Maximizer Mobile App
The old app was failing: 84% of CRM users did not have it installed, and its NPS was -4. I changed the plan from small fixes to a full rebuild, sold it to the Head of Product, and designed a new cross-platform app for financial advisors from the ground up, with AI features such as a note taker.
Impact
Faster
3→1
Taps for core actions
Validated
3
Domain experts confirmed the direction
- Product
- Maximizer Mobile App
- Timeline
- Feb to Apr 2026 (rebuild design)
- Status
- Live (MVP), staged rollout
- Role
- Solo Product Designer
- Platform
- Mobile, cross-platform (iOS and Android)
- Team
- 1 designer, 1 PM, 1 .NET MAUI engineer, shared back-end API team
What I Shipped
01 Architecture
Modular and Expandable
Each module stands on its own, one level deep, so we can add, swap, or remove one without touching the rest.
Independent Layers
Data entry, data access, insights, and AI are separate layers. If one is removed, the rest of the app still works.
Pattern System
Three base modules set the core component patterns. New modules are built from these parts instead of designed from scratch.
02 Navigation and IA
Flattened the Structure
Brought deep controls to the surface, so actions that took 3 taps now take 1.
Learn One, Know All
Every module uses the same layout, filters, and AI placement. Once users learn one module, they know them all.
One Clear View per Module
Removed redundant view toggles. Each module has one layout that shows what matters most.
03 Data Entry
Notes Open When You Need Them
When an advisor starts a call from the app, the note taker opens as soon as they come back to it, ready to capture what was said.
Manual First, AI on Top
Manual entry always works. The AI layer reads the note and pulls out tasks to approve, if the advisor opts in to AI.
Capture in Any Format
Type a note or take a photo of a handwritten one. Both go into the same note taker. Voice notes are designed, not yet shipped.
04 Cuts and Prioritization
Cut Dead Weight
Removed legacy features that were broken or unused, instead of rebuilding them.
Retired the Map Module
Removed the standalone map views. Kept the location API to send meeting addresses to Uber or Lyft and to map apps for directions.
My Role
Championing the Rebuild: Made the case to rebuild the mobile app and won approval from the Head of Product, backed by concept work I created to show what the app could become.
Defining the Build: As the solo product designer, partnered with the PM and Head of Product to decide what we would build, mapping use cases and priorities before design started.
Mobile Design System: Extended the web app's design system to mobile, in light and dark themes: one cross-platform design (.NET MAUI) with native components and the same flows on iOS and Android.
Cutting What Didn't Work: Pushed to drop the parts of the old app that weren't serving users, so the rebuild stayed lean and focused.
Why It Was Time to Change
By 2024, adoption had stalled and the app had no real edge over competitors. The cost of maintaining it was more than the few remaining users could justify. The main reasons:
Not Built for Advisors: Built for a generic salesperson, with no tools, wording, or structure for financial advisors.
No Overview: Advisors had no at-a-glance view of their business.
Painful Data Entry: Capturing notes during and after meetings was so hard that the feature was effectively unusable.
Buried Features: The complex structure hid useful features.
Slow and Heavy: The app loaded all data at once, which made it slow.
No Guided Flows: Multi-step processes had no dedicated flows, so users got lost.
Missing Capabilities: The app lacked features advisors needed for daily work.
84%
of CRM users did not have the app installed
-4
Average NPS of the legacy app
Improve or Rebuild
Question
Leadership had to choose early: keep improving the existing app step by step, or rebuild it from the ground up?
Driving the Case to Rebuild
I ran an early heuristic audit of the existing app and mapped the full application to expose the complexity of its information architecture. I then produced the initial concepts and simplified flows to show what a leaner experience could look like.
Shaping the Decision
I paired those findings with the business and technical signals: low adoption, and the end of Microsoft's Xamarin support, which forced a migration to .NET MAUI regardless. Together, these made the case I brought to the Head of Product: build a new app instead of another round of small fixes. The Head of Product approved the rebuild.
Goals
Business
- Grow Active Adoption: Get advisors to install the app and use it daily, since 84% of CRM users did not have the existing app installed, and many who did rarely used it.
- Refocus on Financial Advisors: Rebuild the app around financial advisors rather than the generic salesperson it was designed for.
- AI-Ready Foundation: Create an architecture that can absorb AI-driven capabilities later without costly rework.
Product Design
Primary: Earn a Place in the Advisor's Day
- Business at a Glance: Surface the state of the advisor's book of business the moment they open the app.
- Scalable Information Architecture: Establish a clear IA and consistent UX patterns across modules so users can move through the product without relearning it.
- Fast, Flat Workflows: Put core actions one tap away, and make note capture quick enough to complete right after a meeting.
- Modern Design System: Extend the web app's design system to mobile, with native components on each platform, scaling to new features without visual debt.
Next: Build Competitive Advantage
- Context-Aware Companion: Layer AI insights onto the new foundation to turn the app from a system of record into a proactive companion.
Product Considerations
Preserve the Backend: The existing back end set the limits of what I could design. For the MVP, I designed inside those limits and asked for no changes. The back end was well established, so the limits were small.
Adopt a New Front-End Platform: I learned how .NET MAUI renders on iOS and on Android, and how different the two look, so I could design one set of flows that works on both.
Design Within Technical Limits: Partner directly with engineering to understand constraints and design within them.
Changing ICP Focus
The rebuild started after Maximizer's late-2024 shift of its ideal customer profile (ICP) to financial advisors, so the new app was designed for them from day one.
Assistants already used the web CRM every day. Advisors did not. The mobile app is the advisor's companion, built to bring them into the CRM.
