Mandatory vs. Optional Onboarding Split
Split onboarding into a spare mandatory setup and a random-access optional layer — model the buyer's mind, not your features.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 85%
An onboarding design principle: onboarding is a scalable substitute for sales, so it must approximate a friendly, approachable human and guide the buyer's mind through value understanding, simplicity check, tire-kicking, and commitment. Practically, split it into two buckets — a spare mandatory setup (ideally under three screens, only the info and essential setup every user needs to succeed) and an optional, random-access layer of beneficial-but-not-essential things people can return to anytime.
Origin
From Oji Udezue's forthcoming PLG book, informed by his time at Calendly and Typeform; he points to the offline retail industry (shelf-stockers, inventory merchandisers) as the true experts in the buyer-psychology this draws on.
Core principles
- 01Onboarding is a substitute for sales and account management at scale
- 02Its domain is social/biopsychology — the mindset of a buyer/evaluator — not product
- 03The mind moves through: value-prop understanding, simplicity check, kicking tires, making the choice
- 04Mandatory setup must be spare — ideally under three screens
- 05Optional-but-beneficial steps should be random-access, returnable to anytime
How to run it
- 1
Frame onboarding as scaled sales
Design onboarding to approximate how friendly and approachable a good salesperson is, so it can process thousands of customers quickly.
Pro tip Draw inspiration from offline retail merchandisers — they are extraordinarily good at guiding an evaluator's mind.
- 2
Map the four stages of the buyer's mind
Support each stage the average mind moves through: understanding the value proposition, judging whether it's simple enough and valuable enough, kicking the tires, and committing.
- 3
Build a spare mandatory setup
Include only what every user needs to succeed — say what the product does and collect the essential setup. Keep it to no more than three screens.
Pro tip At Calendly the mandatory setup was connecting your calendar and setting defaults like 9-to-5 availability — the two things that set up all future success.
Watch out If you make the truly-essential setup optional, more people will fail; conversely, don't pad the mandatory flow with non-essentials.
- 4
Move the rest into an optional random-access layer
Put beneficial-but-not-universally-necessary steps into an optional layer users can return to anytime they want help.
Pro tip Skip elaborate click-through wizards — in Udezue's experience they have limited benefit; just direct people to the one thing they need to do next.
- 5
Show one or two examples of good, and clarify trial state
After mandatory onboarding, show one or two (not ten) examples of what 'good' looks like — it's mimetic — and make crystal clear where the user is in the trial cycle and when payment is expected.
Pro tip Clarity about trial timing (pay now vs. after seven days) materially improves conversion because users otherwise feel 'at sea.'
In the wild
Calendly's mandatory onboarding asked only for connecting the calendar and setting defaults such as 9-to-5 availability. Those two essentials set up all future success, so they were required rather than optional.
→ A minimal mandatory flow that prevented downstream failure while keeping setup short.
Udezue found multi-step click-through wizards had limited benefit. What worked was directing people to the single thing to do post-setup and showing one or two examples of good output, which is mimetic — users see exactly what they're shooting for.
→ Higher trial-to-conversion by reducing onboarding to the essential action plus a concrete model to imitate.
Common mistakes
Making essential setup optional
If the setup that determines future success (like connecting a calendar) is skippable, far more users will fail downstream. The genuinely-essential steps must be mandatory and spare.
Relying on elaborate wizards and long flows
Long click-through wizards have limited benefit and add friction. Fewer mandatory steps plus a couple of concrete 'good' examples convert better than a guided tour.
Is it for you?
Best for
PLG product teams designing self-serve onboarding that must replace human sales at scale.
Not ideal for
High-touch enterprise sales motions where a human handles activation, or products whose complexity genuinely requires deep guided setup.
From the transcript
“onboarding is a substitute for sales”
“one is mandatory the setup it has to be spare I try not to make it more than three screens”
“akali for example we asked them about connecting their calendar and setting some defaults is it nine to five or whatever that is those two…”
“having examples not ten one or two of what good looks like is particularly powerful because it's mimetic”
From the episode
Picking sharp problems, increasing virality, and unique product frameworks
Oji Udezue (Typeform, Twitter, Calendly, Atlassian)