Smart Agency September 2024 – February 2025

Earning Users' Trust

How I helped a new startup increase its onboarding rate by 935%.

My role: Lead UX Designer

Smart Agency was an ambitious spin-off of the Marketing Agency TERRITORY, creating a platform for small businesses to automate their marketing with AI. I joined shortly before the product went live, but I saw three major flaws with the MVP that was about to be released.

The Problem

1

The signup process had 17 steps

  • 1 welcome page
  • 13 briefing process steps
  • 3 registration and payment steps
user flow showing screenshots of every step in the onboarding process
2

Test users were confused by the result

We showed an abstract "brand book" when they expected to see an AI-generated website.

"What's a brand book?"
"Where's my website?"
"What the heck is this?"
Screenshot showing the brand book interface that confused users
3

Payment required before seeing value

Users had to provide business and payment details BEFORE they could see the product. As an unknown entity, Smart Agency was expecting too much user trust before showing any value.

Screenshot of payment screen that appeared before users could see the product

My Proposal

Let users see the platform features for themselves before we ask them to choose a plan or provide payment details.

This is an established best practice for new SaaS products that lack brand recognition or have an unfamiliar kind of product. Smart Agency had both of those problems, so expecting users to provide payment details (even with a free first month) was not going to work.

1

Move account creation to step 1

Reasons:

  • Get users' emails in our system up-front so CRM can contact them even if they don't complete the briefing process
  • Jumping straight into answering briefing questions confused test users. Account creation sets expectations
  • User can come back and complete the briefing later without losing their progress
Proposed new signup flow including a mockup of the new account creation page

Implementation:

  • Create account without a password
  • After email verification, user is logged in
2

Redesign the briefing result to match expectations

Reason:

The main USP on the Smart Agency website and in marketing campaigns stated: "GENERATE YOUR COMPANY WEBSITE IN MINUTES," so users expected to see their website after answering all the briefing questions. Seeing a brand book instead felt to them like a bait and switch.

Mockup of a brandbook picker where elements are styled to look like a website homepage

Note: For legal and technical reasons, we could not do the actual website generation at this step, so we needed to redesign the brand book to look like a website preview.

3

Move registration and payment after onboarding

Reasons:

  • There was no compelling user-centric reason for account creation to be linked to company registration and payment
  • The freemium model allows a user to access the platform quickly, after a short account creation process. The platform itself becomes a marketing tool, as users see the value of basic features for themselves, but also see what more can be unlocked with a paid plan
  • Diagram of a user flow where plan selection and registration/payment come after user has accessed the product's cockpit.

The Challenge

Accepting my proposal would mean rejecting the revenue-first business model that the C-Suite had prepared; a model where users were locked in to an expensive subscription after a 30-day free trial.

My proposal required a paradigm shift to a "Freemium" model where users can play around with feature-limited platform tools for free and upgrade to a paid plan once we had proven our value.

My proposal would touch all the product streams and generate a lot of JIRA tickets.

My proposal was a challenge to the status quo

and Leadership rejected it...

until after the product went live.

Making My Case... Again

A week after launch, I had the data I needed to make my case in terms that leadership and POs would accept.

2%

Only 2% of visitors finished the onboarding process.
And they all subsequently canceled their plans before the 30-day trial was up.

Critical Insight

The highest drop-off rates occurred on the steps I had first flagged and proposed alternatives for.

My proposal was no longer an unnecessarily risky challenge to an agreed-upon set course. It was the SOLUTION to a new existential problem: literally nobody was signing up for the platform.

All the previous stakeholder resistance disappeared. Reframed and repackaged for the new reality, my three small proposed tweaks didn't look so risky anymore.

The Results

Approval

My re-proposed onboarding flow was approved by all stakeholders a month after the product launch.

Implementation

All four product streams worked on all cylinders for three sprints in order to implement all the necessary changes to the front end and back end.

I liaised with all POs during the process, and also worked as an IC redesigning the brand book step.

Impact

935%

5 days after we shipped these updates, the onboarding completion rate had increased by 935%.

Key Takeaways

Data Changes Minds

When stakeholders aren't initially receptive to UX recommendations, real-world data from launch can reframe the conversation entirely.

Trust Must Be Earned

For unknown brands, asking for payment before demonstrating value creates an insurmountable barrier. Show first, ask later.

Persistence Pays Off

Being willing to re-pitch a rejected proposal when circumstances change is crucial. The best idea at the wrong time needs patience.