Best Heatmap Tools in 2026: Tested and Compared

Last verified: July 2026. Every price and free-tier figure in this guide is sourced to the vendor's own pricing page and stamped with the date we checked it — because heatmap-tool pricing changes often, and most comparison articles quietly go stale.

A heatmap tool shows you what visitors actually do on a page — where they click, how far they scroll, what they ignore — instead of what you assume they do. In 2026 the category has split into three camps: free entry-level tools that get you a color overlay fast, qualitative suites that pair heatmaps with session replay and surveys, and developer-and-AI-native tools that let you pull the underlying behavior data out through an API, a CLI, or an AI agent. Which camp you belong in matters more than any single feature checkbox.

We compared 11 tools on the criteria that actually decide the purchase: real pricing (with the date we verified it), free-tier limits, session replay, A/B testing, form analytics, integrations, GDPR posture, AI features, mobile support, and two axes almost no other guide measures — how openly each tool lets you export your own behavior data, and whether an AI coding agent can reach that data directly. HeatMapX is our own product; we've kept it in the comparison but scored it on the same axes as everyone else, including where it deliberately does less than the competition.

How we evaluated (and why these criteria)

Rather than rank on vibes, we scored every tool against a fixed set of criteria. Here's what each one means and why it's on the list.

• Pricing & free tier (verified): the cheapest paid monthly price and the free plan, each read off the vendor's official pricing page on the date shown. If a price is annual-billing-only, we say so.
• Session replay: can you watch individual recorded sessions, not just aggregate heatmaps? A different job from heatmapping, but frequently bundled.
• Form analytics: field-by-field drop-off on your forms — where CRO money often hides.
• A/B testing: can the tool itself run experiments (client-side and/or server-side), or does it only observe?
• Integrations: GA4, CRMs, e-commerce platforms, data warehouses.
• GDPR / privacy: EU hosting options, consent handling, data-retention controls.
• AI features: automated insights, anomaly flags, natural-language querying.
• Ease of setup: one snippet vs a tag-manager dance vs an SDK build.
• Mobile app support: native iOS/Android heatmaps, not just responsive web.
• Enterprise vs SMB fit: where the tool's pricing and depth actually land.
• [Our unique axis] API openness score (0–5): can you get your raw behavior data back out — for a warehouse, a migration, or your own analysis — or is the export shallow or enterprise-gated? This score comes from our own cross-tool research on 19 heatmap and analytics products. No other guide we found measures this.
• [Our unique axis] CLI & MCP (AI-agent access): can a developer — or an AI coding agent via the Model Context Protocol — pull heatmap data directly from the command line or an agent, instead of clicking through a dashboard? This is where tooling is heading, and almost no comparison touches it.

What makes this guide different

Most heatmap round-ups recycle the same feature tables. Four things here you won't find together anywhere else:

1. An API-openness score for every tool, so you can see at a glance whose data you can actually take with you.
2. Dated pricing. Every figure carries the date we verified it against the vendor's page — so you know how fresh it is.
3. A developer / AI-agent lens. We note which tools expose a CLI or an MCP server, because if your workflow runs through an AI agent, that's the difference between "usable" and "not."
4. Current ownership. Hotjar is now part of Contentsquare, and its pricing is presented accordingly — a fact most 2026 guides still get wrong.

At a glance: 11 heatmap tools compared

Prices verified against each vendor's official page on July 7, 2026. In the table, "Quote" means quote-only enterprise pricing, "Usage-based" means metered self-serve pricing with no flat starting price, and "API openness" is our 0–5 score for how freely you can export your own raw data. The rows below focus on where the tools actually differ — every tool reviewed here does click and scroll heatmaps, so those aren't repeated in the table.

The 11 tools, reviewed

HeatMapX — best for developers and AI-driven CRO at a flat price

Our own tool, scored on the same axes as everyone else. HeatMapX is heatmap-first with built-in A/B testing (client and server) and AI analysis, at fully transparent flat pricing: Free, $12 Lite, $29 Plus, $99 Pro (verified Jul 2026). It scores 5/5 on API openness and ships both a CLI and an MCP server, so a developer — or an AI coding agent — can pull heatmap data directly instead of clicking through a dashboard. Honest limits: HeatMapX deliberately does not offer session replay or surveys. If watching individual recordings is core to your workflow, choose a replay-first tool below. Best for engineering-led teams that want heatmaps plus experimentation plus open, exportable data without enterprise pricing.

Microsoft Clarity — best free option, no catch

Genuinely free with no traffic cap (verified Jul 2026). You get click and scroll heatmaps, session replay, funnels, AI insights, and an MCP server. What's missing: native A/B testing (only an experiment-tagging API for tests run elsewhere), surveys, and a CLI. API openness is a respectable 3/5. For a zero-budget team that wants solid observation without a bill, nothing else on this list competes.

