Amplitude Review 2026: The Product Analytics Platform That Puts Behaviour First

The Honest Take

If you are building a digital product and you are not doing serious behavioural analytics, you are flying blind. You know acquisition numbers from your marketing stack, revenue from your finance tools, and support ticket volume from your helpdesk — but you do not know *what users are actually doing* between those events. That gap is where Amplitude lives, and it fills it better than almost anything else on the market.

Amplitude in 2026 is a mature, sophisticated product analytics platform with genuine AI capabilities, a strong experimentation layer, and one of the best session replay implementations available. It is not cheap, and it is not simple. But for product teams that take user behaviour seriously, it is one of the most valuable tools available.

What Amplitude Does

Amplitude is a product analytics platform designed to help product, engineering, and growth teams understand how users behave inside digital products — web, mobile, or desktop. The core question it answers: which behaviours drive retention, revenue, and growth, and which ones indicate churn risk?

In 2026, the platform spans:

• Product analytics — event-based tracking with funnels, cohort analysis, retention, pathfinding, and impact analysis
• Session replay — video-like recordings of individual user sessions correlated with analytics events
• Experiment — A/B testing and feature flagging with statistical rigour and automated analysis
• CDP (Customer Data Platform) — built-in data collection and identity resolution for connecting events across sessions and devices
• AI features — Amplitude AI for automated insight generation, natural language querying, and anomaly detection

The platform's philosophy is event-based: every interaction a user has with your product is tracked as an event (clicked button, completed checkout, viewed feature, etc.), and the analytics engine allows you to slice, dice, and correlate those events in virtually any way you need.

Product Analytics: The Core Capability

Amplitude's funnel analysis, retention analysis, and user path analysis are genuinely class-leading in depth and flexibility.

Funnels allow you to define any sequence of events and measure conversion between each step, with the ability to segment by user properties, cohorts, time windows, and more. The "Explore" feature lets you drill into the users who dropped out at each step — not just the aggregate percentage but the actual users — so you can investigate qualitative reasons for drop-off.

Retention analysis is where Amplitude particularly excels. The N-day and unbounded retention curves are standard; the genuinely powerful feature is the ability to identify which early behaviours predict long-term retention. Amplitude calls this "Compass" — it surfaces which actions in the first session, first week, or first month correlate with users who are still active 90 days later. This is the kind of insight that directly informs product decisions and onboarding design.

User paths and journeys — the Pathfinder chart shows the most common sequences of events before and after any given event, revealing navigation patterns you would never discover from looking at individual screens. Finding out that 40% of users who completed a key conversion first went through a specific help article is the kind of insight that changes product roadmaps.

Cohort analysis allows you to define user cohorts based on any combination of event behaviours and properties, then track how those cohorts behave over time. Behavioural cohorts (users who performed X before doing Y) are more analytically interesting than demographic cohorts and Amplitude handles them elegantly.

Session Replay: The Qualitative Layer

Amplitude's Session Replay sits alongside the quantitative analytics and is deeply integrated — not bolted on. You can move from an aggregate funnel drop-off to individual session replays of users who dropped off at that exact step. This integration between quantitative ("40% dropped off here") and qualitative ("let me watch 10 replays to understand why") is the product's killer workflow.

The privacy controls are solid — sensitive input fields are automatically masked, and you can configure additional masking rules for PII-heavy applications. GDPR compliance is handled through configurable data retention and the ability to delete individual user recordings.

Session replay in Amplitude is genuinely competitive with standalone tools like FullStory and LogRocket, particularly because of the event correlation. For teams that would otherwise pay separately for a replay tool, this integration has real consolidation value.

Experiment: A/B Testing With Statistical Rigour

Amplitude Experiment is a strong A/B testing and feature flagging platform. What differentiates it from simpler tools is the statistical rigour and the integration with Amplitude's analytics layer.

Statistical approach: Amplitude uses sequential testing, which allows you to evaluate experiment results as data comes in rather than waiting for a fixed sample size. This reduces the time-to-decision for experiments while controlling for false positive rates — a meaningful practical improvement over traditional fixed-horizon testing.

Feature flags allow gradual rollouts and targeted releases based on user properties or cohorts defined in Amplitude Analytics. The feedback loop between analytics (identify the cohort) and experimentation (target that cohort with a test) is tight.

The integration with the analytics platform means you can analyse experiment impacts not just on your primary metric but on the full range of downstream behaviours. Running a checkout flow experiment and wanting to know whether it affected 90-day retention? Amplitude can answer that.

Amplitude AI: Practical Intelligence

Amplitude's AI features in 2026 are genuinely useful rather than aspirational.

Ask Amplitude (natural language querying) allows product managers and non-technical stakeholders to ask questions in plain English and receive charts and analysis without needing to configure the query builder manually. The interpretation of intent is reasonable, though complex analytical questions still require a human to frame the question carefully.

