Session replay and heatmaps

Session replay vs. heatmaps: when to use each

Compare session replay vs. heatmaps, learn what each reveals, and see how to use both with product analytics to investigate customer friction.

Table of Contents

                Session replay shows the sequence of actions in an individual customer session. Heatmaps combine interactions across sessions to show where people click and how far they scroll on a page. Use heatmaps to find a pattern, then watch relevant replays to investigate the experience behind it.

                For example, a heatmap might show repeated clicks on a checkout label that isn't a link. A replay can show what the customer did before and after those clicks. Product analytics helps establish how often the behavior accompanies an abandoned purchase.

                How session replay and heatmaps differ

                The main difference is the unit of analysis: session replay follows an individual session over time, while a heatmap summarizes interactions across sessions on a page. Replay preserves the order of actions. A heatmap makes repeated behavior easier to spot at a glance.

                DimensionSession replayHeatmaps
                Main questionWhat happened during this customer's visit?Where do people interact on this page?
                ViewPlayback of a captured sessionAn aggregated overlay or element ranking
                Useful detailNavigation, repeated actions, and the sequence around a problemClick concentrations, element interactions, and scroll reach
                Typical starting pointA reported issue or a session behind a metricA page with an engagement or discoverability question
                Main limitationAn individual session doesn't establish how common a problem isAggregation removes the sequence behind each interaction
                Evidence to addAnalytics to measure the wider patternReplays to investigate the pattern and analytics to measure outcomes

                Consider an onboarding page with an optional setup step. A heatmap can reveal that the skip control receives little engagement. A replay can show a customer scrolling past it, returning to the top, and leaving. Neither observation proves that removing the step will improve activation. Together, they give the team a specific hypothesis to investigate.

                The two views are complementary. The session replay overview explains individual-session investigation; the heatmaps guide covers aggregated visual analysis.

                When to use session replay

                Use session replay when the order of actions matters, especially when investigating a reported bug, an abandoned flow, or an unexpected analytics result. A replay helps you reconstruct the experience around an event instead of guessing from the event name alone.

                Suppose a customer reports that they couldn't submit a checkout form. The final click tells you where their attempt ended. A replay might show repeated submissions, an error appearing below the visible area, and a return to the address field. That sequence gives the team a more precise issue to reproduce.

                In Amplitude, recordings can be viewed alongside an event timeline. Where the relevant capture is configured, error analytics adds console logs and failed network requests, while frustration signals identify behaviors such as repeated clicks. These details help connect a technical issue with the experience the customer encountered.

                Choose recordings relevant to the question. Sessions from customers who abandoned address entry will be more useful for that investigation than a random collection of homepage visits. Include successful sessions as a comparison: a confusing-looking interaction may also appear in visits that end in a purchase.

                A replay records observable behavior. It doesn't tell you what someone was thinking. Use the recording to form an explanation, then look for supporting evidence before treating that explanation as established.

                When to use heatmaps

                Use heatmaps when you want to understand how people interact with a page across captured sessions. They help identify controls people overlook, areas that attract unexpected clicks, and content that sits beyond the scroll depth many visitors reach.

                The map type determines the question it can answer:

                • Click maps: Show where clicks cluster. Repeated clicks on a static illustration can suggest that people expect it to be interactive.
                • Selector maps: Rank engagement with clickable elements. These help compare which controls attract interaction on a page.
                • Scroll maps: Show how far visitors reach. These help assess whether important content appears before or after a common stopping point.

                For a pricing page, a scroll map might show that few captured sessions reach a comparison section near the bottom. That supports a question about placement. It doesn't establish that the visitors read everything above it or that moving the section will increase purchases.

                Separate audiences whose layouts and behavior differ. A mobile visitor may encounter a collapsed navigation menu while a desktop visitor sees every link. Comparing each layout independently gives the interaction pattern a clearer meaning.

                Amplitude's heatmap views support click, selector, and scroll analysis. The useful view depends on whether the decision concerns a location, a specific element, or visibility farther down the page.

                How to use replay and heatmaps together

                Start with a measurable product problem, use heatmaps to locate a page-level pattern, and review relevant recordings to develop an explanation. Then return to analytics to assess the scale and decide what change to test.

