Journey Orchestration
Triggers, waits, branches, goals and channel actions on a visual canvas, with per-node stats, versioning and a 30-day replay.
DecideExplore Journeys
Decide
Prove incrementality, not activity.
A/B and multivariate tests, bandits, control groups, a universal holdout, attribution windows and feature flags.
Conversion · order_completed within 72h
n = 12,400 / armControl
no message
3.10%A
push · 10% off
3.90%B
WhatsApp reminder
4.60%C
push + in-app
4.20%
Universal holdout · 3%
receives nothing, ever
3.01%
n = 4,120 · chosen by deterministic bucket · never messaged
Bandit allocation · traffic shifts toward the winner
cp 1 of 4A 33%
B 33%
C 34%
cp 1cp 2cp 3cp 4winner rolls out to the remainder
Overview
A renewal is won on one number: revenue you caused, measured against people you deliberately left alone. Edverix runs message and journey experiments with automatic winners, keeps a permanent account-wide holdout, and attributes conversions and revenue back to the message that earned them.
Only on Edverix
- Cost per conversion by channelOnly on Edverix
Because you bring your own providers, every send has a cost, and the only number that matters is what a conversion cost.
- Flag state on every eventOnly on Edverix
Segment any funnel, retention grid or revenue report by which variant a person was in.
Capabilities
Everything in Experimentation
11 capabilities in 3 groups, organised the way the product is.
Tests
A/B and multivariate
Up to five variants with automatic split sizing and a winner rolled out to the remainder.
Bandit optimisation
Traffic shifts toward the winner continuously rather than at a fixed checkpoint.
Journey-path experiments
Compare two entirely different sequences and route toward whichever converts.
Control and attribution
Campaign control groups
Hold back a random slice per campaign and compare automatically.
Universal holdout
A small account-wide group that receives nothing, ever, so cumulative impact is measurable.
Conversion and revenue attribution
Designate the conversion, set the window, and see revenue traced to the message that earned it.
Uplift reporting
Conversion, revenue and order lift against control, calculated for you.
Cost per conversion by channelOnly on Edverix
Because you bring your own providers, every send has a cost, and the only number that matters is what a conversion cost.
Flags and config
Feature flags and remote config
Change app behaviour and content without an app-store release, with per-segment overrides and instant rollback.
Deterministic user buckets
A stable per-user number from 0 to 99 that experiments, holdouts and flags all share.
Flag state on every eventOnly on Edverix
Segment any funnel, retention grid or revenue report by which variant a person was in.
FAQ
Questions about Experimentation
How is the universal holdout chosen?
By deterministic bucket, so the same people stay in the holdout across every campaign and journey until you rotate it.
Which attribution model do you use?
Deepest-latest by default: a click beats an open beats a send, ties broken by recency, inside the conversion window you set. First-touch and linear are available per report.
Get started
See Experimentation on your own data.
Send your first event in an hour, simulate your first journey against real history, and keep every byte of it in a database you can query.
- No credit card for the Developer plan
- Bring your own providers
- Export everything, any time