Replays in context
Every session is linked to the flags and experiment variants the user had, with no manual tagging. Filter the list to one variant and compare it with another.
Session Recordings
Metrics tell you what. Recordings show you why.
Session Recordings replays real user sessions with the flags and variants each user saw. The Confidence Agent watches them and reports the bugs, friction, and opportunities behind a metric change.
See the sessions behind every result
Unlimited recordings on paid plans and 100,000 a month on the free plan. You only pay for the recordings the Confidence Agent analyzes.
An agent that watches for you
The Confidence Agent watches sessions and reports the bugs, friction, and opportunities they have in common, each linked to the recordings behind it.
A guardrail moves, and the dashboard can't tell you why
Your experiment wins on its primary metric, but a guardrail metric gets worse. A guardrail metric is a metric you watch to make sure a change doesn't cause harm somewhere else, such as cart abandonment or error rate. Now you have to explain the regression before you can ship.
The numbers tell you which step got worse, but not what users ran into there. A standalone replay tool doesn’t know which variant a user saw unless you send it yourself, so comparing treatment with control takes days of extra work.
How do Session Recordings work with feature flags and experiments?
Session Recordings is built into Confidence, so every recording shows the flags and variants the user saw. You decide what to record using the experiments and rollouts you already have. Recording rules choose who gets recorded; they never change which variant a user receives.
A/B test rule
Records sessions from an existing experiment, using its targeting and allocation. You set what share of those sessions to capture.
Rollout rule
Does the same for a feature rollout.
Conditional targeting rule
Records any audience you define with context attributes and segments, such as users on a premium plan.
What does the Confidence Agent find in recordings?
The Confidence Agent watches hundreds of sessions and tells you what they have in common: bugs, points where users get stuck, and opportunities to improve. Inside an experiment, it compares variants and tells you what users do in treatment that they don’t do in control.
Ask follow-up questions, such as “Do users who leave try to find the shipping cost first?”, or search transcripts by meaning to find sessions where something happened, even if no event or error was logged.
Narrow hundreds of sessions to the few that explain your result
Filter recordings by the flag variant a user saw, or by signals Confidence detects in every session: rage clicks, dead clicks, error messages, failed network requests, and console errors. Filter on how often a signal happens, such as three or more rage clicks.
Combine signals with page URL, duration, device, browser, OS, country, and custom tags. For example: every treatment session on the payment page with at least one failed request.
Recordings feed Confidence Loop
Session recordings are one of the signals Confidence Loop uses out of the box. Loop turns what users run into during sessions into observations, groups them into issues, and investigates them. Recordings explain one experiment. Loop watches your whole product and fixes what it can.
Private by default
The SDK masks every input field by default, so typed values like names, emails, and card numbers aren't captured. Mask or block more elements on top of that, and use recording rules to record only the sessions you need, at the sample rate you choose.
Recordings are kept for 30 days, then deleted. AI findings and transcripts stay, so you can still search and read them after the replay is gone.
How do I set up Session Recordings?
01
Install the package
Install the @spotify-confidence/session-recording npm package.
02
Initialize the recorder
Pass your client secret and the same context you pass to your feature flags.
03
Create a recording rule
For example, an A/B test rule on the experiment you want to understand.
04
Open the Recordings tab
New sessions appear in Confidence a few minutes after they end.
Questions and answers
How much do Session Recordings cost?
Recording is unlimited and free on paid plans, and the free plan includes 100,000 recordings a month. You only pay for recordings that the Confidence Agent analyzes. Replaying, filtering, and searching sessions has no extra cost.
Which platforms do Session Recordings support?
Session Recordings supports web apps today, through the @spotify-confidence/session-recording npm package.
Which flag evaluations show up in a recording?
Recordings include the flag variants evaluated in the same browser by Confidence's JavaScript SDK or its OpenFeature web provider. Flags evaluated on the server don't appear automatically, because they're evaluated outside the recorded browser.
Do recording rules change which variant users get?
No. Recording rules only choose which sessions to record. Flag targeting and experiment allocation stay exactly as they are.
Can I record only users in a specific experiment?
Yes. Create an A/B test rule on the experiment and set the percentage of sessions to record. Confidence uses the experiment's own targeting and allocation.
What does the Confidence Agent do with recordings?
It watches and transcribes sessions, then reports bugs, friction, and improvement opportunities, each linked to the recordings behind it. In an experiment, it compares variants. You can also ask it follow-up questions about what users did.
How long are recordings kept?
Recordings are kept for 30 days, then deleted. AI findings and transcripts are kept, so you can still search past sessions and read what the Confidence Agent found.
Is user input recorded?
Input values aren't. The SDK masks all inputs by default, and you can mask or block more elements on top of that.
Read more about Session Recordings
Introducing Recordings
Confidence Session Recordings lets you replay actual user sessions, with AI-powered analysis to surface bugs and insights right from your experiments.
Zan Markan
Confidence Loop: Building products that continuously improve
Confidence Loop turns signals from real users into product understanding and improvements, with teams choosing how much of the loop runs autonomously.
Johan Rydberg
Get Started