Hotjar (now part of Contentsquare) — best-known, replay + surveys for marketers

The most recognized name in the category — and, as of the Contentsquare acquisition, no longer a standalone brand, a detail most 2026 guides still get wrong. There's a free plan; Growth starts at $39/mo billed annually ($49 month-to-month, verified Jul 2026). Strengths: qualitative replay, surveys, funnels, AI, and an MCP server. Weaknesses that matter: no native A/B testing (Hotjar monitors experiments run in third-party tools), and an API openness of just 1/5 — your data is hard to take with you. Best for marketing teams who want replay and surveys and won't need to export raw data.

Crazy Egg — the click-map veteran with simple split testing

Around since 2005 and still focused on click and scroll maps. No permanent free plan (30-day trial); Starter is $29/mo billed annually (verified Jul 2026). You get heatmaps, session replay, client-side A/B testing, surveys, and funnels. No server-side testing, no CLI or MCP, and API openness of 1/5. Best for small teams that want classic click maps plus straightforward on-page split tests, and don't care about data portability.

Mouseflow — best for form and friction analysis

Free plan (500 sessions/mo); Essential from $25/mo (verified Jul 2026). Mouseflow's calling card is automated friction detection and deep form analytics. It covers heatmaps, replay, surveys, funnels, AI, and an MCP server, at a solid 3/5 API openness. It has no native A/B testing and no CLI. Best for conversion teams whose biggest questions are "where do people stall?" and "which form field loses them?"

FullStory — best for enterprise, retroactive replay

An enterprise experience-analytics platform with a surprisingly generous free tier (30,000 sessions/mo). Paid pricing is quote-only — no public figure on the date we checked. Strengths: high-fidelity, no-sampling replay you can analyze retroactively, plus surveys, funnels, AI, and MCP. It has no native A/B testing and no CLI, and its site confirms click maps but didn't explicitly name scroll maps (so we left that cell unverified). API openness 3/5. Best for larger orgs that live in session replay.

Lucky Orange — affordable all-in-one for small e-commerce

No permanent free plan (7-day trial); Build from $32/mo billed annually (verified Jul 2026). A cheap bundle of heatmaps, replay, surveys, and funnels, with limited AI. No native A/B testing, no MCP, and API openness of 1/5. Best for small e-commerce stores that want a lot of observation features for a low monthly price and aren't planning to export data.

Smartlook — being sunset, plan your migration

We include Smartlook because most other 2026 guides still list it as a live option — but you should know it is winding down. Following Cisco's acquisition, Smartlook is End-of-Sale (no new subscriptions since May 31, 2026) and will be decommissioned on September 30, 2027, with capabilities migrating to Splunk's Digital Experience Analytics (verified Jul 2026). It still offers heatmaps, replay, and funnels, but we can't recommend adopting a product on a published shutdown timeline. If you're already on it, start planning your move.

Contentsquare — enterprise experience analytics (and Hotjar's parent)

The enterprise platform now sitting above Hotjar — and more self-serve than most enterprise tools. Its Experience Analytics line has a genuine Free plan ($0, 200k sessions/mo) and a Growth tier from $39/mo billed yearly ($49 month-to-month, verified Jul 2026), with Pro and Enterprise tiers quote-based above that. It offers heatmaps, replay, surveys, AI (Sense), and MCP on every plan; funnels and CLI are partial. No native A/B testing (it integrates with A/B tools instead). API openness 3/5. Best for teams that want zone-based, revenue-linked behavior analytics and can grow from a free plan into enterprise.

PostHog — the strongest open-source, developer-native option

The most direct competitor to HeatMapX on the developer axis, and we'll say so plainly. PostHog is open-source with a generous usage-based free tier (1M events, 5k recordings/mo, and more) and metered pricing above it — there's no flat starting price. It bundles heatmaps, replay, server-side experiments (client-side via feature flags), surveys, funnels, and AI, and — like HeatMapX — ships both a CLI and an MCP server and scores 5/5 on API openness. The trade-offs versus a flat-priced heatmap tool: PostHog is product-analytics-first (heatmaps are one feature among many) and usage-based billing can be harder to predict. Best for engineering teams that want everything in one open platform and are comfortable metering.

VWO — best for experimentation-first teams

An experimentation-led suite where heatmaps ride alongside a full testing engine. Sales-gated: tier names are public but no dollar figures rendered on the date we checked. It's the most feature-complete row in the table — heatmaps, replay, client- and server-side testing, surveys, funnels, and AI (via its Copilot) — with an MCP server and 3/5 API openness. Best for teams whose center of gravity is A/B testing and who want observation bundled in rather than bolted on.

Which one should you pick?

