Amplitude
Product analytics with AI agents
The deepest product-analytics platform on the market, worth it when you have the data maturity and the budget to match.
Product and growth teams that need behavioral cohorting and funnel analysis deeper than pageview tracking can provide, and who have engineering resources for the setup.
Pricing scales steeply and opaquely. Free caps at 2M events, Growth and Enterprise are custom-quoted with buyers reporting $22K-$64K+, and contract exit friction is a real complaint.
Reach for it when your team has the data maturity and budget to exploit its depth. If you only need basic traffic analytics or can't dedicate an engineer to instrumentation, it's the wrong tool.
What you're actually getting
Amplitude is a behavioral analytics engine. You instrument your product with events, and it answers questions GA4 cannot touch: which features drive retention, where users drop out of funnels, how cohorts behave over time.
The platform now wraps this in an AI layer. The Global Agent lets you ask product questions in plain language and get charts, cohorts, and follow-up investigations back. Specialized Agents monitor dashboards, watch session replays, and synthesize customer feedback on their own.
It has also grown beyond analytics into experimentation and activation. Feature flags, A/B tests, session replay, guides and surveys, and a CDP all live under one roof. The breadth is real, but so is the complexity.
I find the analytics and data governance in Amplitude Analytics significantly more powerful than what we had before. The features of definable events, groups, and the power of reports give us the ability to answer every complex business question we throw at the data. We've been able to spin up dashboards for our OKRs that give leaders fast insights into the weekly and monthly changes to our user base.
Where it earns its keep
Amplitude earns its place when behavioral questions are the core of your product work. Funnel analysis pinpoints exactly where users drop off. Retention cohorts show which onboarding steps predict long-term engagement. Segmentation slices users by behavior, beyond simple attributes.
It also shines when one platform replaces a patchwork. Teams that used to juggle a separate analytics tool, feature-flag service, A/B testing vendor, and session-replay tool consolidate into Amplitude and stop stitching data across systems.
The AI agents get real daily use. The Global Agent answers ad-hoc questions in seconds that used to require a dashboard build. The Session Replay Agent watches thousands of sessions and surfaces friction patterns a human would miss.
What I find most helpful about Amplitude Analytics is its ability to provide clear insights into user behavior and product usage. It makes it easy to track user actions, analyze retention, and identify areas for improvement. It have steep learning curve for new users, especially when working with advanced reports and event tracking. Setting and maintaining a clean analytics also requires careful planning.
Where it'll bite you
The learning curve is steep. Budget two to four weeks for a team to get comfortable. Event taxonomy design and data governance require careful planning upfront. Messy instrumentation produces misleading charts, and cleaning it up after the fact is painful.
Cost is the second trap. Free covers 2M events, which runs out fast once instrumentation expands. Growth and Enterprise are custom-quoted, and buyers report $22K to $64K+ per year. Overage billing and auto-renewal terms have produced real complaints on Trustpilot.
Contract and account-management friction shows up across review sites. Trustpilot sits at 1.7/5 with 46 reviews, many flagging billing disputes and difficulty exiting contracts. The product earns praise, but the commercial relationship deserves scrutiny before you sign.
Beware of their malpractice, it is hard to get out of a contract, they will keep charging you even if they know full well the software is not working for the client and they are not using it.
What it costs, really
There is a genuinely free tier: $0 for 2M events/month, no credit card, unlimited seats, AI agents included. Plus starts at $0 for the first 2M events, then scales with usage up to 70M events.
The real bill for a scaling team lands in Growth or Enterprise territory, custom-quoted and event-based. Buyers report $22K to $64K+ per year. A Startup Scholarship gives qualifying early-stage companies (under $10M funding, under 20 employees) one free year of Growth. Price your own pilot before trusting any single number.
What I like best is how Amplitude brings product analytics and Session Replay together in one place. We track complex quote and bind funnels across web and mobile, and being able to jump from a drop-off in a chart straight into a replay of what the user experienced makes debugging and prioritization much faster. When event properties aren't structured consistently across client and server, cross-platform funnels can fail silently, you see drop-offs without an obvious reason until someone audits the payload shape.
I like that you can assemble the board yourself. I like that, in principle, all the needs I had related to analytics (views and clicks) were met by Amplitude Analytics. When I was creating one of the charts, I encountered a situation where, upon switching to another tab and returning to the previous one, my data was not saved. I mean the filter data, and it felt like they should have been saved.
Same job, other tools
Also filed under Analytics & attribution:
