Hotjar vs Microsoft Clarity: UX Analytics for Commerce Sites in 2026

Nobody has ever shipped a conversion improvement because they watched a session recording.

They shipped it because a recording confirmed a hypothesis they already had, or contradicted one they were about to bet a sprint on. Session replay and heatmaps are not discovery tools in any reliable sense. They are evidence tools, and evidence is only worth what it costs you to collect and how much of it you can trust.

Hotjar vs Microsoft Clarity is usually framed as free versus paid, which makes it sound like a budget decision. It is not. The two products have diverged into genuinely different categories: one is a research platform that happens to include heatmaps, the other is a behavioral instrumentation layer that Microsoft gives away because your data has strategic value to Microsoft. Picking correctly means knowing which of those you actually need.

Hotjar vs Microsoft Clarity Is a Question About Sampling and Intent

Every behavioral analytics tool answers two questions. What did users do, and why did they do it. The tools split cleanly on which one they are built to answer.

Clarity is built for the first question, at full volume. No traffic ceiling, no sampling, no session budget to ration. Every session on your storefront is eligible for capture, which matters enormously when you are debugging a checkout failure that affects two percent of mobile Safari users. A sampled tool will show you nothing. An unsampled tool will show you eleven sessions.

Hotjar is built for the second question. Surveys, feedback widgets, and interview recruiting sit alongside the replay data, so a heatmap anomaly can be followed immediately by asking the affected segment what confused them. That loop is the product, and no amount of free replay volume substitutes for it.

The trap is that both tools demo identically. Both show heatmaps, both show recordings, both flag rage clicks and dead clicks. In a thirty minute evaluation they look like the same product at two prices, and teams choose on price. Six months later the team that picked on price discovers it bought the wrong category.

For commerce specifically there is a third variable that most comparisons skip entirely, which is what the tool does to your page weight and your consent posture. We will come back to that, because on a storefront it frequently outranks both feature sets.

Microsoft Clarity: Unlimited Volume, and the Question of Why

Clarity is free, and has been consistently described by Microsoft as free forever with no traffic limit and no sampling. For a commerce site doing meaningful volume this is not a small thing. Paid replay tools price on sessions, so the sites that most need behavioral data are exactly the sites that can least afford to capture all of it.

What you get: session recordings, click and scroll heatmaps, rage click and dead click detection, JavaScript error correlation, and segmentation by device, browser, referrer, and custom tags. The GA4 integration is genuinely useful, letting you jump from a GA4 segment into the matching recordings rather than hunting manually. Copilot style summarization over recording sets has matured into something that saves real hours when you are triaging a hundred sessions rather than watching three.

The constraint that actually bites is retention. Playback data is kept for roughly 30 days, with favorited or labeled sessions and heatmap data retained substantially longer at around nine months. Thirty days is fine for active debugging and useless for seasonal analysis. If you want to compare this year's Black Friday checkout behavior to last year's, Clarity will not have the recordings, and the workflow that saves you is remembering to favorite sessions before the window closes. Teams discover this in November, which is the worst possible month to discover it.

The second consideration is strategic rather than technical. Clarity is free because behavioral data at internet scale is valuable to Microsoft, and their terms permit use of collected data for their own purposes. For most merchants this is an acceptable trade. For a brand whose merchandising behavior is genuinely proprietary, or one operating under procurement rules that scrutinize data processors, it is a conversation with legal rather than a conversation with the analytics team.

Free tools are not cheaper. They move the cost from your budget line to your risk register, which is fine as long as somebody is actually reading the risk register.

Clarity's masking defaults are reasonable but not sufficient for commerce out of the box. Anything that renders customer data in the DOM, order confirmations, saved addresses, partial card details, loyalty balances, needs explicit masking rules before you enable capture on authenticated pages. The default behavior is safer than most teams assume and less safe than compliance will assume.

Hotjar: Research Workflow, Now Inside Contentsquare

Hotjar was acquired by Contentsquare, and by 2026 the integration is no longer cosmetic. What was one product with one bill has been reorganized into separately priced lines covering experience analytics, voice of customer, and product analytics. Published entry pricing for the experience analytics tier sits in the neighborhood of $49 per month for a small session allowance, scaling to several hundred per month in the fifty thousand session range and past $700 per month at two hundred thousand sessions. The free tier was raised substantially in late 2025 and is now genuinely usable for a small storefront.

That repackaging is the single most important fact in this comparison, and it cuts both ways.

The upside is a real upgrade path. A merchant who outgrows heatmaps can move into zone based analytics, impact quantification, and journey analysis without replatforming their instrumentation. If you expect to be running a dedicated experience team in two years, starting inside that ecosystem saves a migration.

The downside is that budget line unbundling tends to precede budget line inflation, and the survey tooling that made Hotjar distinctive now prices separately from the replay tooling that made it familiar. Evaluate the combination you actually need, not the entry price of the piece you looked at first.

