Skip to content

Adaptive Journeys

Build the path. Prove it changed behavior.

Build event-triggered customer journeys across email, inbox, web push, and mobile push. Wait, branch, send, and run an Experiment from the behavior already reaching Apex. A workspace-wide Global Holdout shows the lift each Journey added.

No credit card to start.

The Journey builder

Customer behavior decides what happens next.

Start an Adaptive Journey from a customer event, a schedule, or Segment entry. Build the path with Wait, Branch, Send, Experiment, Webhook, and Exit steps.

An Engagement Branch can route on a click, bounce, push open, or no action after the window closes. The same customer events start Journeys, resume waits, change branches, and credit Outcomes.

Activation onboarding

Skill Shot sample workspace

Choose a customer moment

Canvas steps

Trigger
Wait
Branch
Send
Experiment
Webhook
Exit

Communications built in

The Journey chooses when. The Composer owns the message.

Every Send or Experiment step points to a communication you build in the same Communication Composer.

Open the letter from the canvas, edit every word and variable, preview email, inbox, web push, or mobile push, then return to the same step. There is no second, smaller editor to learn.

Cart recovery · linked to the 72-hour Send step

Edit the communication here. The journey keeps the timing and path.

add_to_cart is live

Variables

Found in Skill Shot and ready to merge.

Three ways to insert

  • Type {{ in any field for autocomplete.
  • Drag a variable onto a field.
  • Click one to drop it where you’re editing on the canvas.
specwirednot wiredsysbuilt-in

Added to cart(app)

Apex found these details in the moment that starts this message.

Customer

Details Apex knows about the person receiving this message.

Apple Mail
Skill Shotsupport@skillshotgolf.com9:41 AM

To: me

Header
Skill Shot
Heading
Text
Image
John Daly holding the Skill Shot app
Text
Image
Skill Shot tournament with a $2,500 prize pot
Button

Experiment inside the path

Compare the message against the action you want.

Use an Experiment step to compare Variants of one communication against the Conversion that matters.

Opens can tell you that someone looked. The Experiment asks whether they signed up, purchased, upgraded, or completed the product action you chose. Apex names the Winner when every evidence gate clears. You decide whether to apply it to the Communication.

Trial conversion · Experiment step

Skill Shot sample workspace

Previewing Deadline-led

3 days left on your trial, Alex

Conversion: subscription_started

Compare the message

Traffic
50 / 50
Conversion window
14 days
Winner
You approve it

Journey templates

Start with the whole Journey, not an empty canvas.

Start with a complete journey for welcome, trial conversion, cart recovery, or win-back.

Choose a template and Apex creates a draft with the starting event, every step, and the communications it needs. Review everything before you turn it on.

Experiment included

Welcome (new signup)

Email

Greet every new signup and compare two subject lines in a ready-made Experiment.

Trigger
user_signed_up
Segment
Every new signup

Prior impact

+18% welcome click-through

Experiment included

Trial conversion (T-7d)

EmailIn-app

Send at seven, three, and one day before trial expiry, with the Conversion already chosen.

Trigger
trial.expiring (T-7d)
Segment
Trial users, 7 days before expiry

Prior impact

+12% paid conversion

Experiment included

Abandoned Cart Recovery

PushEmail

Send push first, email later, and exit the moment the cart empties or purchase happens.

Trigger
add_to_cart
Segment
Anyone who adds a product to cart

Prior impact

+15% cart-recovery conversion

Bring what already works

Import the graph. Review the details.

Import exports from AWS Pinpoint, Customer.io, Iterable, or SendGrid into draft Journeys and Segments.

Nothing imported sends until you review the canvas, attach the right communications, and publish. Apex shows what imported successfully and what you still need to review. It does not guess at missing letter content.

Migration preview

Review the translation before creating any drafts

Import from

What Apex brings over

  • Journey steps and connections
  • Segments and trigger contracts
  • Warnings for anything that needs a person
skill-shot-pinpoint.jsonpinpoint
Segments
1
Journeys
1
Trigger contracts
2
Translation warnings2 total
1 manual review1 info
  • [step: send-day-3] Attach or create this letter in the Communication Composer before publishing.
  • [global] Everything was imported as a draft. No customer can enter yet.
Translated journeys (1)
  • Trial conversion
    pinpoint-trial-conversion · 7 steps · trigger=trigger-trial-expiring
Translated segments (1)
  • Trial customers, 7 days from expiry
    pinpoint-trial-customers · predicate kind=event_fired
Journeys land as drafts. Nothing fires until you publish.

Publish without surprises

Try it in Dev. Open it to Beta. Turn on Production.

Turn a Journey on for Dev, Beta, Production, or more than one at once. Every publish creates a new version while current customers finish the version they started.

Before a marketing message sends, Apex checks suppression, opt-out, frequency caps, and the Global Holdout. Apex also checks the trigger, every variable, every linked communication, and overlapping Journeys before it publishes.

Activation onboarding · Draft ahead of Production v3

Skill Shot sample workspace

Production v3 keeps running

Publishing creates v4. People already inside finish on v3.

ProductionBetaDev

Turn on for

Environments

Before Apex publishes

  • Every trigger and letter is ready
  • Every variable has a value
  • Overlapping triggers get a warning

Global Holdout + Calibrated Impact

Know what happened because the Journey ran.

A Global Holdout keeps a fixed group from every marketing Journey, and Calibrated Impact compares its Conversion rate with the customers who received the Journey.

One send report tells you what recipients did. The control group estimates what they would have done anyway. The difference is the lift the Journey added.

Trial conversion · Production

Skill Shot sample workspace

What counts

Subscription startedsubscription_startedLive
Conversion window
14 days
Global Holdout
5%

Measure the Conversion, not the open

The held-out customers receive no marketing sends from any Journey. Their behavior is the control.

Calibrated impact

Production onlyv4
Exposure
Exposed
1810
95.0% of triggers
Holdout
95
5.0% — control group
Daily exposure (last 3 days)2026-09-14 → 2026-09-16
ExposedHoldoutDaily lift
Conversion + lift
Exposed rate
12.98%
235 converted
Holdout rate
10.53%
10 converted
Lift
23.3%
vs. holdout

People who received the Journey converted at 12.98%. The randomly assigned control group converted at 10.53%. Apex sent that group no Journey messages. Confidence: high based on 1,810 exposed.

This is an early read. Confidence improves as more customers reach the Conversion.
How these numbers are computed

Single-touch, journey-scoped. Each conversion event is credited to journeys that exposed (or held out) this subject within the configured attribution window.

Window. Events withingoalEventWindowDays(fallback attributionWindowDays, default 7) of the journey’s start credit. Journeys that opt intopostExitAttributionWindowDaysalso credit events within that many days of successful exit.

Multi-journey overlap. When the same conversion qualifies for multiple journeys and at least one sets attributionPriority, we credit a single winner per a deterministic stack (highest priority → most recent exposure → highest version). Otherwise every overlapping journey credits independently.

Censoring. Failed, stopped, and timed-out executions are dropped from both the numerator and denominator — the cohort split you see above is the censored census.

Your move

Build a Journey that proves it worked.

Start with a template or bring the graph you already run. Attach the communications, choose the Conversion, and keep the Global Holdout honest. You do not need a sales call to begin.