I rebuilt creator onboarding from a wall of dropdowns into a progressive, personalized flow.

A decision about where complexity should live, not a simplification: expanded where creators struggled, compressed where they lost time to value.

Figma

Claude

Sigma & Heap

Before
After
Before
After

10%

Increase in seller activation*

20%

Increase in product creation

90%

Increase in Brand and domain set-up

*Activation at Teachable is defined by a scoring model: a school is considered activated once it reaches a 60-point threshold across a defined sequence of early behaviors, including domain and branding customization.

CONTEXT

Teachable is a B2B SaaS platform where creators sell online courses and onboarding is the first real impression to increase seller activation

DURATION

3 Months

MY ROLE

User Researcher

Product Designer

Data analyst

TEAM

2 x Product Designers

1 x Product Manager

4 x Engineers

PROBLEM AREAS

Value came after the work, not before

I mapped the existing flow against user interaction data and support tickets, and found the gap:

What the flow showed.

A long intake form followed by a five-step course creation sequence, separate screens for course details, thumbnail, pricing, purchase description, and outline, before a creator saw anything specific to their business.

What the data confirmed

Creators who customized their brand activated faster, but that setup was hidden in the dashboard, generating 85+ support tickets. Whereas, course creation forced users to take pricing decisions before they were ready.

Original Teachable onboarding form: a single screen with nine stacked dropdown questions and lengthy creator set-up flow before a creator sees any value.

COMPETITOR ANALYSIS

Three of four competitors I looked at led with structured, personalized onboarding, showing product value early.

I audited four direct competitors to see how they handled the same trade-off I was navigating: how much to ask a new user upfront versus how much to defer until they'd already seen value.

Miro board screenshot of four competitor onboarding flows, showing structured, personalized onboarding.

DESIGN GOALS

Getting creators to their own brand faster, and protecting the fastest path to their first product.

Reorder, not just redesign

Move brand and domain setup to onboarding flow, so the first thing a creator does is specific to their business.

Make ownership visible

I wanted creators seeing their own brand, name, and colors as a preview in first three clicks

Get to a first product fast

Strip course creation down to what's required to focus on low effort product creation.

HYPOTHESIS 1 - DOMAIN AND BRAND SETUP

If brand and domain setup moves earlier into onboarding, then activation and retention improve,
because seeing their own brand sooner helps creators recognize Teachable as a fit for their business.

DESIGN DECISION PHASE 1 - completion

I replaced a single screen of ten stacked dropdowns with a guided, 4-step progressive flow, including brand and domain set-up.

Before and after comparison: the original single dense form versus the new four-step guided flow, including the new dedicated brand and domain step

Step progress indicator

Shows creators exactly how far they are and how much is left, replacing the overwhelm of ten fields with a visible, finite path.

Grouped, contextual questions

Fewer questions per screen, and grouped by theme, so each step feels like answering one thing at a time, not filling out a form.

Brand & domain step

Step 3 didn't exist in the old form at all, it's a moment added specifically to build ownership early.

Phase 2 - ENGAGEMENT & ACTIVATION

Making it feel like their school by redesigning branding and domain setup step introduced in phase 1 with a live preview.

Before and after comparison of the brand and domain setup step: a static form with no feedback versus the redesigned step with a live preview.

1. School name field

Ties the name a creator just chose directly to what renders in the preview a moment later.

2. Primary color picker

Lets creators see their brand's actual color, not a placeholder, live as they pick it.

3. Live preview panel

Updates in real time, school name, color, and subdomain all reflect instantly.

HYPOTHESIS 2 - COURSE CREATION

If we simplify course creation during onboarding by breaking it into one clear, guided step with simple decision making, new users will get to a “first lesson published” moment faster and feel less overwhelmed, increasing activation and trial-to-paid conversion.

Phase 3 - ACTIVATION

Product creation collapsed from a five‑step modal into a single "Create your first product" screen inside the flow.

Steps to add thumbnail and pricing plans were removed from the initial flow rather than consolidated, deferring that setup to later.

Before and after comparison: the original five-step course creation modal versus the redesigned single-screen.

1. Simplify product creation

A guided course setup flow during onboarding that asks only essential questions upfront with an optional AI-assisted course generator.

