onboarding-cro

Overview of onboarding-cro skill

What onboarding-cro is for

The onboarding-cro skill helps you improve post-signup onboarding, activation, and time-to-value. It is best used when users create accounts but fail to reach the first meaningful outcome that proves your product’s value.

Who should install onboarding-cro

This skill fits product marketers, growth teams, founders, PMs, and UX writers working on SaaS or product-led growth flows. It is especially useful if your main problem is not acquisition, but weak activation after signup.

The real job-to-be-done

Use onboarding-cro when you need more than generic “improve onboarding” advice. The skill pushes the conversation toward a clear activation event, friction diagnosis, faster first value, onboarding checklist design, empty-state guidance, and testable experiments.

What makes this skill different

Compared with a normal prompt, the onboarding-cro skill has a tighter operating frame:

• it starts with product and activation context
• it focuses on the aha moment, not broad UX opinions
• it turns onboarding into measurable experiments
• it includes specific patterns surfaced in the repo, such as checklist design, empty-state use, and time-to-value reduction

Best-fit and misfit cases

Best fit:

• trial users sign up but do not complete setup
• new users do not reach the first success milestone
• you need onboarding experiments tied to activation metrics

Misfit:

• you are optimizing pre-signup conversion only
• you mainly need lifecycle email strategy
• your issue is retention far beyond the first-run experience

For those cases, the repo itself points to related skills like signup-flow-cro or email-sequence.

How to Use onboarding-cro skill

Install context for onboarding-cro

Install from the parent repository, since onboarding-cro lives under skills/onboarding-cro in coreyhaines31/marketingskills:

npx skills add https://github.com/coreyhaines31/marketingskills --skill onboarding-cro

If your environment uses another skill loader, the important part is the repo URL and the exact skill slug: onboarding-cro.

Read these files first

For a fast decision, inspect:

• skills/onboarding-cro/SKILL.md
• skills/onboarding-cro/evals/evals.json
• skills/onboarding-cro/references/experiments.md

Why this order matters:

• SKILL.md gives the operating method
• evals/evals.json shows what good output should contain
• references/experiments.md gives a useful bank of test ideas once diagnosis is done

What inputs the skill needs

The onboarding-cro usage quality depends heavily on four inputs:

1. product type and audience
2. your activation definition or aha moment
3. current onboarding steps
4. known drop-off points or baseline metrics

If you omit these, the model will still answer, but the output usually becomes generic and less testable.

Check for product-marketing context first

The repo explicitly tells the agent to read .agents/product-marketing-context.md first, or .claude/product-marketing-context.md in older setups. If your workspace has one of those files, provide it or make sure the agent can access it before asking for recommendations.

This matters because onboarding recommendations often fail when they are detached from product positioning, audience, and value proposition.

Turn a rough request into a strong prompt

Weak prompt:
“Help me improve onboarding.”

Strong prompt:
“We run a B2B project management tool for agencies. Our activation event is ‘create first project and invite one teammate.’ Only 30% of trial users create a project in week one. Current flow after signup: email verify, workspace setup, template choice, project creation, invite step. Biggest drop-off is after workspace setup. Use onboarding-cro to diagnose friction, redesign the first-run path, suggest a 3-7 item checklist, improve empty states, and propose experiments with metrics.”

That prompt gives the skill enough structure to produce an action plan instead of broad suggestions.

What good onboarding-cro prompts usually include

Include as many of these as you can:

• activation event definition
• current onboarding flow steps
• mobile or desktop context
• self-serve or sales-assisted onboarding
• user segment differences
• analytics baseline
• screenshots or copied UI text
• constraints like engineering limits or legal requirements

The skill is much stronger when it can separate critical steps from nice-to-have setup.

Recommended workflow in practice

A practical onboarding-cro guide looks like this:

1. define the activation event
2. map the current path from signup to first value
3. identify avoidable friction and unnecessary steps
4. decide whether to use product-first, guided, or value-first onboarding
5. design a short checklist with clear completion actions
6. improve empty states and stalled-user moments
7. pull experiment ideas from references/experiments.md
8. attach success metrics before shipping changes

This sequence matches the repository’s core emphasis on time-to-value and one goal per session.

How to use the checklist pattern well

The evals strongly suggest a checklist pattern of roughly 3 to 7 items. That is a useful constraint, not just a formatting preference. If your checklist is longer, users experience setup as work; if it is shorter, it may not create momentum.

Good checklist items are:

• observable
• tied to product value
• sequenced toward the aha moment
• completable in one session

Bad checklist items are administrative tasks with no visible payoff.

Use empty states as onboarding surfaces

One of the most useful parts of onboarding-cro for Conversion is its treatment of empty states. If a user lands on a blank dashboard, empty project list, or empty workspace, that screen should actively push the next best action:

• explain the value of the screen
• show one primary CTA
• reduce decision load
• offer templates, examples, or dummy data if appropriate

This is often a faster win than redesigning the full signup flow.