Automated Insights proactively surfaces anomalies — sudden drops in conversion, unexpected spikes in error events, cohort behaviours that deviate from expected patterns — without requiring anyone to check dashboards manually. This is Amplitude's answer to the "analytics that nobody looks at" problem.

Predict — Amplitude's predictive analytics feature — uses ML models trained on your product's event history to score users on likelihood to convert, churn, or complete a specific action. These scores can feed into marketing automation tools or be used to trigger in-product interventions. The predictive quality is solid for high-volume products with rich event histories; it is less valuable for products with limited event data.

Pricing

Amplitude operates a usage-based model where pricing scales with monthly tracked users (MTUs):

The Starter plan is genuinely capable — it includes core analytics, session replay, and some AI features. It is one of the more generous free tiers in enterprise SaaS and means meaningful products can get started without procurement involvement.

Growth and Enterprise pricing is where the complexity starts. Session replay is charged separately above the Starter tier based on session volume. Experiment is an add-on. CDP features carry additional costs. A fully-loaded Amplitude deployment for a meaningful B2C product with millions of users can run to tens of thousands of pounds per year.

The practical note: Amplitude is materially more expensive than PostHog at equivalent feature coverage, and competitive with Mixpanel depending on usage patterns.

Comparison: Amplitude vs. Alternatives

Mixpanel is the closest competitor. The core analytics capabilities are comparable; Amplitude has an edge in experiment depth and session replay integration, while Mixpanel's self-serve experience and Flows feature are strong. The decision between them often comes down to team preference after evaluation rather than objective superiority.

Heap takes a different approach with autocapture — it records every user interaction automatically without requiring explicit event instrumentation. This means historical data exists for events you did not think to track in advance, which is a genuine advantage. The trade-off is data volume, noise, and the need for retroactive event definition. For organisations with limited engineering bandwidth for instrumentation, Heap is worth serious evaluation.

PostHog is the open-source challenger that has expanded rapidly. It combines product analytics, session replay, feature flags, A/B testing, and CDP in a single platform with an open-source core. For product teams comfortable with self-hosting (or PostHog Cloud's competitive pricing), the feature breadth is remarkable. The analytics depth for advanced retention and cohort analysis is behind Amplitude and Mixpanel, but the gap is closing.

Who It's For

Amplitude is the right choice if:

• You are building a digital product (app or web) and user behaviour data is central to your product strategy
• You have a product team that will genuinely use the analytics, not just generate dashboards for quarterly reviews
• You run regular product experiments and want statistical rigour integrated into your analytics platform
• Session replay is important and you want it correlated with quantitative events
• You have engineering resource to instrument events properly — Amplitude's value scales directly with instrumentation quality

Amplitude is not the right choice if:

• You need autocapture without engineering instrumentation (consider Heap)
• You are cost-constrained and want open-source flexibility (consider PostHog)
• Your product has very low user volumes and you cannot reach meaningful statistical significance in experiments
• You primarily need marketing analytics rather than product analytics
• You want a simple, low-complexity analytics setup for a straightforward web property

How to Get Started

1. Start with the free Starter plan — Amplitude's free tier is generous enough to instrument a real product and get meaningful data flowing. Do not over-engineer the instrumentation at the start.

2. Define your event taxonomy before you instrument — The biggest mistake teams make is instrumenting hundreds of events without a clear naming convention or event hierarchy. Spend a day defining your core events (what counts as a conversion? what is your activation event?) before writing any tracking code.

3. Install the SDK — Amplitude has SDKs for every major platform (JavaScript, iOS, Android, React Native, Python, etc.) and a broad set of integrations with tools like Segment, Rudderstack, and customer data platforms.

4. Build your North Star chart first — Before building funnels and cohorts, define and build your organisation's North Star metric chart. This creates alignment and is usually the first chart worth sharing with leadership.

5. Implement session replay for key flows — Start with your critical conversion funnels (signup, onboarding, checkout) rather than capturing all sessions. This controls costs and focuses attention.

6. Set up Automated Insights alerts — Configure anomaly detection for your key metrics so drops in activation or retention surface automatically rather than waiting to be discovered.

The Verdict

Amplitude is one of the best product analytics platforms available in 2026, and for product teams that are serious about understanding user behaviour and running a disciplined experimentation programme, it is worth the investment. The combination of deep funnel and retention analytics, integrated session replay, and a genuinely strong experimentation platform creates a workflow that supports the full product development cycle.

The costs are real and scale with usage. Teams need to evaluate whether their usage pattern — particularly session replay and MTU volumes — keeps them in a sensible pricing tier. And the platform's value is directly proportional to the quality of event instrumentation, which requires engineering investment upfront.

For teams willing to make that investment, Amplitude pays back substantially.

Digital by Default rating: 8.5/10

Building a digital product and wondering which analytics platform is right for your team? We help UK product teams implement and get maximum value from analytics and experimentation tools. [Talk to us at Digital by Default](/contact) — and bring your event taxonomy questions.