2. Skip to dashboard CTA

The old flow had no exit. The new one gives creators permission to go to their dashboard and explore the product.

3. Brand preview carried forward

Keeps the ownership feeling from the domain step alive, showing creators their course card in their own brand colors.

IMPACT

Creators who completed the flow activated faster

10%

Increase in seller activation*

20%

Increase in product creation

90%

Increase in Brand and domain set-up

*Activation at Teachable is defined by a scoring model: a school is considered activated once it reaches a 60-point threshold across a defined sequence of early behaviors, including domain and branding customization, product creation.

CHALLENGES & LEARNINGS

Product strategy, measurement, and the constraints that shaped my decision making

No controlled A/B test

The team didn't rule out A/B testing for lack of ability. A before/after comparison was judged to deliver a strong enough signal that investing the time to build a full controlled test wasn't worth it relative to the team's three-month learning window.

Defending metrics

Flow completion declined on the surface. The stronger case: users who completed the new flow were more prepared, having already set up their brand and domain, and that preparation showed up as a positive result in the business metrics that mattered, activation and retention.

Shipping in three phases

I knew from the outset I had roughly a three-month investment period and planned three phases to fit inside it. Each phase needed its own validation before the next began, a pace that only works when the timeline is set ahead of time, not retrofitted under pressure.

See other projects

Designed a personalized feature report card to drive upgrades through AI powered workflow

AI powered

0 ENGINEERING EFFORTS

Growth Design

Designed a personalized feature report card to drive upgrades through AI powered workflow

  • AI powered

  • 0 ENGINEERING EFFORTS

  • Growth Design

Designed a provider comparison tool to reduce decision-making time and increase appointment bookings

Mobile-First Design

COMPETITOR EDGE

Healthcare UX

Designed a provider comparison tool to reduce decision-making time and increase appointment bookings

  • Mobile-First Design

  • COMPETITOR EDGE

  • Healthcare UX

Designed an experience for childcare assistance program that help parents check eligibility and track applications in an accessible, transparent way

Accessibility

Concept app

Civic design

Designed an experience for childcare assistance program that help parents check eligibility and track applications in an accessible, transparent way

  • Accessibility

  • Concept app

  • Civic design

Designed a personalized feature report card to drive upgrades through AI powered workflow

AI powered

0 ENGINEERING EFFORTS

Growth Design

Designed a personalized feature report card to drive upgrades through AI powered workflow

  • AI powered

  • 0 ENGINEERING EFFORTS

  • Growth Design

See other projects

Driven by detail and clarity, I design websites, products, and meaningful experiences.

Designed a personalized feature report card to drive upgrades through AI powered workflow

AI powered

Design

Design

Designed a provider comparison tool to reduce decision-making time and increase appointment bookings

AI

Design

Design

Designed an experience for childcare assistance program that help parents check eligibility and track applications in an accessible, transparent way

  • Accessibility

  • Concept app

  • Civic design

CONTEXT

Teachable is a SaaS platform where creators build and sell online courses, and onboarding is the first moment a new creator forms an impression of the product.

DURATION

2 Months

MY ROLE

User Researcher

Product Designer

TEAM

Bia Frenkel - Senior Product Designer

Matthew Malia - Product Manager

David De Los Santos - Engineering Manager

Sarah Driscoll- Director of Design

PHASE 1 - completion

We replaced a single screen of ten stacked dropdowns with a guided, 4-step progressive flow, including brand and domain set-up.

The bet:

Less to face at once → more people finish.

Trade-off:

This increased the total number of steps, a deliberate trade-off, but meant creators were introduced to branding and domain as a natural part of getting started, not something they'd stumble into later in Settings.

Impact

Completion held flat at the population level (14.7% → 15.0%), but among creators who actually entered the flow, 75.5% completed it end to end, the population number just couldn't see that signal on its own.

Phase 2 - ENGAGEMENT & ACTIVATION

Making it feel like their school by redesigning branding and domain setup step introduced in phase 1 with a live preview.

The bet:

Early ownership and investment → more free trials convert to paid.

Trade-off:

Higher design and engineering complexity than a static form, plus extra copy work to make early decisions feel low-stakes and easy to change later.

