The same setup step keeps failing.

Connected

Error events, retries, time-on-step, support searches.

Example eligibility rule

Trigger failure-aware help after the same error code occurs three times within 15 minutes or time on the step exceeds twice its normal completion time. Suppress personalized help during a known outage and route the event to incident messaging instead.

The same error, the same field, the same OAuth screen, three times. This is not a motivation problem. It is a design or implementation problem showing up as a person.

Avoid this assumption

If the same step fails for a cohort, that is a product bug with a name, not a targeting opportunity.

Personalization here looks like help: a different explanation, a bypass, a human. Measuring it as "engagement" is how you miss churn.

Reach

How many eligible people can you reach in the chosen channel and time period?

Cost

What will content, delivery, QA and ongoing maintenance cost?

Return

What incremental outcome would cover that cost, and what is it worth?

Planning inputs, not measured results. Estimate reach × incremental outcome rate × value per outcome, then subtract cost. Use a range and a consistent time period.

Example format: Experience

An alternate path (skip, later, different auth, sample data).

Compare lighter and deeper responses

Choose the lightest response that solves the need.

Assist: Offer useful help

Inline error copy that names the likely cause and a doc jump.

Adapt: Change a meaningful part

An alternate path (skip, later, different auth, sample data).

Transform: Tailor the journey

A live assist offer after N failures, plus a ticket with the step prefilled.

What to watch out for

• A chatbot that restates the error.
• Forcing the same path because "the tour must complete".
• Not logging the failure into the learning memory.

Control

Default error state.

Treatment

Failure-aware help after repeated attempts.

Success measure

Step completion after failure, and activation.

Guardrail

Support volume. Help should drop tickets, not just relocate them.

Win or lose, record what worked, for whom and why. What you learn becomes the next signal.

• Pendo in-app guides Product experience platform combining usage analytics with in-app guides targeted by user or account segments.
• Appcues In-product engagement platform for building targeted flows, tooltips, checklists, and announcements without engineering.
• Customer.io Event-triggered lifecycle messaging across email, push, SMS, and in-app, built on a customer profile and behavior-data layer.
• Twilio Segment Unify Builds unified, real-time customer profiles from event and identity data (Unify), which other tools consume for activation.

They started setup, then disappeared mid-step. Remind them of the unfinished step. Then stop.

They paid. Then they never started setup. Give them one first step. Not a tour of everything.