Amplitude for Designers: What It Is, What It Measures, and When to Use It
Amplitude is a product analytics platform. What designers can do with funnels and retention, which plans it has today, and how to get started.
What is Amplitude?
It is a product analytics tool. Your development team installs an SDK, defines which actions to record (for example “payment button pressed” or “account created”), and Amplitude stores each event with its user, date, and properties. From its interface you can then build charts without writing code.
The difference from a session recording tool is the approach. Amplitude aggregates: it shows you percentages and trends across thousands of users. A heatmap or a recording shows you one specific person.
What plans does it have today?
According to Amplitude’s official pricing page (November 2026), there are four plans: Free, Plus, Growth, and Enterprise.
- Free: free with no credit card, with up to 2 million events per month. The page’s comparison table also lists limits such as 10 saved charts per organization, one year of data access, 10,000 session replays per month, and no support or SLA.
- Plus: starts at $0 with the first 2 million events per month free and goes up to 70 million events and 700,000 monthly tracked users (MTU). It adds custom events and formulas, behavioral cohorts, alerts, heatmaps, and two years of data on top of Free.
- Growth and Enterprise: custom pricing based on event volume, with a contact-sales option. Growth adds advanced behavioral exploration and SSO; Enterprise adds data access controls and advanced user management.
Plans and limits change often, so before deciding, read the current official page rather than a third-party comparison site. To estimate your volume, ask development how many events the product records per month. That number tells you which tier you are in, far better than any industry average.
What can a designer do with it?
Two charts cover most design questions:
- Funnel. Measures how users move through a defined sequence of events and where they drop off. A user counts as converted when they trigger the events in the order you set. You can choose “this order” (other events may happen between steps), “any order”, or “exact order” (no events in between). It is the ideal view for a checkout or an onboarding. To understand this kind of analysis better, read funnel analysis.
- Retention. Measures how often users come back after an initial event. It compares the date of that starting event with the date of the return event, and shows each return event as a separate segment in the chart. It tells you whether a design improvement makes people return or only makes the first visit prettier. More detail in cohort analysis.
You can also compare segments (for example, mobile users versus desktop users) and save charts to track them over time.
Why should a designer care?
Because it lets you check whether your design changed anything. Without data, a redesign conversation ends in taste. With a funnel from before and after, the conversation shifts to which step lost people and whether that step improved.
For example, if you redesign sign-up, Amplitude can show whether the share of people reaching “account created” went up, and how many of them come back the following week. Those are two different questions, and each has its own chart.
Example
Illustrative example: a booking app has a four-step funnel: “search”, “time slot selection”, “customer details”, and “booking confirmed”. In the funnel, 42% of people who reach customer details don’t confirm. The team suspected price, but the chart shows the drop-off concentrates on the details step, mostly on mobile. The hypothesis changes: the form asks for too many fields on a small screen. Now the designer has a concrete finding to propose a test of progressive disclosure in the form and measure the result with the same funnel.
When NOT to use it
- To find out why something happens. It tells you that people leave at a step, not the reason. For that you need user interviews or usability tests.
- If the events aren’t well defined. A funnel built on badly named events gives numbers nobody understands. Before opening a chart, someone on the team has to define the event plan.
- If your volume is very low. With few users, a two-point change means nothing. In that case, watch sessions or talk to people.
- To make privacy decisions without review. Events can contain personal data. Check with the right people what gets recorded.
Common mistakes
- Looking at the overall average without segmenting.
- Comparing periods with different seasonality without saying so.
- Building funnels on events that don’t reflect what the user did.
- Trusting a figure from a price comparison site without checking it.
- Treating a chart as the answer to a question it did not ask.
How to start tomorrow
Ask for read access to your product’s Amplitude project. Find the event that marks the start of a flow you designed and the one that marks its end. Build a two- or three-step funnel and look at drop-off by device. Write the number down before you change anything. After your next change, repeat the same chart with the same comparison period. If you can’t find the events, that is the first conversation you need to have with development.