Impact

Brand and domain customization rose from 30% to 90%+ of new creators inturn increasing user confidence to activate

Phase 3 - ACTIVATION

Product creation collapsed from a five‑step modal into a single "Create your first product" screen inside the flow.

The bet:

Fewer steps → more people create a product during onboarding (and, downstream, convert).

Trade-off:

Fields like thumbnail and pricing were removed from the initial flow rather than consolidated, deferring that setup to later.

Impact

Seller activation improved by 10%, the most consistent result across the whole redesign, holding up across multiple time windows.

IMPACT

Customer value over confidence paid off, and it's how I still approach uncertain decisions.

90%

Increase in Brand and domain set-up

10%

Increase in seller activation

45%

Reduction in support tickets

*Activation at Teachable is defined by a scoring model: a school is considered activated once it reaches a 60-point threshold across a defined sequence of early behaviors, including domain and branding customization.

PHASE 1 - completion

We replaced a single screen of ten stacked dropdowns with a guided, 4-step progressive flow, including brand and domain set-up.

The bet:

Less to face at once → more people finish.

Trade-off:

This increased the total number of steps, a deliberate trade-off, but meant creators were introduced to branding and domain as a natural part of getting started, not something they'd stumble into later in Settings.

Impact

Completion held flat at the population level (14.7% → 15.0%), but among creators who actually entered the flow, 75.5% completed it end to end, the population number just couldn't see that signal on its own.


Phase 2 - ENGAGEMENT & ACTIVATION

Making it feel like their school by redesigning branding and domain setup step introduced in phase 1 with a live preview.

The bet:

Early ownership and investment → more free trials convert to paid.

Trade-off:

Higher design and engineering complexity than a static form, plus extra copy work to make early decisions feel low-stakes and easy to change later.

Impact

Brand and domain customization rose from 30% to 90%+ of new creators inturn increasing user confidence to activate

Phase 3 - ACTIVATION

Product creation collapsed from a five‑step modal into a single "Create your first product" screen inside the flow.

The bet:

Fewer steps → more people create a product during onboarding (and, downstream, convert).

Trade-off:

Fields like thumbnail and pricing were removed from the initial flow rather than consolidated, deferring that setup to later.

Impact

Seller activation improved by 10%, the most consistent result across the whole redesign, holding up across multiple time windows.

IMPACT

Customer value over confidence paid off, and it's how I still approach uncertain decisions.

90%

Increase in Brand and domain set-up

10%

Increase in seller activation

45%

Reduction in support tickets

*Activation at Teachable is defined by a scoring model: a school is considered activated once it reaches a 60-point threshold across a defined sequence of early behaviors, including domain and branding customization.

CHALLENGES & LEARNINGS

Product strategy, measurement, and the constraints that shaped my decision making

No controlled A/B test

The team didn't rule out A/B testing for lack of ability. A before/after comparison was judged to deliver a strong enough signal that investing the time to build a full controlled test wasn't worth it relative to the team's three-month learning window.

Defending metrics

Flow completion declined on the surface. The stronger case: users who completed the new flow were more prepared, having already set up their brand and domain, and that preparation showed up as a positive result in the business metrics that mattered, activation and retention.

Shipping in four phases

Flow completion declined on the surface. The stronger case: users who completed the new flow were more prepared, having already set up their brand and domain, and that preparation showed up as a positive result in the business metrics that mattered, activation and retention.

CHALLENGES & LEARNINGS

Product strategy, measurement, and the constraints that shaped my decision making

No controlled A/B test

The team didn't rule out A/B testing for lack of ability. A before/after comparison was judged to deliver a strong enough signal that investing the time to build a full controlled test wasn't worth it relative to the team's three-month learning window.

Defending metrics

Flow completion declined on the surface. The stronger case: users who completed the new flow were more prepared, having already set up their brand and domain, and that preparation showed up as a positive result in the business metrics that mattered, activation and retention.

Shipping in three phases

Flow completion declined on the surface. The stronger case: users who completed the new flow were more prepared, having already set up their brand and domain, and that preparation showed up as a positive result in the business metrics that mattered, activation and retention.