Research → information → action
Wilder
Visitor questionWhere do I go for help or research?
Next moveFind a service, resource, or way to act
Shipped nonprofit platform
Research · UX/UI · WordPress development
Wilder had thousands of pages. People still needed one clear next step. I rebuilt the experience around what they came there to do.
I led stakeholder and user interviews, UX analysis, information architecture, wireframes, interface design, responsive WordPress development, content migration, and the Research Library integration. The goal was to make a complex organization easier to understand and make services, research, giving, volunteering, and contact easier to find.
See the decisions behind the rebuild. Research, structure, migration, integration, and post-launch behavior.NonprofitServices + research experience
2 systemsMain site + Research Library
End-to-endResearch through development
01 People
The old site asked people to understand Wilder before it helped them.
Wilder.org had grown into a large Drupal site with thousands of pages, heavy menus, dense content, and several different audiences competing for space. The redesign had one practical job: help people find services, research, and ways to support Wilder without first learning the organization chart.
This was not a small marketing site. Someone might arrive looking for mental-health support, child-development services, caregiver resources, a research report, a partnership contact, a volunteer role, or a place to donate. Those visitors were not starting from the same question, and some were arriving under real pressure on a phone.
I led the UX, UI, and front-end work from planning through launch. I used stakeholder and user conversations, analytics, heatmaps, traffic patterns, content audits, and the existing site structure to find where people were getting lost. I synthesized that work into a UX analysis and presented the findings and redesign direction to stakeholders. Then I rebuilt the information architecture, wireframes, interface, WordPress system, content model, responsive behavior, and Research Library experience around clearer first moves.
I also stayed inside the implementation. The work included the Drupal migration, custom WordPress development, Elementor and ACF components, search and filter behavior, the Presto research connection, XML imports, taxonomy cleanup, QA, and post-launch analytics. That kept the user experience connected from the first interview to the live page.
The question behind the redesign
Can someone find the right next step without understanding how Wilder is organized?
What I studied before changing the structure
The old site contained too much history and too many competing needs to redesign from opinion. I needed to see how staff described the work, how visitors moved through it, what they noticed, and what the content system could realistically support after launch.
Stakeholder and user conversations
I interviewed people across the organization and listened for the questions visitors asked before they could act. That exposed the gap between internal language and the way someone seeking help described the same need.
Analytics and traffic paths
I reviewed high-volume entry pages, exits, internal movement, device use, search behavior, and the routes people took toward services, research, contact, volunteer, and donate pages.
Heatmaps and behavior review
I looked at where attention stopped, what people tried to click, how far they scrolled, and where dense desktop layouts became especially difficult on smaller screens.
Content and governance audit
I mapped page families, duplicate ideas, outdated routes, naming conflicts, editor needs, and the content that had to survive migration without carrying the old structure forward.
Four experiences had to work at the same time
The redesign could not optimize one audience by hiding the others. It needed a shared system that let each person recognize a path quickly, then move into the right level of detail.
A person seeking services
May be stressed, uncertain about terminology, and using a phone. The experience has to identify a useful route before asking for patience.
Needs to know: Where do I go, and what happens next?
A community partner or researcher
Needs to find evidence, reports, authors, topics, and partnership information without fighting a separate database or guessing Wilder’s internal labels.
Needs to know: Can I find the right research and understand how it connects?
A donor, volunteer, or advocate
Needs a credible picture of Wilder’s work and a clear action. Giving and involvement should feel connected to the mission, not buried in utility navigation.
Needs to know: Where can I help, and why does it matter?
A Wilder editor
Needs safe, repeatable page patterns and a content model that can be maintained without rebuilding layouts or recreating the sprawl the redesign removed.
Needs to know: Can I keep this current without breaking the system?
02 Structure
I rebuilt the site around first moves, not departments.
The old navigation tried to expose the institution. The new experience needed to expose the next useful action. I used the research to turn a large page tree into a smaller set of recognizable paths, then gave each path a consistent way to unfold.
01
What brings me here?
Services, research, partnership, giving, volunteering, or learning about Wilder.
02
Which path sounds like me?
