Call the winner once it has won.

You send the variant on your events and ClickClacks does the maths. Every result carries an always-valid interval, so you can check it every morning without inflating the error rate.

Start tracking free All features
percent confidence, held across every comparison
90/95/99
arms: one control and up to seven treatments
8
split check across every arm that hides a broken result
p<0.001
guardrail events watched beside the success metric
20

What it does

ClickClacks doesn’t assign variants or manage feature flags. It reads the variant from an event you already send, counts who converted inside the window, and says when the difference is significant.

Plan

Know how long it will take before you start.

Choose the smallest change worth finding and a confidence of 90, 95 or 99%. ClickClacks works out the people needed per arm from your own baseline, and a likely finish date from your own traffic. Add treatments and the estimate grows to match, because each comparison has to clear a stricter bar.

  • A trade-off table shows what a smaller change costs in days.
  • Setup checks it has seen every arm’s value in the last 24 hours.
  • A guided five-step setup finds tests already running with PostHog, GrowthBook or Segment.

Check any day

The intervals hold however often you look.

Results use always-valid (sequential) confidence intervals, built to hold at every look. With several treatments, the error rate is split evenly across the comparisons (a Bonferroni split), so 95% means 95% for the whole test, not for each arm. Until an interval clears zero, the verdict is “Not enough data yet”.

  • Checking daily does not inflate the error rate.
  • Each treatment is compared with the control on one shared axis.

Guardrails

See what the winner did everywhere else.

Watch up to 20 other events next to the success metric. Guardrails don’t pick the winner. They show, with their own intervals, whether a treatment is doing damage somewhere else.

  • Every metric gets each arm’s rate, the lift and its interval.
  • A treatment that harms a guardrail is flagged on its own row.

Split check

A broken split blocks the decision.

Each test compares the real split with the allocation you set, an even split or a share per arm, using a chi-squared test across every arm. At p < 0.001 the result is hidden, the likely causes are listed, and recording a decision is refused until the assignment is fixed.

  • People per arm sit next to each arm’s expected share.
  • The banner notes the day the mismatch was first flagged.

Questions it answers

Ask it the way you’d ask a colleague.

How ClickClacks answers it

The verdict turns significant only once the always-valid interval for the lift clears zero. Until then it says “Not enough data yet”, whatever the observed number looks like.

+2.5 pts

Lift in sign-ups · +22.3% relative

Who it’s for

Three ways teams use it.

  • Agencies

    The client’s developers already split the new landing page with their own tooling. Send the variant on the pageview and bring the client an interval and a finish date.

    Variant on an existing pageview

  • Indie developers

    Assign the variant however you like, even a random pick in your own code, and send it on the event you already fire. ClickClacks shows when the result is significant.

    variant: b on signup_started

  • Product teams

    Try three onboarding flows at once against today’s. Guardrails keep refunds and support tickets on the same page, and the split check stops a broken rollout from being shipped.

    A control and up to seven treatments

How it works

Under the hood

  1. 01Send the variantAdd it as a property on an event you already send, e.g. experiment: pricing_layout_v3, variant: b. A person’s first exposure decides their arm.
  2. 02Describe the testPick the exposure event, the property holding the variant, the control and up to seven treatments, the expected share per arm, and the success event with a window of 1 to 90 days.
  3. 03Check it whenever you likeEach treatment’s interval narrows as people arrive. When one clears zero, its verdict turns significant.
  4. 04Stop or shipStop at any time. Record any arm as the decision, control included, once every arm has a usable result. It’s refused while the split looks broken.

Spec sheet

from the product

Statistics
Always-valid intervals
Confidence
90, 95 or 99%
Arms
1 control + up to 7 treatments
Several treatments
Bonferroni-split error rate
Expected split
Even, or a share per arm
Split check
Chi-squared, all arms, p < 0.001
Guardrails
Up to 20 events
Success window
1–90 days

Same events, different question

Works well with

All features →

FAQ

Experiments, in detail.

  • No. Your app decides who sees what, however it does today, and sends the variant as a property on an event. ClickClacks reads it and does the maths.

  • No. The guided setup looks through your events for tests that are already running, with presets for PostHog, GrowthBook and Segment as well as a format of your own, and offers them as a starting point. You map each variant value to an arm, renaming, merging or ignoring values as needed, and the mapping stays editable with a log of what changed.

  • Up to eight arms: one control and up to seven treatments. Each treatment gets its own always-valid interval against the control, and the error rate is split evenly across those comparisons, so the confidence you pick holds for the test as a whole.

  • Their first qualifying exposure decides their arm, permanently. Conversions only count if they happen at or after that exposure and inside the window.

  • The intervals are always-valid (a sequential method), built to hold at every look. A fixed-horizon test loses its guarantees when you peek; this one doesn’t.

  • It compares the observed split across every arm with the allocation you set, using a chi-squared test. Below p < 0.001 the result is hidden, the likely causes are listed, and recording a decision is refused until you fix the assignment and restart.

  • Any of them, the control included. ClickClacks records the decision once every arm has a usable result and the split check passes. Rolling it out stays with your own code or flag tool.

Start hearing your clicks.

Install one script tag and see your first report in minutes.

Start tracking free Sign in

About 5 KB · pageviews, clicks and scroll depth, automatically