For many teams that's acceptable. For anyone with strict data-residency or third-party-processing requirements, it's a conversation with legal before you install the script. Worth setting against that: Clarity's default-on input masking, from Microsoft's docs and the ClarityInsights guide, is a real point in its favor. It hides typed content out of the box rather than making you opt in.
The second flag is LogRocket's capture surface. Recording API payloads and Redux/Vuex state, per Lokker, means sensitive fields can end up in a replay unless masking is configured deliberately. The capability that makes LogRocket great for debugging is the same one that makes governance teams ask harder questions.
Even-handed read: Clarity's concern is where data goes, LogRocket's is what data gets captured. Neither is disqualifying. Both should be on the checklist before a contract, not after. how session-replay masking actually works is worth a read if this is the axis that decides your purchase.
Where mobile-native replay fits (and one option worth knowing)
Here's the honest gap in this whole comparison: it's web-first. All four tools built their reputations on web session replay, and native iOS/Android replay is thin across the set. If your product is a mobile app before it's a website, none of these four is an obvious fit for watching what happens inside the app itself.
That's the context for mentioning Kixo, evaluated on the same rubric, not as a spine. Kixo does web replay via rrweb and adds native iOS and Android session replay alongside it, plus heatmaps and privacy masking, sitting next to a product-analytics layer that covers events, funnels, retention, cohorts and user flows. Its angle is chat-first: you ask questions in plain language and it generates the charts, with a visible reasoning trail. Pricing is per-project across FREE, GROWTH, and ENTERPRISE tiers, bracketed by monthly active users.
The maturity caveat, stated plainly: Kixo is a newer, AI-native platform, not a decade-old category incumbent like the four above, and I haven't run it through the same month-long eval. So treat it as one option to shortlist if native mobile replay plus product analytics in one place is your requirement, not as a proven replacement for FullStory's fidelity or Clarity's zero cost on web. For teams whose whole question is mobile, it's at least worth a demo where these four aren't. If mobile-first is where you're headed, the mobile session replay landscape covers the wider set of options.
Verdicts: who should buy each (and who should not)
Microsoft Clarity. The default for anyone who wants to watch real sessions without a bill, with default-on masking as a genuine bonus. Don't buy it if you need error logging, advanced filtering, or the ability to look back past 30 days. The retention window will cost you exactly when you need history most.
Hotjar. Still the marketer's heatmap-and-feedback suite, and the December 2025 limit increases made the free tier more usable. Don't buy it if you want every session recorded or a single clean line of pricing. Post-merger it samples replays and bills across Contentsquare modules.
FullStory. The fidelity winner, full stop, thanks to 100% capture and cached assets that keep old replays intact through redesigns. Don't buy it if you're budget-constrained or you only need marketing heatmaps. You'll pay premium money for capture depth you may not use.
LogRocket. The engineer's replay tool, with console logs, network payloads, and app-state capture that make reproductions fast. Don't buy it if your users are non-technical or your data governance can't accommodate payload and state capture.
Here's the thing this all comes down to. The four split into two camps. Clarity and Hotjar are the free-or-freemium, sampled-or-time-limited camp for teams watching behavior on a budget. FullStory and LogRocket are the paid, full-capture camp for teams that treat replay as core infrastructure. Pick your camp first. The tool falls out of that decision.
FAQ
Which session replay tool is best? There's no single winner. FullStory is best for replay fidelity with 100% no-sampling capture, while Microsoft Clarity is best for cost since it's free and unlimited. Pick by which of those two axes matters more to you.
Does Hotjar record every session? No. Per FullSession's analysis, Hotjar's advertised 200,000 free-tier sessions is an analytics figure, and only about 5% of sessions are actually recorded as replays, capped at 10,000.
Is Microsoft Clarity really free? Yes, Clarity is free with unlimited recordings, but per Microsoft's docs it retains recordings for only 30 days by default (favorites up to nine months), and your data routes through Microsoft's infrastructure.
FullStory vs LogRocket, which for developers? LogRocket is the more developer-focused of the two. Per Lokker it captures console logs, network payloads, JS exceptions, and Redux/Vuex state. FullStory leans toward product-team fidelity with 100% capture and cached assets for durable historical replay.