THE FULL CASE STUDY

TL;DR Scaling a no-code SaaS: Improving retention, adoption, growth & more

TL;DR Scaling a no-code SaaS: Improving retention, adoption, growth & more

TL;DR Scaling a no-code SaaS: Improving retention, adoption, growth & more

How UX-driven feature design and a rebuilt design system reduced churn by 26% and accelerated product development at Candu, a product adoption platform for growth teams.

Industry:

Tech; No-code; Digital adoption platform

Scope:

Product design; Web design

Client:

Candu

No-code editor. SaaS, Product design & Web design
THE IMPACT

Outcomes that moved the needle

Outcomes that moved the needle

Outcomes that moved the needle

23%

23%

23%

Reduced user churn

New editor features directly addressed the friction points causing users to disengage.

25%

25%

25%

Faster product development

Design system expansiong & restructure, cut component duplication and aligned handoff across teams.

18,5%

18,5%

18,5%

Higher user acquisition

A/B tested marketing site redesign improved conversion from visitor to trial signup.

CONTEXT

Live product with real users and real expectations

Live product with real users and real expectations

Candu is a product adoption platform built for growth and product teams who need to ship in-product experiences, onboarding flows, checklists, in-app announcements, growth experiments, and more, without touching the codebase. Its core is a drag-and-drop, no-code web editor/builder.

Candu was already live with active users when I joined which meant every design decision carried real stakes. I owned feature expansion, design system maintenance, and a full marketing site redesign, across a remote team in London and San Francisco.

Friction removal

FEATURE DESIGN . Editor

Image cropping: closing a costly workflow gap

Image cropping: closing a costly workflow gap

The Problem

The Problem

Users regularly uploaded images that didn't fit their layout. Without a native cropping tool, the only workaround was leaving Candu entirely, open external software, re-export, re-upload.

Every product exit mid-session is a retention risk.

This surfaced repeatedly in user feedback. It wasn't a feature request; it was a frustration signal.

Design objectives

Design objectives

  • Allow image adjustments without leaving the platform

  • Reduce decision load, make cropping feel effortless

  • Match industry-standard patterns so learning time is near-zero

  • Integrate naturally into the existing editor interaction model

No-code editor. Image crop-custom

Custom crop: free-form drag handles

Custom crop: free-form drag handles

No-code editor. Image crop-ratios

Preset ratios: one-click common formats

Preset ratios: one-click common formats

Preset ratios: one-click common formats

Activation & Retention

FEATURE DESIGN . Editor

Pre-built content blocks: solving the blank-canvas problem

Pre-built content blocks: solving the blank-canvas problem

The Problem

The Problem

Candu's users aren't designers. They're product managers, growth leads, and customer success teams, people with strong goals and limited design bandwidth.

A blank canvas isn't empowering for them; it creates friction that delays time-to-first-publish.

And when users don't see value fast, churn risk spikes.

Design objectives

Design objectives

  • Shorten time-to-first-publish for new users

  • Drag-and-drop content blocks with sensible brand defaults

  • Full customisation: colors, fonts, spacing, images, copy

  • Maintain design integrity without requiring design skills

No-code editor. Blockes library
No-code editor. Blockes library

Blocks library:categorised sections, ready to drop in

Blocks library:categorised sections, ready to drop in

No-code editor. Drop, customize, publish
No-code editor. Drop, customize, publish

Drop, customise, publish: the core loop in 3 clicks

Drop, customise, publish: the core loop in 3 clicks

User account expansion

FEATURE DESIGN . Collaboration

Share functionality: bringing the whole team into Candu

Share functionality: bringing the whole team into Candu

The Problem

The Problem

Content created in Candu stayed inside Candu. Sharing for review meant exporting to external tools, losing context, and slowing down cross-functional feedback loops.


Beyond the UX friction, this was a growth constraint: individual users couldn't pull teammates in organically, limiting account expansion.

