Logo Maker: Raising export rate from 13% to 16.5% through E2E flow improvements

The Opportunity

Logo Maker is a free, no-signup entry point into Adobe Express, built for people with zero design background — small business owners, side-hustlers, anyone who needs a logo fast. Most users arrive from search, not from Adobe directly, so this is often someone's first hands-on proof of what Express can do — and the bet is that experience turns an anonymous visitor into a registered user.

It's also Adobe Express's highest-traffic acquisition surface, drawing 3.5x more unique users than the next most popular Create page. But only 13% exported a logo, and the backend technology powering generation wasn't changing — no new AI, no rendering engine, no rebuild on the roadmap.

The question was: how much could design move the needle on a system that wasn't going to change?

What I Was Solving

Working with PM and data science partners, I mapped the full drop-off funnel to identify where design — not engineering — could make the biggest difference.

Three failure patterns emerged.

• A required field before the first impression. Users wanted to see what made Logo Maker different from other logo tools. Instead, before they could see a single logo, we required them to enter a business name. It was unnecessary friction that deterred potential users before Logo Maker had a chance to make its case.
• Color came too late. Users couldn't filter by color on the listing page. The only way to change color was a random shuffle at the final step — after they had already formed a clear idea of what they wanted.
• The final page was a dead end. At the final editing screen, users could either download their logo or enter the full editor with no guidance. If they weren't satisfied with the result and wanted to tweak the color or font, there was no clear way forward — it felt like their only option was to stop there. The link to the editor didn't explain what they could do next.

Before Logged Out Editor (LOE) was available, we optimized the SEO flow for export completion because our data showed that users who exported were more likely to become Adobe Express users.

Once LOE became available, the opportunity changed. A user who exported a logo and left could remain a one-time Logo Maker user — and the limited experience could leave them with a poor impression of Express. We wanted to route users into LOE instead, where they could experience much more of Adobe Express's capabilities and become higher-potential Express users. The jump points on the final screen weren't just a UX convenience. They were a business strategy.

Six Design Decisions

1. Making the entry field optional

The required business name field was causing early drop-off. Users who hadn't settled on a name — or simply wanted to look around first — were blocked before they saw any value. We removed the requirement, but for users who left the field blank, we generated a temporary fallback business name so they could still see a fully rendered logo at the end of the flow. Almost 95% entered a name anyway, and entry-page drop-off fell by 7%.

Before

After

Exposing the input field on SEO so users can get a head start with the Logo Maker experience

Logo Maker can also be used in-product. We provided fallback business information in the fields.

‹›

Image gallery

2. Skipping Style — and making color a first-class filter

Control flow: business name → style → icon → logo design → rendering. Redesigned flow: business name → optional icon → logo design → rendering. We removed the Style step and moved style into filters on the logo design screen, alongside a new color filter.

Backend color detection wasn't on the roadmap, so the color filter queries Tempo (Adobe Express's template storage system) filenames instead — the blue swatch surfaces results tagged "blue logo." Skipping Style reduced drop-off. 23% of users used the color filter, and that cohort exported at 30%. The same pattern works across other makers without backend color detection.

‹›

Image gallery

3. Auto-advance on selection

The flow required clicking "Next" after every selection — a wizard-form pattern, not how people actually browse visual options. Clicking a tile is the decision; confirming it again is friction without purpose. Auto-advance removed that friction across the entire flow.

4. Segment-aware CTAs at the final step

The original final step had only two CTAs: "Download" and "Open in Adobe Express." They were the same regardless of sign-in status, and many users didn't know Adobe Express itself. As a result, "Open in Adobe Express" didn't communicate what would happen next.

We redesigned the final step around user state. Authenticated users saw "Edit more," while unauthenticated users saw "Sign up to save."

The test results showed a fundamental behavioral difference: authenticated users were ready to keep customizing, while unauthenticated users were primarily looking to download.

Authenticated users

Unauthenticated users

‹›

Image gallery

5. Editing jump points with pre-opened panels

We replaced "Open in Adobe Express" and turned the final step into three clear jumping-off points: "Change color," "Change fonts," and "Change icons." Each tile routed users into the Logged Out Editor with the relevant panel already open, so they could start editing immediately.

For both authenticated and unauthenticated users, Change Color was the most selected action: 25% and 38%, respectively. Among authenticated users, Change Font was the fourth-most-used action at 15%, but delivered the second-highest conversion rate at 53.7%.

6. Continuous template browsing from the final step

Previously, comparing alternatives meant going back to the listing page. In the control, users repeated that back-and-forth and often dropped out along the way. We added a carousel to the final step so they could browse alternatives without leaving, while carrying the business name through automatically.

Results