Find where users drop off
before your aha moment.
Describe your onboarding in steps. Get a drop-off risk score for each stage, the friction causing abandonment, and a prioritised fix list — specific to your product type and user goal.
No screenshots needed. Tell us your product type, target user, desired aha moment, and the steps in your onboarding — in plain English.
Claude evaluates every step against known drop-off patterns for your product type — scoring risk, estimating abandonment rate, and identifying friction points.
A risk score per stage, the specific friction causing drop-off, and a concrete fix for each — plus your overall health score and estimated time to aha moment.
Most churn is decided in the first session.
If a new user doesn't reach value before they leave, they almost never come back. Onboarding is not a feature — it's the entire product for the first 30 minutes.
Wrong sequence kills retention
Asking for a credit card before delivering value, or showing advanced features before core ones, causes 40-70% of trial users to never return. Sequence is everything.
Most teams never measure their own onboarding
Teams that build onboarding rarely go through it as a new user. They optimise for familiarity, not first-time comprehension — which are completely different experiences.
Time to aha moment predicts churn
The longer it takes a new user to reach the moment they feel the product works, the more likely they are to churn. Every unnecessary step between signup and value is a churn driver.
Friction compounds across stages
A 20% drop at stage 1 and a 30% drop at stage 3 leaves you with under 56% of your original cohort before anyone has seen the core feature. Small improvements at each stage compound.
Useful at any stage — not just when retention is broken.
The grader tells you where users leave and why. That's useful before you build, after you rebuild, and at any point where retention metrics are not what you need them to be.
SaaS product managers
When D7 retention is low despite reasonable acquisition
Most D7 churn happens because users never reached the aha moment. The grader identifies exactly which onboarding stage is losing users and why — without having to wait for session recordings.
Founders at pre-product-market-fit stage
Before investing in growth or paid acquisition
Acquiring users into a broken onboarding is expensive. The grader surfaces structural sequence problems early, before you spend on driving traffic into a flow that loses 60% of users in the first 10 minutes.
Growth teams
After an onboarding redesign, to stress-test the new flow
After rebuilding onboarding, teams often improve one problem and introduce another. The grader gives a fast structured read of the new flow before releasing it to a full cohort.
Product designers
During discovery or before presenting an onboarding proposal
Running the grader on the current flow before a redesign proposal gives you a concrete problem list to frame the business case. After the proposal, it gives you a benchmark to improve against.
The friction we find in most onboarding flows.
These four patterns account for the majority of avoidable user drop-off at the onboarding stage. They are not unique to any product type or company size — they appear consistently because the people who design onboarding already know the product.
The grader flags which of these apply to your flow and at which stage — with a specific fix rather than a general recommendation.
Asking for too much before delivering any value
The most common critical finding: signup requiring name, company, role, team size, and use case before the user has seen a single screen of the product. Each field at this stage has a dropout cost. The fix is to push data collection to after the first value moment.
Empty state with no path forward
New users land in a blank dashboard and have to figure out the first action themselves. "Create your first [thing]" is not helpful when the user doesn't yet know what [thing] does for them. Products that demonstrate value before requiring work convert significantly better.
Aha moment buried behind setup steps that should be optional
Inviting team members, connecting integrations, setting up a profile, and choosing a plan are all things that can wait. Making them mandatory before the user can experience the core product is one of the highest-friction onboarding mistakes — and one of the easiest to fix.
No progress indication during multi-step setup
Users who don't know how much setup remains are more likely to abandon. A simple step counter ("Step 2 of 4") reduces abandonment meaningfully because it sets an expectation. The absence of progress signals is interpreted as an unknown commitment.
Common questions
How does the onboarding flow grader work?
You describe your product type, target user, and onboarding steps in plain English. Claude (Anthropic's AI) analyses your flow against known drop-off patterns for that product and user type — scoring each stage by risk, estimating abandonment rate, and identifying the specific friction causing users to leave.
Do I need screenshots or a live product link?
No. The grader works from your plain-text description. You describe what happens at each step — what the user sees, what they are asked to do — and the AI analyses the structure and likely friction from that description.
How accurate are the drop-off estimates?
The estimates are based on industry benchmarks for the product type you select. They are directional — intended to rank stages by risk and quantify the likely impact — not a replacement for your actual analytics. The most valuable output is the friction point identification and fix prioritisation.
What is an "aha moment" and why does it matter?
The aha moment is the first time a new user genuinely experiences the core value of your product — not just learns about it, but feels it working. Research consistently shows that the faster users reach this moment, the higher their long-term retention. The grader uses your aha moment description to evaluate whether your onboarding sequence efficiently delivers users there.
How many steps should I include?
Include between 2 and 7 steps — each one representing a distinct stage where the user has to do something or make a decision. You do not need to list every micro-interaction; focus on the major gates (signup, setup, first action, first value, etc.).
My product has a complex setup — is that a structural problem or just reality?
Often both. Some products require configuration before they can deliver value — that is real. But most teams front-load that setup rather than deferring it, because building the setup flow feels productive. The grader distinguishes between genuinely necessary early steps and steps that can be deferred until after the user has experienced the core value. Deferred setup almost always improves retention.
Can I use this to evaluate a mobile app onboarding flow?
Yes. Describe your mobile app onboarding steps the same way — what the user sees and does at each stage. Mobile onboarding typically has different drop-off patterns than web (smaller screens, interruptions, permission requests) and the grader accounts for those when you specify a mobile app product type.
How is this different from running user interviews?
User interviews are qualitative and slow — a single round takes 2–4 weeks and gives you the perspective of 5–10 users. The grader is structural: it evaluates your onboarding design against validated patterns and benchmarks in under 30 seconds. It is the starting point, not the replacement. Use it to identify where to focus research, not to replace it.
We redesigned onboarding 6 months ago — should we run this again?
Yes, especially if your retention metrics have changed. Onboarding that was well-designed for your product at one stage often becomes a bottleneck as the product adds features, the target user evolves, or the competitive context shifts. Running the grader on your current flow is a 2-minute investment that surfaces whether the redesign held or introduced new problems.
AI spots the drop-off.
We redesign the flow.
Our product designers validate every friction point against your actual user data, then deliver a redesigned onboarding flow — with clear effort vs. impact priorities so you know where to start.
No commitment · We'll scope it together