Use direct labels and task-based choices instead of internal program structure.
03
Can I understand it quickly?
Lead with plain language, readable hierarchy, and enough context to choose.
04
Can I trust the next step?
Use stories, program detail, research, and clear expectations near the decision.
05
What do I do now?
Find services, request an appointment, read a report, contact, donate, or volunteer.
Principles that drove the reset
These rules helped the team decide what to keep, what to combine, where to place it, and how the same logic should behave on a phone and in the content-management system.
Start with the task
Navigation and page openings used the visitor’s question first. Organizational detail remained available, but it no longer had to be decoded before someone could move.
Give each page a job
A page could inform, route, reassure, or convert, but it could not do all four equally. One primary next step kept dense content from turning into another menu.
Rebuild the mobile order
Mobile was not a compressed desktop layout. I reconsidered navigation, section order, reading length, tap targets, and the distance between a need and the action that answered it.
Treat maintenance as UX
The public experience would decay if editors had to improvise every page. Reusable templates, fields, content types, and guardrails were part of the design, not an administrative afterthought.
The site exposed its complexity.
- Large menus attempted to show too much at once.
- Services, research, giving, and organizational content competed in the same hierarchy.
- Long pages made important actions difficult to find.
- Separate systems made the Research Library feel detached from Wilder.org.
- Desktop structure carried into mobile even when it created long, fragile paths.
The site exposed useful choices.
- A smaller set of task-based sections made the first decision easier.
- Services, stories, research, impact, and involvement used consistent page patterns.
- Primary actions appeared closer to the need they answered.
- The Research Library became a searchable part of the same experience.
- Mobile layouts were reordered around comprehension, reach, and action.
A reset, not a reskin
The visual change is obvious. The important change is underneath it: fewer competing paths, clearer page jobs, a more direct service hierarchy, and a connected research experience.
Previous site
Redesigned site
I used the existing site as evidence, not as the blueprint. The redesign kept Wilder’s depth while changing how that depth was organized and revealed.
03 Build
The UX stayed connected because I carried it into code.
I did not stop at a Figma file. The content model, responsive order, component behavior, migration rules, search states, editor controls, and build QA all changed the final experience, so I stayed responsible for them through launch.
01
Understand the real system
I brought interviews, analytics, heatmaps, content audits, platform constraints, and stakeholder priorities into one working picture before proposing a new structure.
02
Model the information
I defined navigation, page families, custom post types, taxonomies, field groups, and relationships so the design could scale without recreating the old sprawl.
03
Wireframe the first moves
I worked from the visitor’s next question: where do I go, what does Wilder offer, can I trust this, and what action belongs here?
04
Design a reusable interface
Hero bands, service cards, stories, calls to action, research cards, filters, typography, spacing, and states became a shared visual and interaction system.
05
Build and migrate
I built the WordPress theme and components, moved content from Drupal, connected structured research data, and cleaned up the edge cases that did not fit a simple import.
06
Test the whole path
I reviewed responsive behavior, search and filters, content states, accessibility, links, forms, analytics, performance, and editor workflows before and after release.
Representative reconstruction
The wireframe was a sequence of decisions.
The original wireframes were working artifacts. For this case study, I reconstructed the core layout in the same clean system used across the portfolio so the structure is easier to compare from project to project.
The frame shows the logic I designed: a clear orientation point, immediate routes for services and support, human proof, a focused services section, and a final action. On mobile, those same ideas were reordered rather than merely stacked.
The question was never “what section looks good next?” It was “what does this person need to decide next?”
Human proof
Action
The system behind the pages
The public interface and the editor experience had to support each other. I built reusable rules so a large institution could keep adding and changing content without losing the clarity of the redesign.
Task-based navigation
Clearer labels and page groupings brought services, research, impact, advocacy, involvement, and organizational content into a more predictable hierarchy.
Reusable page templates
Services, programs, stories, events, staff, and research used consistent structures so visitors could learn the interface instead of relearning each page.
Content types and taxonomies
The WordPress model matched the content instead of forcing everything into pages, supporting structured services, programs, news, events, staff, and research.
