Turning Web Traffic into Active Mobile Users A conversion funnel analysis for a Fortune 500 financial institution
Client: Fortune 500 Financial Institution | Digital Tools Department
Role: UX Lead & CRO Strategist
Collaborators: UX Designer, SEO, Content
Overview
A Fortune 500 financial institution had a problem: their website wasn’t converting visitors into mobile app users at the rate they expected. Some “Download the App” CTAs were performing well. Others were being ignored entirely. The Digital Tools department needed to understand why and what to do about it.
I led this engagement in collaboration with another UX designer, partnering across SEO and Content teams to analyze the full user journey from the first landing page to final app download. Together we built a data-backed strategy to help the client stop guessing and start making intentional decisions about their CTAs and landing page content.
Visual suggestion: A sanitized flow diagram showing the user journey from blog article → product landing page → download CTA → app store — with drop-off points highlighted.
My Approach
Before making any recommendations, I needed to understand where users were dropping off and why. That meant looking at the full journey, not just the download page, and understanding what users were thinking and feeling at each step.
I oriented myself to the client’s digital ecosystem, reviewed whatever documentation and site metrics were available, and worked with our SEO team to map the pages most likely to influence app download behavior. From there, we identified which content types and user intents to prioritize for research.
Visual suggestion: A sanitized screenshot or diagram of the Digital Tools section of the site — with branding removed — showing the pages included in the analysis.
Research Methods
We used two primary methods to build our strategy.
User Interviews We recruited current customers who were unfamiliar with the mobile app. During the interview sessions, we shared blog articles and product landing pages directly with participants and observed how they engaged with the content. This gave us a firsthand look at what information felt confusing, what was missing, and what was keeping them from taking the next step toward downloading the app.
Website Data Analysis We analyzed existing site metrics and user behavior data to understand how visitors moved through the Digital Tools pages and interacted with the download CTAs. Because we didn’t have access to their primary analytics platform, we partnered with the SEO team to piece together the data we needed (page views, drop-off rates, and CTA engagement patterns) from the sources available to us.
Visual suggestion: Sanitized user interview artifacts — a discussion guide excerpt or anonymized synthesis notes showing the themes that emerged.
Visual suggestion: A sanitized data table or chart showing page-level engagement patterns across the Digital Tools section.
Finding the Real Problem
The data told a clear story. Users weren’t ignoring the download CTAs because they didn’t want the app. They were ignoring them because the content around the CTAs wasn’t giving them enough reason to act. The copy was generic, the value proposition was unclear, and there was nothing building trust or reducing the friction of switching from their existing banking habits.
The deeper issue was motivation, not awareness. Users knew the app existed. They just didn’t know why it was worth downloading over what they were already doing.
Visual suggestion: A side-by-side comparison of a high-performing CTA placement versus a low-performing one — with branding removed and key differences annotated.
Challenges
Keeping the goal in focus It’s easy to optimize for clicks. But a click on a download CTA doesn’t mean anything if the user doesn’t follow through. We kept the team focused on the actual download as the true measure of success, not intermediate engagement metrics that felt good but didn’t move the needle.
Working within real constraints Our recommendations had to be actionable for the content team without requiring a full technical overhaul from the product team. We scoped everything with implementation feasibility in mind so our work wouldn’t sit in a backlog indefinitely.
Making sense of messy data Without access to the client’s primary analytics tool, we had to be resourceful. We partnered with the SEO team to piece together the metrics we needed from available sources. It took longer than expected, and it’s something I’d approach differently with AI-assisted analysis on a future project to help speed things up at the very beginning.
Recommendations and Prioritization
Our research produced 18 actionable recommendations organized into four categories:
- Improve CTA visibility and placement
- Align messaging with user intent
- Build confidence and reduce friction
- Motivate with incentives and context
We built a prioritization matrix to help the client determine which recommendations to pursue immediately versus which to add to their quarterly roadmap. The matrix also flagged which changes were low-risk to implement directly and which would benefit from A/B testing before a full rollout.
Although we ran out of time during the final presentation to walk through the matrix together, it was included in the deliverable as a framework the client could use independently to sequence their next steps.
Visual suggestion: The sanitized prioritization matrix showing recommendations mapped across an impact vs. effort grid — with branding removed.
Visual suggestion: A sanitized slide or summary showing the four recommendation categories with one or two example recommendations per category.
What I’d Measure
If I were tracking performance post-implementation, here’s what I would have prioritized:
- App download conversion rate from Digital Tools pages
- CTA click-through rate by page type and placement
- Drop-off rate at each step of the download flow
- Mobile vs. desktop conversion differences across the journey
- Return visit rate from users who clicked but didn’t download
Visual suggestion: A clean hypothetical KPI dashboard — framed as a measurement plan rather than reported results.
What I Learned
Data quality shapes everything upstream. Disorganized analytics slowed the project down significantly. On future engagements I’d use AI to run a first pass on pattern extraction before the team invests time in manual analysis.
Presentation timing is a deliverable problem, not just a scheduling problem. Running out of time before the prioritization matrix conversation meant the client left without a clear starting point. In hindsight, I’d restructure the presentation to lead with prioritization rather than ending with it so the most actionable part of the conversation happens while there’s still time to have it.
Recruiting screeners need to be tested. Participants slipped through our screener who had already downloaded the app, which skewed some of our interview data. Tightening the qualifying questions and testing them before launch is now a standard step in my research planning.
Mobile and desktop are not the same funnel. We tested primarily on desktop and assumed parity with mobile. CTAs that redirect to the App Store or Google Play behave differently than browser-based flows, and that distinction has real implications for conversion recommendations.