TL;DR

Texter is CM.com’s global tool for testing SMS routes and operators. After seven years, users no longer trusted it, and the relationship between users and the product team had broken down. The product team expected a visual refresh. I showed the real problem was trust, and built a collaborative process where users, the product team, and the business made decisions together, based on real data instead of assumptions.

Results: 30% fewer workarounds · 25% more scheduled tests · 40% fewer false positives · 80% more users actively shaping the product · Data tracking and user testing became a standard

Context

Texter helps teams choose the best SMS routes and operators by balancing price, quality, and delivery success worldwide. It’s used across Europe, APAC, the USA, and Latin America by support, operator, smart routing, and wholesale teams, each with their own culture, biases, and knowledge of local operators and clients.

When I joined, every group was pulling in a different direction:

• Users didn’t trust the tool. They double checked results in other apps, databases, or with clients, and relied on personal experience to judge routes. After past redesigns ignored their feedback, they were skeptical about getting involved again.
  
  

• Teams had built their own Power BI dashboards, databases, and manual tools to get their work done, and were protective of them. They feared a redesign would make all that work worthless.
  
  

• The product team believed a visual refresh would solve the problem, and made decisions based on assumptions about how users worked.
  
  

• The business was wary of exposing technical complexity to users.

It was clear this wasn’t a visual redesign. I was brought in to lead the end to end UX redesign, and the real challenge was fixing the collaboration between users and the product team. I started with three key decisions.

Decision 1: Treat it as a trust problem, not a visual one

Task: Understand how users actually decided on routes and operators, and get everyone to see the same picture.

I ran interviews and shadowing sessions with 6 user teams across four regions, selected for their different usage patterns and backgrounds, and mapped the manual workflows, dashboards, and spreadsheets they used outside Texter.

The findings were clear:

• The scheduled test feature, one of Texter’s most valuable functions, was barely used. The last visible scheduled test was over three years old.
  
  

• Results were hard to interpret: difficult to sort and filter, with no easy way to compare expected and received messages, and no visibility into test coverage.
  
  

• Users constantly cross checked Power BI and other tools to understand real route quality.

Instead of presenting conclusions, I brought users and the product team together to analyze them. I presented the current user flows, showing every step users took outside Texter to make a decision, and we ran collaborative workshops with empathy maps and personas. Together we prioritized four problems: trust, collaboration, result visibility, and batch test usability.

Impact: For the first time, the product team saw the real workflows behind the complaints, and agreed that many UI fixes were obvious but the real barrier was the lack of transparency.

Decision 2: Make collaboration the product strategy

Task: Repair the relationship between users and the product team.

The biggest problem wasn’t on the screen. Misalignment between users and the product team was causing communication barriers, double work, manual workarounds, and was blocking Texter from evolving. So I designed a process with user data at the center, so both sides could work from the same facts:

• A public user feedback database. Collecting feedback wasn’t enough; users needed to see it being used. I created a database visible to both users and the product team, which became the central place for product decisions.
  
  

• Recurring alignment sessions with both sides. With the product team, to keep user feedback in the roadmap. With users, to show improvements and share insights across teams. Many users said they picked up useful ideas from seeing how other teams worked.
  
  

• Respecting existing work. Instead of replacing teams’ dashboards and spreadsheets, I used them as research input and design references, showing users their work would shape the new Texter rather than disappear.

Impact: Users went from skeptical to actively shaping the product, with 80% more users taking part in feedback sessions. The feedback database became the shared source of truth for the roadmap.

Decision 3: Let the data settle disagreements

Task: Convince a product team that relied on assumptions, without making it a fight.

The product team wasn’t convinced by this approach yet, so instead of arguing, I picked one small, visible problem to test it on: showing the difference between expected and received messages.

1. Gather assumptions. I asked the product team how they thought users were reading the data.

2. Brainstorm together. We generated possible solutions as one team.

3. Test with users. In a user workshop, different groups tried different prototypes. We measured how accurately they spotted the difference and how long it took them.

4. Share back. I presented the users’ results and feedback to the product team.

5. Reflect. We discussed the gap between assumptions and real user needs, and how real insights should guide design decisions.

Because Texter is complex, low fidelity prototypes didn’t work for users. I used Figma Make to build interactive prototypes with realistic data, including batch tests, multiple operators, and coverage stats, so users could give feedback on something that behaved like the real tool.

Impact: The product team saw the gap themselves and adopted data-backed decisions as the way of working. Users could finally hear how the product team understood their feedback, which made both sides more open to collaborating.

Redesigning Texter

With trust and a shared process in place, we worked in small, user focused iterations, with users reviewing every step.

Making coverage visible

Users didn’t understand that sending a test to one node ten times isn’t the same as sending it once to ten nodes. Only the second measures real route quality, so users were unintentionally creating false positives. The business was initially against exposing technical details to users, so I used the user test results to make the case, and added a Coverage section to test creation that shows operator and country distribution.

Impact: False positives dropped 40% in the beta group, proving that transparency improved both data accuracy and trust.

Clearer results

The results page got sortable, filterable batch tests and a clear comparison between expected and received messages, so users could judge route quality without leaving Texter.

Impact: 30% fewer workarounds outside the tool.

Bringing scheduled tests back

The old tool made scheduling easy but showed no progress or outcomes, so users forgot tests and wasted traffic. The new Test Schedule page shows results, editable recurrence, and content settings, plus insights on each batch’s health.

Impact: 25% more scheduled tests, improving route monitoring and cost efficiency.

McMoney App: growing Texter’s coverage

With trust rebuilt and the team focused on the right problems, we finally had room to look beyond Texter itself, at McMoney.

McMoney is the app that powers Texter’s coverage. People around the world install it and earn money for receiving test messages, and every active user becomes a node Texter can test on. More active users means more nodes, more countries and operators covered, and more reliable test results worldwide. Making McMoney successful was the most direct way to make Texter better.

I led the redesign of McMoney with one goal: make it rewarding enough that people keep it installed and invite others. The current app was legacy without information or engange mechanism so we focus on them in the first iteration: referral bonuses, levels, achievements, progress tracking.

Impact: when I leave McMoney was in development so I have no tracked impact of the app.

Results

Texter went from a tool users double checked to one they rely on, and the product team now builds with users instead of for them.

✓ 30% Fewer workarounds
✓ 25% More scheduled tests
✓ 40% Fewer false positives
✓ 80% More users actively shaping the product

The rebuilt trust opened conversations that weren’t possible before: merging three or more related apps into Texter, and freeing team capacity for McMoney, an app that sources global users to send test messages into Texter and improve test coverage worldwide.

“They didn’t just change everything, they kept what worked and added things that actually made our jobs easier.”
— Wholesale user, CM.com

What I learned

Trust takes longer to rebuild than a UI. Users who had been ignored before needed to see their feedback used, again and again, before they invested in the product.

Transparency works in both directions. Users needed to see how the product team understood their feedback, and the product team needed to see real user data. Making both visible turned two frustrated groups into one team.

What I’d do differently: This project is one of the ones I enjoyed more, I gain the trust to use real user insights in the process. And all the features and changes we did created way more impact that any blind redesign. What I would do differently is go faster to more B2C part of the platform, as when we got to the app redesign a lot of that feedback became really valuable and missed by our clients.