Where Hotjar clearly wins is the qualitative loop. You spot a funnel drop on a product listing page, trigger an exit survey on that page for the affected segment, and have open text responses within a day. Clarity has no equivalent and Microsoft has shown no sign of building one. If your team's bottleneck is understanding motive rather than observing behavior, the free tool does not compete.

Hotjar's sampling behavior deserves scrutiny at commerce volume. When you exceed your session allowance, capture is distributed rather than simply stopped, which sounds generous and quietly undermines rare event debugging. A sampled dataset is excellent for aggregate patterns and close to worthless for the specific failure affecting a narrow segment. Know which problem you are buying the tool to solve.

The Page Weight Question Most Comparisons Skip

On a storefront, a behavioral script is not free even when the license is. Both tools ship a client side agent that observes DOM mutations, and both add measurable overhead on low end mobile hardware. On a Core Web Vitals sensitive site, interaction latency added by a replay agent can cost you more conversion than the insight returns.

The defensible pattern is the same for both products. Load the agent after hydration, exclude it from checkout unless you have a specific active investigation, sample deliberately on your highest traffic templates rather than capturing everything because capture is free, and measure the delta in field data rather than lab data. Teams building headless storefronts should think about this alongside the rest of their client side telemetry budget, which we cover in more depth in our comparison of headless analytics stacks for Hydrogen storefronts.

Consent is the paired issue. Both tools collect behavioral data that most European regulators treat as requiring consent, and both integrate with consent management platforms. The failure mode is not the integration, it is the gap between when your page loads and when consent resolves. Verify that the agent genuinely holds capture until consent fires, on a real page load, on a real device, before you assure anyone that it does.

The Decision Framework: How to Choose

Dimension	Microsoft Clarity	Hotjar
Cost at commerce volume	Free, no traffic ceiling	Free tier, then roughly $49 to $700 plus per month by session volume
Sampling	None, full capture	Distributed sampling above allowance
Playback retention	About 30 days, longer for saved sessions and heatmaps	Varies by plan, generally longer than Clarity
Qualitative research	None	Surveys, feedback, interview recruiting
Upgrade path	Flat product, no enterprise tier	Into Contentsquare experience analytics
Data usage posture	Microsoft may use data per their terms	Standard vendor processor terms
Best at	Rare event debugging, unsampled coverage	Understanding motive, closing the research loop

The framework reduces to three questions.

First, is your bottleneck observation or explanation? If your team already has hypotheses and needs evidence, take the unsampled free tool and spend the budget elsewhere. If your team keeps arguing about why users abandon, buy the research loop.

Second, how rare is the behavior you care about? Rare failures require full capture. A tool that samples cannot find a defect affecting one percent of one browser, and no amount of AI summarization compensates for data that was never recorded.

Third, what is your retention horizon? If your analysis is seasonal or your compliance function requires evidence retention, a 30 day playback window is a structural constraint rather than an inconvenience, and the workaround of favoriting sessions manually is a process that will fail at exactly the wrong moment.

There is a fourth answer that most mature commerce teams land on, which is both. Clarity runs everywhere at full capture as the standing instrumentation layer. Hotjar or another research tool runs on specific templates during specific studies. The combined script cost is the real objection, which is why the deployment discipline above matters more than the license comparison.

What This Means for Your Business

Behavioral analytics has a persistent failure mode that has nothing to do with which vendor you choose: the data gets collected, dashboards get built, and no decision changes. Hours of recordings accumulate and nobody watches them, because watching them is not anybody's job and the output is not a metric anybody is accountable for.

The tool that changes decisions is the one wired into an existing loop. That means a heatmap review attached to a scheduled merchandising meeting, replay evidence required in the ticket template for any conversion bug, and a standing rule that any experiment proposal cites either quantitative evidence or a specific recorded behavior. Without that, the free tool and the expensive tool produce identical business outcomes, which is none.

It also means treating behavioral analytics as one input among several rather than the source of truth. Replay tells you what happened on one session. Your experimentation platform tells you whether a change actually moved the number across all of them, and that distinction is where most conversion programs go wrong. We work through the platform side of that in our review of experimentation tooling for headless commerce.

How Contra Collective Bridges the Gap

We deploy behavioral instrumentation on headless and Shopify Plus storefronts where the script budget is already tight and the consent requirements are real, so the work is as much about what not to load as what to capture. We wire the output into decision loops that already exist rather than building another dashboard nobody opens, and we measure the performance cost of the agent in field data before it ships to your highest traffic templates. Ready to make the right call for your stack? Book a free technical audit with no sales pitch, just clarity.

Final Thoughts

If you are choosing today with no other context, run Clarity. It is free, it does not sample, and the cost of being wrong is a script tag you remove. Use the budget you did not spend on a research tool when you have an actual research question, which will be sooner than you think and more specific than you expect.

If you already run Hotjar and the Contentsquare repackaging just moved your renewal number, the question is not whether the new price is fair. It is whether you are paying for the replay you could get free, or for the qualitative loop that has no free equivalent. Those are very different invoices, and only one of them is worth defending.