Solution scope

Solution scope

  • Shareable links: unique URL's for direct browser access

  • Email invitations with contextual notifications

  • Permission controls: view, edit, or comment roles

  • Password protection for sensitive content

  • Stakeholder preview page with inline commenting

No-code editor. Link sharing

Link sharing: copy+paste

Link sharing: copy+paste

No-code editor. Password protection

Password protection: optional for sensitive content

Password protection: optional for sensitive content

No-code editor. Email invitation

Email invitation: just 3 clicks away

Email invitation: just 3 clicks away

No-code editor. Stakeholder preview
No-code editor. Stakeholder preview

Stakeholder preview: inline commenting

Stakeholder preview: inline commenting

Drop-off Prevention

FEATURE DESIGN . Onboarding

SDK installation: making the first technical step feel easy

SDK installation: making the first technical step feel easy

The Problem

The Problem

Getting Candu into a product requires a developer to install the SDK. For many users, this handoff to engineering is where momentum dies. A confusing or intimidating installation screen delays activation, and delayed activation increases the risk of users never fully committing to the product.


The goal was to reduce perceived complexity and make the path to "installed" feel achievable for both technical and non-technical users.

Solution scope

Solution scope

  • Reduce cognitive load at a high-drop-off moment

  • Give users a clear choice of installation method upfront

  • Make it easy to loop in a developer without leaving the product

  • Surface live installation status so users know when they're done

No-code editor. SDK installation
No-code editor. SDK installation

Pick: Embed code, Integration, or loop in a developer via email

Pick: Embed code, Integration, or loop in a developer via email

No-code editor. Password protection
No-code editor. Password protection

Copy snippet, add user ID, deploy. Status updates live.

Copy snippet, add user ID, deploy. Status updates live.

Activation

FEATURE DESIGN . Onboarding

Onboarding flows: iterating toward the right first experience

Onboarding flows: iterating toward the right first experience

The challenge

The challenge

New users needed to understand what the product could do for them, set it up, and reach their first meaningful moment, all before losing patience.


There's no single right answer to that problem, so we explored multiple onboarding approaches to find what actually moved users forward.

The approach

The approach

Rather than committing to one flow, we designed and tested different models, each prioritising a different entry point into the product.


Some led with inspiration (live examples, templates), others with action (create first, install later). The screens below are selected moments fromthose different explorations.


Rather than committing to one flow, we designed and tested different models, each prioritising a different entry point into the product.


Some led with inspiration (live examples, templates), others with action (create first, install later). The screens below are selected moments fromthose different explorations.

No-code editor. Onboarding flow
No-code editor. Onboarding flow

Step 2: See live examples. Real customer results shown upfront to build confidence before any setup.

Step 2: See live examples. Real customer results shown upfront to build confidence before any setup.

Step 2: See live examples. Real customer results shown upfront to build confidence before any setup.

No-code editor. Onboarding flow
No-code editor. Onboarding flow

Step 2 (variant): Explore templates filtered by use case. Reduces blank-canvas anxiety early.

Step 2 (variant): Explore templates filtered by use case. Reduces blank-canvas anxiety early.

Step 2 (variant): Explore templates filtered by use case. Reduces blank-canvas anxiety early.

No-code editor. Onboarding flow
No-code editor. Onboarding flow

Step 3: Choose how to publish: embed in existing page or create new. Decision made before installation.

Step 3: Choose how to publish: embed in existing page or create new. Decision made before installation.

Step 3: Choose how to publish: embed in existing page or create new. Decision made before installation.

No-code editor. Onboarding flow
No-code editor. Onboarding flow

Step 5: Installation via Webflow integration. Split-screen guides the user through setup with contextual documentation inline.

Step 5: Installation via Webflow integration. Split-screen guides the user through setup with contextual documentation inline.

Step 5: Installation via Webflow integration. Split-screen guides the user through setup with contextual documentation inline.

Analytics & Security adoption

ENTERPRISE FEATURES

Analytics visibility & enterprise-grade security

Analytics visibility & enterprise-grade security

Analytics overview

Analytics overview

Designed the analytics dashboard for tracking visitor and event data across all content published through the editor. The challenge: surfacing actionable insight without overwhelming non-analytical users.