Pull experiment ideas from the reference file

references/experiments.md is worth reading because it converts strategy into test candidates. Useful categories include:

• reducing friction
• changing step order
• testing value-first paths
• using pre-filled templates
• adjusting required vs optional steps
• recovery for stalled users
• performance, accessibility, and mobile onboarding improvements

Use it after diagnosis, not before. Otherwise you risk random experiment lists with no link to your actual bottleneck.

Metrics to ask for every time

Do not use onboarding-cro without a measurement layer. Ask for:

• activation rate
• time to activation
• step completion rate
• checklist completion rate
• drop-off by step
• cohort differences by segment or acquisition source

This skill becomes much more valuable when each recommendation maps to a measurable change.

onboarding-cro skill FAQ

Is onboarding-cro only for SaaS products

No. The onboarding-cro skill is most obviously useful for SaaS, but it can also help with any product that has a meaningful first-use path: marketplaces, fintech apps, collaboration tools, creator tools, or AI products. The key requirement is a definable activation event.

Is this better than a normal CRO prompt

Usually yes, if your problem is activation rather than general UX critique. The skill gives a more disciplined frame: define the aha moment, cut time-to-value, focus each session on one goal, improve checklist design, and generate experiments tied to metrics.

When should I not use onboarding-cro

Do not use onboarding-cro install as your first move if your problem is clearly before signup. If users never start registration, another conversion skill is a better fit. It is also not the right tool for long-term retention programs or pure email nurture strategy.

Is onboarding-cro beginner friendly

Yes, if you can describe your product, users, and current flow. You do not need deep CRO expertise, but beginners get better results when they bring screenshots, funnel metrics, and an explicit activation definition instead of asking abstractly.

Does the repository include implementation code

No major automation or scripts are surfaced here. This is a thinking and workflow skill, supported mainly by SKILL.md, evals, and an experiment reference. Install it for better analysis and recommendation quality, not for plug-and-play code.

How do I know the skill is working well

Compare the output against the eval-like expectations from evals/evals.json. A strong answer should:

• check for product marketing context
• define activation clearly
• diagnose time-to-value friction
• recommend an onboarding approach
• suggest a tight checklist
• use empty states intentionally
• propose experiments
• include measurement

How to Improve onboarding-cro skill

Give onboarding-cro a precise activation target

The single biggest upgrade is to state exactly what user action means “this user got value.” Examples:

• “imports first CSV and sees a live dashboard”
• “creates first project and invites one teammate”
• “uploads first design and receives one comment”

Without this, the skill cannot prioritize steps properly.

Provide the current flow as a numbered sequence

Do not just say “our onboarding is clunky.” List the steps in order:

1. signup
2. email verification
3. workspace naming
4. template selection
5. data import
6. dashboard view

This helps the skill spot where value is delayed and which steps can be postponed, skipped, or merged.

Share real friction, not assumptions

Strong inputs include:

• “60% drop after email verification”
• “mobile users abandon template selection”
• “trial users skip integrations and never return”

Weak inputs sound like opinions:

• “users seem confused”
• “the flow feels long”

The skill gets better when you provide observed behavior.

Ask for one redesign plus three experiments

A good way to improve onboarding-cro usage is to structure the output request:

• one recommended core onboarding path
• three experiments to test against it
• expected impact and tradeoffs
• metrics for each

That keeps the response actionable instead of sprawling.

Force prioritization by constraints

If you have real limits, say so:

• no major backend work this quarter
• mobile only
• cannot remove compliance steps
• design team unavailable
• can only ship copy and ordering changes

Constraint-aware prompts produce sharper recommendations and reduce unrealistic advice.

Common failure mode: solving setup, not value

Teams often ask the skill to optimize completion of internal setup steps instead of user value. A shorter setup flow is not enough if users still fail to experience the core benefit. Tell the skill to optimize for first value, not mere form completion.

Common failure mode: too many goals in one session

The source material emphasizes one goal per session. If your onboarding tries to teach every feature at once, ask the skill to separate:

• required actions for activation
• secondary configuration
• later education

This usually improves completion and clarity.

Improve outputs with artifacts

If available, attach:

• screenshots of first-run screens
• copy from modals and tooltips
• analytics screenshots
• event names from your product analytics
• user interview snippets from recent signups

These artifacts let onboarding-cro critique concrete moments instead of inventing them.

Iterate after the first answer

After the first pass, ask follow-up questions like:

• “Rewrite the checklist for enterprise admins”
• “Now optimize for mobile users”
• “Keep email verification but move it later”
• “Rank these experiments by ease and likely impact”
• “Turn this into an A/B test plan”

That second iteration is often where the onboarding-cro guide becomes genuinely implementation-ready.

Use the experiment reference selectively

Do not dump the full references/experiments.md list into your roadmap. Improve the skill’s usefulness by asking it to filter experiments by your exact bottleneck, such as:

• reducing friction before first project creation
• improving empty-state conversion
• recovering users who stalled after signup day one

Selective use creates a tighter test program and better odds of measurable lift.