                For an onboarding investigation, the workflow can look like this:

                1. Identify where customers leave. A funnel shows a drop between choosing a workspace name and creating a first project.
                2. Examine that page. A heatmap shows engagement with a help icon near the workspace field and little interaction with the continue button.
                3. Review the sequence. Relevant replays show customers opening the help text, editing the name repeatedly, and leaving after a validation message.
                4. Measure the affected behavior. Compare completion among sessions with and without the validation event, keeping customer segment and time period consistent.
                5. Write a hypothesis. Clearer naming instructions may reduce invalid submissions and increase first-project creation.
                6. Evaluate the change. Measure the intended outcome and monitor related problems before deciding on a wider release.

                In Amplitude, heatmaps use session replay capture. Click and selector map interactions can lead to the underlying events, relevant recordings, or a customer cohort. Scroll maps provide access to recordings that reached the selected depth.

                This lets an investigation move between aggregate behavior and individual examples. Keep the page, audience, device layout, and observation period aligned so the evidence describes the same experience.

                Mistakes that lead to the wrong conclusion

                Visual evidence becomes misleading when you treat interaction as intent, an example as a population, or a before-and-after change as proof of causation. Both replay and heatmaps work better when you make the limits of the evidence explicit.

                Avoid these interpretation mistakes:

                • Treating clicks as success: A busy button can indicate interest, confusion, or failed attempts. Pair interaction with the outcome, such as a completed purchase.
                • Treating scrolling as reading: Reaching a section establishes exposure to that area of the page, not comprehension of its contents.
                • Generalizing from one replay: An unusual session can reveal a real bug without showing how many customers encounter it. Quantify the associated behavior before prioritizing it against other work.
                • Ignoring the capture population: A sampled heatmap describes captured sessions. Recording rules and sampling affect which experiences are available for analysis.
                • Mixing page states: An overlay can be hard to interpret when a URL contains several layouts or application states. Align the background and the audience with the experience under investigation.

                For example, a redesign might increase clicks on a plan selector while completed upgrades decline. Calling the change successful based on the heatmap alone would miss the business result.

                Amplitude's heatmap troubleshooting guidance documents capture and page-state limitations, including sampled data and single-page application views that share a URL. Those details affect interpretation even when the map itself looks complete.

                Connect visual evidence with product outcomes in Amplitude

                Start with a drop-off in Amplitude, use a heatmap to locate a page-level pattern, and open relevant replays to examine what happened. Keep the investigation tied to the customer outcome you want to improve.

                For a checkout problem, product analytics establishes the conversion trend. Session replay adds the sequence around an unsuccessful attempt. Heatmaps help show where captured interactions concentrate on the page.

                Test the proposed fix in Statsig, Amplitude's experimentation platform. Use a feature flag to deliver the variants and an A/B test to measure the effect. A clearer error message, for example, can be evaluated on completed checkouts rather than clicks alone.

                Choose the evidence that answers the decision. Visual investigation helps you develop a plausible explanation; measurement establishes whether the proposed change improves the outcome.

                Related content:

                Amplitude is the leading all-in-one analytics platform for building better products. Product and engineering teams use Amplitude to connect product analytics and web analytics with session replay, heatmaps, experimentation, feature flags, and guides and surveys in one unified workflow, with AI agents working across all of it. Learn how Amplitude is building self-improving products with Wave, helping teams turn product data into PRs.

                Frequently asked questions

                Session replay and heatmaps show behavior at different levels. Replay follows an individual session in sequence. Heatmaps aggregate interactions across sessions on a page. A heatmap can identify an area attracting unexpected clicks, while relevant replays show the actions surrounding those clicks.

                Use both when an investigation needs a page-level pattern and the individual experiences behind it. A reported error may initially call for replay alone. A question about content visibility may start with a scroll map. Product analytics adds the frequency and outcome measurements needed to prioritize either finding.

                Heatmaps can identify patterns associated with abandonment, such as little engagement with a control or limited scroll reach. They don't directly reveal intent or establish the cause of a departure. Relevant replays, outcome metrics, and customer feedback help test the explanation suggested by the map.

                Support depends on the product. Amplitude heatmaps support web experiences, including mobile web, rather than native mobile apps. Amplitude session replay supports web and supported mobile platforms. A mobile website and a native app have different capture requirements, so their availability should be considered separately.

                Capture coverage and privacy controls answer different questions. Sampling determines which sessions are collected; masking and exclusion determine which content is captured. Amplitude provides controls for page text, inputs, and excluded elements. Validate the captured experience against your organization's requirements before collecting production sessions.