Case Study
Culikid uses cooking to build independence for neurodivergent individuals, but their website wasn't reflecting that mission. An end-to-end website redesign and Information Architecture restructure, led as Design Manager across a remote team, to create a site as welcoming as the programs it represents.
Culikid's existing site had fallen behind modern accessibility and usability standards with unintuitive navigation, generic stock imagery, and an unclear mission. Buried navigation made it difficult for families and volunteers to find programs or take action. A lack of authentic imagery and impact data made it hard to build trust, and accessibility gaps excluded the neurodivergent users Culikid exists to serve.
As Design Manager on a Develop for Good project team, I coordinated a remote, globally dispersed team through discovery, IA restructuring, iterative design, and a full Webflow build, while managing client relationships, scope alignment, and project deadlines across every phase.
Research began before any design work. We audited the existing Culikid site for usability and accessibility gaps, reviewed comparable nonprofit sites to benchmark expectations, and aligned with the client on visual direction for a specific color palette, engaging and dynamic navigation, and an updated logo. This upfront discovery agreed with the client upfront on what they defined as success, though that definition was later radjusted with late-stage scope changes.
Due to limited resources from the client in their ability to provide baseline metrics and a user sample, the team did their best to screen and recruit users through the UserTesting platform and their own social networks, as well as performing AI-assisted research to supplement the limited participant pool.
Phase 1 evaluated the low-fidelity prototype using two methods. A 12 participant structured survey measured mission clarity, program findability, trust, and donation confidence. AI usability testing across five representative personas ran in parallel, independently surfacing the same three priorities: stronger trust signals (authentic imagery, impact stats, testimonials), more scannable program details, and clearer CTAs.
Phase 2 ran an AI-assisted A/B test comparing two high-fidelity visual directions, a darker and a lighter color scheme, evaluated by the same five personas. The findings pointed toward a hybrid direction that optimized warmth, readability, and user flows, which became the final design.
01
The original site had navigation labels that were vague and inconsistently grouped, making it difficult for first time visitors to locate programs or understand how to get involved. We considered improving the visual treatment of the existing structure, but a UX audit showed the problem was categorical, not cosmetic. I rebuilt the sitemap from scratch around user goals rather than organizational structure, iterated the header and footer with the client three times, and validated the hierarchy before visual design began.
02
Survey data showed trust and donation confidence scoring 3.5/5, the lowest of any measured dimension, while program discoverability scored 4.3/5. Users could find programs, but didn't feel confident enough to act. We debated prioritizing visual polish in accordance with client requests, but qualitative feedback revealed that users wanted impact statistics, authentic program photography, and real testimonials rather than stock imagery and generic copy. We redesigned the visuals and content around establishing evidence.
03
We could have hardcoded the site content for a faster initial build, but Culikid staff would have needed developer support for every future update. We chose to build the entire site around 7 Webflow CMS collections so that programs, testimonials, donation tiers, blog posts, and partner logos could all be updated independently after handoff. This added complexity during the build phase but handing off a site the client can't maintain would have undermined the value of the redesign.
Wireframes established layout and information hierarchy for key pages like the homepage, programs, get involved, donate, and volunteer, defining content placement and CTA flow before any visual styling began.
Then, a style guide was created by enhancing the client requested color scheme of navy and beige, with the additio n of guidelines for typography, spacing, border radius, etc. The initial logo design was iterated on, eventually culminating in a more refined and simple logo that can be seen in the final design.
Four hi-fi iterations (v1.0 – v1.3) refined the visual system across typography, spacing, component design, and color. Each round incorporated research findings and client feedback, culminating in the hybrid visual direction selected after A/B testing.
%20(1).png)
An interactive Figma prototype captures the final hi-fi design and key user flows across desktop and mobile. The prototype covers the full homepage, programs, get involved, donate, and volunteer flows in the final v1.3 visual direction.
The redesign delivered 23 fully built pages across a restructured IA, 36 CMS templates for ongoing self-management, and resolution of 6+ accessibility issues present in the original site. The new site features modernized visual design with authentic program imagery, intuitive dropdown navigation, and embedded impact statistics and testimonials to build visitor trust.
Google Analytics was implemented site wide to track program registrations, volunteer applications, and donation completions. Post launch metrics are pending.
Status: Design direction and build framework approved by the client. Final desktop/mobile QA and a handful of interaction additions remaining before public launch, with the client agreeing to assume responsibility for those outstanding items.
Scope & ownership: As Design Manager, I shaped the research plan and design direction, contributed screen designs across wireframes and four hi-fi iterations, and managed the coordination that kept a remote, globally distributed team on track. I contributed heavily to the Project Requirements Document, built and maintained the project plan, created meeting agendas, and led both client and internal team communication throughout.
Balancing client vision with research findings: The client came in with a clear visual direction, which gave the team strong guidelines, but it sometimes contradicted what research was surfacing about user needs or established design best practices. Presenting findings as shared evidence, rather than a challenge to the client's instincts, was what moved those conversations forward. A design library built on a standardized system of color, typography, spacing, and auto-layout components also gave us a common framework between designers and engineers, making implementation more efficient.
Trade-offs made: Site development required more engineering resources than expected, so certain interactions and page-level functionality had to be reduced or scoped out. Knowing when to accept a constrained outcome rather than chase perfection was a skill I developed throughout this project. As the session came to a close, I worked with the Technical Manager to produce handoff documentation covering in-scope and out-of-scope items, outstanding issues across all pages, and CMS navigation guidance so that the client could keep building on what the team had started. Despite deviations from the initial requirements, the client approved the design direction and build framework.
Coordination challenges: The most consistent challenge was team communication and working across time zones with teammates who had different skill sets, priorities, communication styles, and availability. I implemented after meeting action item summaries, regular mid-week check-ins, internal meetings amongst the design and engineering teams, and tagging individual contributors on Slack for accountability, which kept the team engaged and more efficient by reducing extended discussion during group meetings.
What I'd do differently: I'd push to establish a more defined and detailed scope and requirements sign-off earlier in the timeline, and provide a synopsis of client design requests so that changing design priorities and and development resource gaps could have been revealed sooner, rather than forcing scope cuts towards the end of the program.
Metrics I'd track: Culikid didn't have baseline analytics on their previous site, so success wasn't measurable from the beginning. If I were scoping this project again, I'd propose tracking:
“Just great. Thank you for everybody who participated in this project and for the great work you did.”
Burhan Halilov, Director of Operations, Culikid
