Driven Decisions Made Designs

The first decision was to rebuild everything from scratch. The web product's visual language was clinical and cold, professional in the wrong direction for an audience navigating one of the most emotionally charged experiences of their lives. We replaced it entirely.

A Design System Built for Warmth

The new design system was built around pastels, rounded edges, and illustration. Hard lines and navy gave way to soft backgrounds, blob shapes, and custom illustrations mixed with modified open-source assets. The logo was redesigned. Every corner was rounded. The visual language was deliberately warm, approachable, and human, the opposite of medical.

This wasn't aesthetic preference. It came directly from research. Mothers described the web product as cold and intimidating. The redesign was a response to that feedback, not a stylistic choice made in isolation.

Designing Around the Baby

One of the most validated features was also one we initially underestimated. Seeing the baby week by week, your baby is now the size of an avocado, tested as a nice-to-have in early research. In practice it became one of the most emotionally resonant moments in the product.

We didn't just add it. We made it central and interactive. As the baby developed week by week mothers could explore exactly what was forming, see a 3D representation of the baby, understand what symptoms to expect, and track developmental milestones. It was front and center on the home screen. Mothers shared it with partners, with older children, with family. It became a moment of connection the product was built around rather than a feature tucked into a corner.

Workflows That Teach Without Feeling Like Homework

The birth plan, hospital bag, postpartum plan, breastfeeding guide, and sleep training workflows were all built around the same principle, interactive and guided, never passive and informational.

The birth plan walked mothers through every decision they would face in the delivery room: comfort preferences, epidural choices, who they wanted present, whether they wanted music, lighting, gender reveals, vaccination preferences, cord clamping. Each item was presented one at a time, completable with a partner, and exportable as a PDF that could be shared directly with their doctor.

The hospital bag workflow broke packing into categories, for the baby, for the mother, for the partner, with checklists that could be checked off directly in the app with reminders as the due date approached.

Postpartum planning went further. Who would cook the first few days. Who would clean. Breastfeeding guidance with posture instruction and formula alternatives for mothers who chose not to breastfeed.

What made these workflows different from a checklist or an article was pacing. Items were presented progressively, one decision at a time, in a way that felt low pressure and completable. Mothers finished workflows in a single sitting, something they consistently said they hadn't expected to do.

Solving for the Unknown Unknowns

The resource center addressed the deeper problem research had surfaced, mothers not knowing what they didn't know.

We surveyed mothers about what they wished they had known postpartum and where they wished they had encountered that information. We interviewed doulas, OBGYNs, and other medical professionals to map the gaps between what mothers knew and what they needed. That research became the content architecture.

The resource center had two modes. A personalized homepage surfaced articles relevant to each mother's specific stage, conditions, and history, content that found her rather than waiting to be searched for. A fully searchable library allowed exploration by topic, stage, or keyword for mothers who knew what they were looking for.

Personalization went deeper than stage. Onboarding assessments captured prior C-sections, existing medical conditions, known fetal conditions, and other factors that shaped what information was surfaced and when. Those assessments could be updated throughout the journey as circumstances changed.

The Design I Defended

One workflow generated significant internal pushback, a health habits feature built around a Tinder-style swipe interaction. Mothers were presented with health habits one at a time and swiped right if they felt the habit was positive or left if they were unsure. The swipe didn't score or judge. It surfaced information and resources related to each habit and ended with an overview of everything covered.

The pushback was predictable. It felt experimental. It felt like a game in a product meant to be taken seriously.

It tested better than almost anything else we built.

Mothers found it intuitive immediately. Swiping was already a native behavior. It was fast. It felt low pressure. It didn't feel like a test. The information landed because they were engaged rather than reading passively. They could move at their own pace and return to it later without losing progress.

I defended it when it was questioned internally and I would defend it again. The research was clear. The interaction pattern served the content. The fact that it felt like a game was exactly why it worked.

Continuity by Design

One consistent finding across research was that mothers started things and didn't finish them. Not from lack of interest, from interruption. Life with a pregnancy, especially a first, doesn't allow for uninterrupted sessions.

Every element of the product was designed around this reality. Unfinished tasks surfaced directly on the homepage so mothers could see exactly where they left off. Workflows saved state at every step. Progress indicators were persistent and visible. The product never required a mother to start over.