The mobile app has no manager view of all branches and teams. Managers use the web app for that. On mobile, advisors and assistants handle their daily work.
User Personas
| Persona | Primary Objective | Strategy |
|---|---|---|
| Advisor | Stay on top of the business at a high level: how much they're earning, whether any clients are at risk, and the actions that need attention right now. Show up for the meetings that matter. | Provide a high-level read on the state of the business plus prioritized, action-ready notifications: whether today needs them or can run without them, and exactly what to do when it does (this meeting, this call, this dinner). They should open the app and instantly know how to spend their day. |
| Assistant | Keep daily operations moving. Manage processes and email. | Speed up high-volume work: fast record creation and editing, bulk actions, and inline validation. |
My Approach
Stable Foundation: Treated the legacy system as stable infrastructure and rebuilt the experience on top of it.
For example, we kept the existing APIs and replaced the interaction model, IA, and interface.
Role-Based Experience: The app should fit each user's role and needs.
For example, advisors and assistants get different IA, data-entry methods, and insights, so the same app feels different to each.
Golf Rule: One look at the app should tell an advisor whether they can go play golf or need to be back at the office.
For example, I designed the home screen so the advisor sees KPIs, alerts and nudges at a glance, and can plan the day.
Flexible AI Adoption: AI is optional. The app works fully without it.
For example, an advisor can type a note and be done, with no AI involved. Or they can opt in to AI: take a photo of a handwritten note (live) or record a voice note (designed, not yet shipped). The AI does the rest, and the advisor reviews, edits and accepts.
Research and Validation
Old Note Capture: In the old app, advisors had to type notes or rely on the phone's speech-to-text, then add each task by hand, one at a time. The app was so clunky that any other method was easier.
Adoption Data: Compared the number of people with the app installed to the total number of Maximizer seats. 84% of CRM users did not have the app installed, and many who did rarely used it.
Customer Feedback: The Customer Success report showed an average NPS of -4 for the legacy app.
Expert Review: Presented the concepts, planned use cases and flows, scope, and what AI could do for advisors to 3 retained domain experts in CRM, investment management, and insurance.
Staged Rollout: Released to the internal team first, then to trusted business partners, then to the public, so each group could surface problems before the next one saw the app.
Spotlight: AI on Mobile
The Problem: Capturing a meeting took too long. Advisors typed the note, then created every task by hand.
The Design: One note taker where advisors type a note or take a photo of a handwritten one (live). The AI turns the input into a clean note and pulls out the next actions.
AI Suggests, the Advisor Decides: Every task the AI proposes is a preview. The advisor edits it or accepts it with one tap.
For example, after a call with James Robson, "Call him next week" becomes a task preview: "Call James Robson next Tuesday, 2 pm." James comes from the call or meeting the advisor just had. Tuesday 2 pm is the first free slot in the advisor's calendar a week out. The task for the right contact is live. The suggested time slot is designed and in the pipeline. It only appears when the company uses IQ Boost (Maximizer's AI feature set) and the user keeps it on.
Privacy by Design: An anonymizer service built by our engineers removes names, dates, and other private data before anything reaches the AI. Those details come back from our own data, so the AI never handles them.
Honest Answers: I worked with the front-end engineer who owned the prompts to add an instruction that tells the AI to correct a false premise instead of accepting it.
For example, if asked how much a client paid in a divorce, the AI should answer that the client is not divorced as of the last update, and link to the record in case that has changed.
Known Limit: The AI still accepts a false premise sometimes. For example, asked "How much did he pay in the divorce?", it can answer "$0". That is technically true, but misleading: the client never divorced. I proposed a small agent that checks AI answers against CRM data through the Global Search API. It did not fit the MVP's scope.
Clear Scope in the MVP: Household Q&A lets advisors ask questions about one client household, meaning a family's accounts and members. It works on one household at a time. I kept the selected household pinned at the top, so advisors always know what they are asking about. Switching to another household is quick.
Client Summaries: IQ Boost summaries are live in the Contact list.
Feedback Loop: Thumbs up and down on AI answers (live) show us where the AI falls short.
Designed, Not Yet Shipped: Voice notes, plus call and meeting recording, are designed and wait on back-end and permission work before release. Before recording starts, the app reminds advisors in two-party consent states and provinces to tell every participant they are being recorded.
Built to Grow: Some features wait on the back end, such as matching names in a note to existing contacts on the fly. The design is ready, and I update it as each one lands.
Outcomes and Impact
Expert Validation: The 3 domain experts confirmed the direction before build.
"You hit the nail on the head," said a former CEO of a national wealth-management firm.
MVP Live: The note taker, AI note extraction, suggested next actions, IQ Boost summaries and Household Q&A are live, after a staged rollout to the internal team, trusted business partners, and the public.
Faster Core Actions: Core actions that took 3 taps now take 1.
Adoption Targets: 25% active use at 12 months and 75% at 3 years. Post-launch data will come after the first marketing cycle.
Reflection
The Rebuild Was the Safer Bet
The web app and the mobile app faced the same question and got opposite answers. On the web, users' income depended on the app, so it was renovated. On mobile, few people used the app. A simple app teaches itself, and the platform change forced new work anyway. Money already spent on an old app is not a reason to keep it.
Design for the Many, Not the Few
Most of the old app's complexity served a handful of high-value clients, and it made the app harder for everyone else. We cut it and designed for what most users needed.
Cut the Feature, Keep the Capability
Before removing a feature, ask what job it does. The feature can go while the job stays.
On Mobile, Speed Is a Feature
A fast, focused app beats a feature-rich one that slows people down.
Removing Is Harder Than Adding
Good design work means questioning past decisions, not only improving them.
Next Case Study