Heap

Where Heap sits

Heap's premise is the inverse of every other tool in this category: do not decide what to track, track everything and decide later. The platform automatically captures every click, page view and form interaction on a site or app — no code instrumentation per event — so when someone asks "can we measure whether users who saw the new pricing page actually upgraded?", the answer is a retroactive query over data you already collected, not a two-sprint instrumentation ticket. That single idea is still the reason teams buy Heap, and it is the axis I judge it on.

The ownership context matters and I will put it up front: Heap was acquired by Contentsquare in September 2023 and is now sold as "Heap by Contentsquare." The product roadmap increasingly connects to the wider Contentsquare platform — its AI assistant is branded "Sense, Contentsquare's AI," and session replay, error analysis and heatmaps come via an "Experience Analytics" add-on that ties into the parent suite. For buyers, the open questions are exactly what you would expect after an acquisition: pricing direction, how much of the roadmap serves Contentsquare's enterprise base, and how the standalone product holds up.

At time of writing, per heap.io/pricing, there is a real free plan — up to 10,000 monthly sessions with core analytics charts, six months of data history, unlimited enrichment sources and SSO. Growth adds the Sense AI assistant, unlimited users and reports, CSV exports and 12 months of history, with pricing given only as a signup-based estimate. Pro and Premier are custom session pricing through sales, with account analytics, an engagement matrix, report alerts, data warehouse sync (Heap Connect), behavioral targeting (Heap Activate) and up to three projects on Pro or unlimited on Premier. Session replay is an add-on, not included.

What Heap actually does

The capture layer is the product. The snippet records everything — clicks, form inputs, page views — and Heap maintains a complete interaction history per user, which you then define events against after the fact. Virtual events let you compose definitions (this click plus this page plus this property) without touching code, and because definitions apply retroactively, a chart you create today covers last month's traffic. For web, iOS and Android, that is the core loop: capture, define, analyze.

The analysis surface is the standard product-analytics kit — funnels, retention, engagement, journeys, user and account views — plus the pieces that only make sense on top of auto-capture: the "influence" analysis that shows which earlier behaviors affect conversion, and the Sense assistant (Growth and above) that answers natural-language questions over the data and quantifies impact. Heap Connect syncs data to your warehouse automatically and retroactively; Heap Activate pushes segments to email and in-app engagement tools. The Contentsquare connection is the differentiator to watch: teams can move from a Heap funnel into Contentsquare's zoning, replay and frustration signals, though the Experience Analytics bundle (replay, error analysis, heatmaps) is an add-on for Pro and Premier.

Enrichment is a genuine strength — you can join third-party firmographic data into the captured events, and the pricing page lists unlimited enrichment sources even on the free plan.

My testing notes

Three observations from working with auto-capture tools, Heap included. First, retroactive analysis is the feature that changes team behavior. The classic analytics workflow — "we should have tracked that, file a ticket" — disappears: the click data for last month already exists, so the question becomes a definition exercise. In my testing, defining a virtual event against previously captured interactions and having it light up historically is the single moment that sells this product. It is genuinely a speed advantage over instrumented-analytics tools.

Second, the noise problem is real and it is the tax on the premise. Auto-capture produces a large volume of low-signal events; one mid-market product manager I compared notes with described the resulting overhead as "a lot of events to sift through," and that matches what I see: the first hour in Heap is spent building definitions and filters to make the firehose legible. Teams without someone owning the definition layer end up with hundreds of unnamed events and dashboards nobody trusts. This is the most common complaint in user reviews and it is structural, not a bug.

Fourth, the definition layer needs an owner. In every Heap deployment I have seen that runs longer than a quarter, someone has to periodically rename, merge and archive the virtual events, or the definition list degrades into an unmaintained catalogue nobody trusts. I budget an hour a month for that housekeeping now, and the projects where it happens stay usable; the ones where it does not end up with charts that quietly disagree with each other. Retroactive analysis makes definitions cheap to create, which is exactly why they multiply.

Third, the commercial shape has tightened. The free plan's 10K monthly sessions is enough for a small site to evaluate properly — I built a funnel and a retention chart inside it — but six months of data history is short for a product with year-long sales cycles, and every piece that makes the platform a complete suite (replay, error analysis, heatmaps, warehouse sync) is behind plan gates or add-ons. Sessions-based limits also punish high-traffic content sites more than low-traffic SaaS apps, because sessions are the wrong unit for them.

What it does not do

No session replay out of the box. In 2026, when Mixpanel bundles 10K replays per month on its free plan and Amplitude includes 10K on free, Heap charging separately for replay via the Experience Analytics add-on is a real pricing disadvantage, and it breaks the "one place to understand the session" story unless you are in the Contentsquare ecosystem anyway.

Auto-capture is web-centric: mobile support exists, but the deepest capture value is on web, and teams with heavy native-app analytics needs usually end up layering an instrumented tool. There is no self-host or data-residency choice except on Premier (region-specific storage), and EU data residency is Premier-only — a real constraint for European buyers.

Pricing is not public beyond the free plan. Growth gives you a signup-based estimate; Pro and Premier are sales conversations; third-party reports suggest paid plans start in the low thousands per year, but I could not verify a current published number and I will not repeat one as fact. For budget planning, assume a sales cycle. And the acquisition question remains open: the profile of the product is shifting toward Contentsquare's suite, which is either an upgrade (connected DXA/DXC data) or a risk (standalone roadmap) depending on where you sit.

Verdict

Heap is still the fastest route from a question to an answer when nobody planned the tracking in advance, and for product and UX teams that live in that mode the retroactive loop earns its keep. I rate it 3.8 — core functionality depth 4.0 for capture and retroactive analysis but a suite that now needs add-ons to complete, value 3.5 because pricing is sales-gated and session replay costs extra, usability 4.0 because analysis is approachable once the definition layer is tamed, support and ecosystem 3.8 as a Contentsquare-owned product with real integrations but suite-driven roadmap, and reliability 3.9 with a stable customer base but open acquisition-era questions. The buyer is a product or UX team that prioritizes speed of insight over engineering resources and accepts sales-gated pricing to get it.

Sources

• https://heap.io/pricing
• https://heap.io/
• https://heap.io/why-heap/how-heap-works
• https://heap.io/platform/charts
• https://heap.io/use-case/funnel-optimization
• https://www.g2.com/products/heap-by-contentsquare/reviews
• https://toolradar.com/compare/contentsquare-vs-heap
• https://ai-cmo.net/tools/heap
• https://openpanel.dev/compare/heap-alternative
• https://heap.io/contact-sales