• Zero budget: Microsoft Clarity — free, unlimited, genuinely capable.
• Marketing / qualitative research: Hotjar (Contentsquare) for replay + surveys, or Mouseflow if forms and friction are your focus.
• Enterprise, replay-centric: FullStory, or Contentsquare for zone-based revenue analytics.
• Developers & AI-agent workflows: HeatMapX (flat pricing, heatmaps + testing + open data) or PostHog (open-source, usage-based). Both score 5/5 on API openness and both ship a CLI + MCP — pick HeatMapX for predictable flat pricing and built-in heatmap-native testing, PostHog for an all-in-one open platform.
• Experimentation-first: VWO.
• Avoid for new projects: Smartlook, which is on a published shutdown path through 2027.

What heatmaps can't tell you (and how to cover the gap)

Heatmaps are diagnostic, not prescriptive. They show you where attention goes, not why, and never on their own prove that a change will make you money. Three limits worth naming before you buy:

• Correlation, not causation. A cold zone on a heatmap is a hypothesis ("nobody sees this CTA"), not a verdict. Confirm it with an A/B test before you rebuild the page — which is why several tools here bundle experimentation, and why a heatmap-only tool often needs a testing tool beside it.
• Aggregation hides individuals. A click map averages thousands of sessions; the one rage-clicking user who churned disappears into the color. Session replay and friction scoring are how you get that person back.
• Sample size matters. A heatmap built on a handful of sessions is noise. As a rule of thumb, wait for a few hundred to a few thousand sessions on a page before trusting the pattern — the exact threshold depends on your traffic and how decisive the pattern is.

The practical takeaway: a heatmap is the start of an investigation, not the end. The tools that pair observation with testing — or that let you pull the data into your own analysis — shorten the distance from "interesting" to "shipped."

Frequently asked questions

What is a heatmap tool, and how does it work?

It's a script you add to your site that records anonymized visitor interactions — clicks, mouse movement, scroll depth — and renders them as a color overlay on your pages: warm colors where attention concentrates, cool colors where it doesn't. Some tools also record full sessions you can replay.

What's the difference between click, scroll, and move maps?

A click map shows where people tap or click (and where they click things that aren't links — a sign of confusion). A scroll map shows how far down the page people get, so you can see what's below an invisible "fold." A move map tracks cursor movement as a rough proxy for where eyes go; treat it as directional, not precise.

Is a free heatmap tool good enough?

For many small sites, yes. Microsoft Clarity is genuinely free with no traffic cap, and several paid tools include a free tier. The free options tend to give you less data retention, shallower exports, and no A/B testing — so you outgrow them when you need to act on the data at scale, not just look at it.

How many sessions do I need before the data is reliable?

Enough that the pattern is stable rather than a few loud sessions. A common working range is a few hundred to a few thousand sessions per page before you trust a heatmap; low-traffic pages simply take longer to gather a trustworthy picture.

Will a heatmap script slow my site down?

Modern tags load asynchronously and typically add a small amount of overhead, but script weight varies between tools, and heavy replay capture costs more than lightweight heatmapping. If performance is critical, measure the before/after with Lighthouse and prefer a tool that publishes a small script size.

Are heatmap tools GDPR-compliant?

The reputable ones can be, but compliance is a shared responsibility: you still need a consent mechanism and a lawful basis. Look for EU data-hosting options, automatic masking of sensitive inputs, and configurable retention. "Free" does not mean "compliant by default."

Can a heatmap alone improve conversions? How is it different from A/B testing?

A heatmap tells you where to look; an A/B test tells you what actually works. Heatmaps generate hypotheses ("this CTA is being missed"); experiments verify them. The strongest setups use both — which is why some tools in this guide combine observation and testing in one product.

Which tools support mobile-app (not just mobile-web) heatmaps?

Native iOS/Android heatmapping is a narrower field than web. If you need in-app maps, confirm native SDK support specifically — some tools that market "mobile" only mean responsive mobile web.

Can I export my data or reach it from code or an AI agent?

This varies enormously, and it's the axis most guides ignore. Some tools give you a full API and even a CLI or an MCP server an AI agent can call; others expose only shallow aggregates or gate raw export behind enterprise plans. If you plan to warehouse your data, migrate later, or run analysis through an AI agent, check the API-openness column before you commit.

Can I run more than one heatmap tool at the same time?

Technically yes, and many teams trial two in parallel. The costs are additive script weight and split data. It's fine for a short evaluation; for the long run, pick the one whose data you can most easily get back out.

Bottom line

If you want free and fast, start with Microsoft Clarity. If you want observation plus experimentation in one place, look at the qualitative suites. And if your workflow runs through code or an AI agent — where the winning question is "can I get my data out and act on it programmatically?" — weigh the API-openness and CLI/MCP columns heavily, because that's exactly where the category is heading. Whichever you choose, pick the tool whose data you can take with you.

Related comparisons

• 7 Hotjar Alternatives Worth Switching To
• 6 Crazy Egg Alternatives (Free and Paid)
• Best A/B Testing Tools for Developers