Editor-safe components
Elementor widgets and ACF field groups gave staff flexibility inside controlled patterns, reducing the need to recreate layout decisions by hand.
SEO, performance, and accessibility
Semantic structure, readable hierarchy, contrast, responsive media, page speed, metadata, and crawlable content were built into the templates rather than added at the end.
Measurement and QA
Forms, key pages, paths, responsive states, search behavior, and post-launch analytics gave the team a practical way to find problems and keep improving.
Design into code
The build was part of the design.
Because I designed and developed the system, I could protect the hierarchy while solving real implementation problems instead of letting the experience drift between comp and launch.
- Custom WordPress theme and front-end components
- Elementor widgets and ACF field groups
- Structured content types and taxonomies
- Drupal migration and manual cleanup
- Responsive QA, accessibility, SEO, and performance foundations
Mobile path
A person looking for help should not have to fight the desktop site on a phone.
Mobile made the cost of the old hierarchy obvious. I shortened openings, simplified navigation, increased tap space, moved task choices forward, and kept the path to services, appointments, research, contact, and support visible without turning every section into a button wall.
- Task choices moved closer to the top
- Navigation collapsed around the highest-value paths
- Headings and paragraphs were rewritten for smaller reading windows
- Cards and calls to action were built for touch
- Section order changed when the mobile decision required it
04 Research Library
The Research Library was a product inside the website.
A search box was the visible part. Under it sat an external Presto database, more than a thousand source records, PDFs, dates, authors, topics, clients, studies, duplicate content, and years of metadata that had to become reliable WordPress content without losing meaning.
The old research experience felt separate from the rest of Wilder.org. Visitors could know the kind of report they needed and still struggle to find it because the data lived in another system and the labels did not always behave consistently.
I treated the library like a product: define the search tasks, model the content, build the import and connection, normalize the data, design the filters and cards, handle empty and error states, and make every result lead to a useful detail page or PDF.
The hard part was not drawing the interface. It was making years of research behave like one coherent, searchable system for visitors and one maintainable content pipeline for Wilder.
Read the source
Presto and XML supplied research records, metadata, files, dates, authors, topics, clients, and study references.
Filter public records
XPath rules admitted only documents marked public instead of trusting the position of a field in a changing XML record.
Normalize the data
Custom functions cleaned dates, filenames, PDF paths, author values, taxonomy terms, and inconsistent metadata.
Map the relationships
ACF fields, WordPress taxonomies, existing studies, client organizations, topics, and files were connected to the right posts.
Make it searchable
AJAX search and filters turned the structured content into a usable library without forcing a full page reload for each choice.
{PropertyValues/PropertyValue[@name="Public document"]/text()} = yes
One small example of the implementation thinking: the import used the field name, not a brittle numbered position, to include only records Wilder had marked for public use.
The problems hidden behind the search interface
The import logs moved beyond record 1,000. At that scale, a small mapping error becomes a site-wide content problem, so I built rules for updating, reusing, cleaning, and connecting data instead of treating the migration as a one-time copy.
Field and content mapping
Abstracts, report IDs, page counts, public descriptions, excerpts, publish dates, social images, PDFs, study references, and other ACF fields were mapped into a consistent research model.
Date and file normalization
I worked through timezone and format problems, cleaned PDF paths, created or reused attachments, and made sure the visible date and downloadable file matched the source record.
Duplicate and update protection
Unique keys and update behavior allowed existing research posts to be found and updated rather than recreated, while duplicate taxonomy terms were detected and reused.
Authors and taxonomy cleanup
Names containing commas, overlapping topic labels, client organizations, project types, subtopics, and other metadata required normalization before filters could be trusted.
Study relationships
Research reports were connected to related study posts and shared metadata so visitors could move between a publication and the broader body of work around it.
Staged API activation
I set up and tested the Presto connection during the build, then held the automatic pull until the full site was stable instead of letting live synchronization complicate unfinished templates.
A research system that felt like Wilder.org
Visitors could search by keyword and narrow results by topic or author, scan useful metadata, and open a report detail without being sent into a disconnected interface.