Clear hierarchy, progressive disclosure, and segment-level breakdowns, within the existing design system.

Multi-factor authentication

Multi-factor authentication

Designed MFA flows for both authenticator apps and SMS, including org-level deployment for admins. Security UX is notoriously friction-heavy.


The goal was to make MFA feel like a normal setup step, not a compliance obstacle that causes users to abandon the flow.

No-code editor. Analytics
No-code editor. Analytics

Analytics: visitor and event tracking

Analytics: visitor and event tracking

No-code editor. Password protection
No-code editor. Password protection

MFA: authenticator app, SMS

MFA: authenticator app, SMS

MFA: authenticator app, SMS

No-code editor. Styleguide

Scalable foundation

DESIGN SYSTEM

The infrastructure underneath everything

The infrastructure underneath everything

The situation

The situation

The original design system was built for a smaller, simpler product. As Candu scaled, the cracks showed: duplicated components, inconsistent patterns, and a library that slowed design rather than enabling it.


Inconsistency at the system level creates inconsistency in the product. And an inconsistent product erodes user trust, which eventually shows up in retention data and revenue.

What I did

What I did

  • Full audit of existing components, patterns, and docs

  • Retired duplicate and outdated components

  • Rebuilt token structure for scalable theming

  • Expanded library to cover new features

Acquisition

MARKETING SITE . Redesign

Testing our way to a higher-converting website

Testing our way to a higher-converting website

The challenge

The challenge

The existing site was product-focused in the wrong direction, it described features, not outcomes. As Candu's capabilities grew, the site failed to communicate its value proposition to the decision-makers most likely to convert: product leads and growth teams with limited time and high expectations.

The redesign needed to bridge the gap between what Candu does and why it matters to someone who controls the budget.

My process

My process

01

Audit existing site

Audit existing site

Identified messaging gaps and conversion blockers.

Identified messaging gaps and conversion blockers.

02

Multiple concepts

Explored different hierarchy and messaging approaches.

03

A/B testing

Compared performance across key conversion points.

04

Maze validation

Usability tested the winner before final build.

02

Multiple concepts

Explored different hierarchy and messaging approaches.

03

A/B testing

Compared performance across key conversion points.

04

Maze validation

Usability tested the winner before final build.

No-code editor. Homepage version A
No-code editor. Homepage version A

Version A

Version A

No-code editor. Homepage version B
No-code editor. Homepage version B
No-code editor. Homepage version B

Version B

Version B

Final A/B testing winner: Homepage version A

Final A/B testing winner: Homepage version A

No-code editor. Homepage version A
No-code editor. How it works page

How it works page

How it works page

No-code editor. How it works page

Page for solutions or case studies

Page for solutions or case studies

Key takeaways from scaling a live product

Key takeaways from scaling a live product

Feedback loops are a design signal, not just qualitative data

Image cropping and other features came directly from user feedback. Building a system to catch these signals early, and route them into the design process is as important as the design itself.

Design systems are a retention lever, not just a dev convenience

Inconsistency in the product creates subtle distrust. Investing in system coherence contributes to the user confidence that drives long-term retention and revenue.

Non-designer users need opinionated defaults

The Blocks feature taught me that empowerment doesn't mean maximum flexibility, it means reducing the burden of decisions that users shouldn't have to make.

The marketing site is the top of the retention funnel

Acquisition and retention aren't separate. A site that sets wrong expectations brings in users who will churn. A well-tested marketing site attracts the right users who stay.

LET'S WORK TOGETHER

Better UX got Candu 23% less churn.
What could it do for you?

If you're building a SaaS product and users aren't sticking, let's have a conversation.

LET'S WORK TOGETHER

Better UX got Candu 23% less churn. What could it do for you?

If you're building a SaaS product and users aren't sticking, let's have a conversation.

LET'S WORK TOGETHER

Better UX got Candu 23% less churn. What could it do for you?

If you're building a SaaS product and users aren't sticking, let's have a conversation.