The interface became simple because the data work underneath it was not. Search quality depended on content modeling, import rules, normalized metadata, and reliable relationships.
The hard part was not the search box. It was making years of research behave like one product.
The visible result was a cleaner library. The work behind it was content architecture, data cleanup, relationships, files, imports, filters, error handling, and a staged connection between systems.
05 Results
The first 30 days showed which paths people were actually using.
I compared GA4 data from July 7–August 5, 2025 with the previous 30 days. The strongest signal was not one site-wide number. It was the growth on the exact paths the redesign had worked to surface: services, research, volunteering, donating, contact, and mobile use.
+26%
16,416 page viewsHome
The redesigned homepage recorded 3,435 more views than the previous period.
+365%
8,710 page viewsMental Health Services
One of Wilder’s most important service paths became far more visible and heavily used.
+238%
2,476 page viewsResearch Library
The new library landing page drew more than three times the previous period’s traffic.
+28.5%
4.8K active usersMobile
Active users on mobile increased after the responsive redesign, with tablet use up 52.9%.
The paths closest to action moved too.
Page-level growth does not prove that one interface decision caused every change. It does show that people were reaching and using the service, support, research, and contact routes the redesign brought closer to the surface.
+1,174%
Volunteer page views reached 3,312.
+235%
Donate page views reached 734.
+317%
Contact Us page views reached 1,798.
+25%
Referral sessions reached 1,696.
+707%
Tracked visits to Get Involved reached 5.5K.
+199%
The main Services landing page reached 2,985 views.
Source: GA4 page-path, channel, engagement, and device comparisons for the first 30 days after launch versus the previous 30 days. Percentages are rounded for display.
What the results told me
I used the first post-launch period as a behavior check, not a victory lap. The numbers helped separate the parts of the redesign that were gaining traction from the parts that still needed better measurement or refinement.
01
Service findability improved
The large lifts on mental health, childcare, supportive housing, aging, and the main Services page showed that people were reaching content that had been buried in the old structure.
02
Research became part of the main site
Growth on the Research Library and Wilder Research pages showed that the connected experience was being found, not treated as a separate utility.
03
Mobile deserved its own decisions
The increase in mobile and tablet active users supported the choice to rebuild the smaller-screen hierarchy instead of treating responsive work as a final formatting pass.
04
Measurement still needed context
A launch changes URLs, campaigns, search visibility, and behavior at the same time. I used the metrics directionally and kept the next round of tracking focused on completed actions, not just page traffic.
Clearer service paths
People could start with the help they were looking for instead of navigating Wilder’s internal structure.
A usable Research Library
Structured data, filters, result cards, detail pages, and PDFs became one searchable experience inside Wilder.org.
A maintainable system
Templates, content types, ACF fields, Elementor components, and import rules gave staff a safer way to keep the site current.
A measurable foundation
Post-launch analytics made the redesign testable and gave the team a clearer next list than another round of internal opinion.
My contribution
What I actually did on Wilder
This was a full UX, content, platform, and integration project. I connected research, design, development, migration, structured data, and measurement so the public experience and the system behind it could work as one thing.
Stakeholder and user interviews
UX analysis and stakeholder presentation
Analytics, traffic paths, and heatmaps
Content inventory and governance audit
Information architecture and navigation
User journeys and task flows
Figma wireframes and interface design
Responsive and mobile-first design
UX writing and content restructuring
Custom WordPress theme and front-end build
Elementor widgets and ACF field groups
Custom post types and taxonomies
Drupal migration and content cleanup
Presto API connection and staged activation
XML, XPath, WP All Import, and custom PHP
AJAX Research Library search and filters
SEO, accessibility, performance, QA, and analytics
The work in one sentence
I did not just move Wilder from Drupal to WordPress. I rebuilt how people found help, research, and a reason to act.
The finished site connected a clearer public experience to a maintainable content and research system behind it. That meant more than a new visual design. It meant tying interviews, analytics, information architecture, mobile behavior, content modeling, migration, API and XML work, front-end development, and post-launch measurement to the same practical question: what does this person need next?