Hey Period — a brand women would trust with their health
Hey Period lets women track their health through blood tests that use menstrual blood. During R&D the startup needed a waitlist, which meant a brand, a marketing site, and an app people would believe in. I was the only designer.
CLIENTHey Period · femtech startup
ROLEProduct designer · only designer
TIMELINE2 months · 2021
TOOLSFigma
What I owned end to endThe brand identity and the marketing website, from strategy to final assets, marketing copy included.
What I built on others' workThe app UI. Low-fidelity wireframes existed before I joined; I designed the visual system on top and changed several UX decisions (detailed below).
Status & outcomePre-launch Handed off during R&D. [ADD: waitlist numbers, launch status, or client feedback]
HOW TO READ THIS PAGE — artifacts carry disclosure labels: MEASURED real, checkable data · RETROSPECTIVE made after the project to document reasoning · SIMULATED illustrative, no real participants · PROPOSED a goal or plan, never an achieved result.
THE PROBLEM
A blood test that uses period blood is genuinely new, and new ideas in healthcare have a trust problem. The brand had to feel credible enough for medicine and warm enough that women would talk about it.
RESEARCH
User access was limited during R&D, so I leaned on secondary research: teardowns of Flo and Clue, community threads on period tracking, and the team's customer conversations. [ADD: if you ran interviews yourself, how many and how recruited; if not, delete the interview claim for good]
RETROSPECTIVE What the competitor teardown established:
Competitor
What it owns
Trust signal
Implication for Hey Period
Flo
Category default; the pink
Ubiquity: "everyone uses it"
Borrow the familiar pink; differentiate on lab data, which Flo lacks
Clue
Science-forward tracking
Clinical tone, citations
Matching Clue's coldness would erase the warmth women's-health forums say they miss
Mail-in test kits (Everlywell-type)
The mail-a-sample behavior
Medical packaging, results dashboards
Proof the barrier is the unfamiliar sample type, once the kit feels professional
The personas below are assumption personas, pointing a fast-moving team at the same two women. The conditional phrasing ("she would use…") flags assumptions, not interviews.
Persona ABrandy, 21, tracks everything already; Hey Period has to earn a slot next to her Apple Watch.Persona BEmily, 29, has a real medical need and no time; the product replaces a doctor visit, so trust beats fun.
BRAND IDENTITY
The riskiest call was how far to push playful before a blood test stops feeling serious.
The pink deliberately echoes Flo.
Flo's reddish pink is the category's most familiar color. Borrowing it buys instant "period product" recognition; the rest of the palette goes where Flo doesn't.
Supporting colors are whimsical on purpose.
Periwinkle, marigold, mint, and blush do the warmth job. You mail this product your period blood; the packaging has to feel more like a skincare subscription than a lab requisition.
Poppins, for both jobs.
Geometric enough for UI sizes, round enough for display sizes. One family across brand, web, and app kept a two-month timeline realistic.
RETROSPECTIVE Three directions the category offered, and why this one won:
Direction
What it buys
Why it lost / won
Clinical teal (the diagnostics look)
Instant medical credibility
Lost: indistinguishable from telehealth brands; kills the community half of the brief
Millennial beige (the DTC wellness look)
Shelf-pretty, safe
Lost: reads as skincare; buries the real-medicine claim the product depends on
Familiar pink + whimsy (chosen)
Category recognition via the Flo echo, warmth via the supporting palette
Won: it serves both halves of the trust problem
[ADD: if actual mood-board or logo explorations survive, show them here and this table becomes documented evidence instead of reconstruction]
Brand systemPalette, type, lockups, and the kit box, which arrives first and had to sell the brand alone.
MARKETING WEBSITE
The site had one job: waitlist signups from a cold audience for a product that doesn't exist yet. I designed every asset and refined the copy with the Head of Product.
The biomarker map is the credibility centerpiece.
Real clinical names (AMH, TSH, hs-CRP, HPV) grouped by what women care about: fertility, hormones, disease risk. It says "this is real medicine" without a paragraph of disclaimers.
Goal-based sections instead of feature lists.
"Solving your period problems / Optimizing your fertility / Tracking your sexual health", each with a chart card previewing the data. People buy the outcome, not the biomarker.
How-it-works in exactly three steps.
Choose what to track, receive the kit, learn about your health. Skepticism lives in the mail-in step, so it's framed as the easiest.
Full site(Typo in the goals headline corrected from the earlier export.)
APP UI
The team had tested low-fidelity wireframes; my brief was "make it look like the brand." In practice the hi-fi pass changed real UX decisions too:
HomeThe wireframe's "Factors" list became tappable biomarker chips; each score is an entry point. Cycle-day context makes a health score like 67 mean something.InsightsOutlined bubbles became color-coded cards with icons, one insight each, readable on a lock-screen-length visit.AssistantThe wireframe's human coach "Megan" became an explicitly branded bot; a health product shouldn't pretend a person reads your results. Biomarker cards render inline, and quick-reply chips replace free typing.
I restructured the bottom navigation: Home / Discover / Chat / Profile became Home / Insights / Explore / Account. "Insights" is the core promise, so it earned a tab; the assistant is a helper, so "Chat" became a floating entry point.
EARNING TRUST
Which elements carry credibility, and which carry warmth:
Credibility: clinical biomarker names, evidence charts, the Flo-familiar pink, an assistant that is honestly a bot
Untested: whether the balance lands. [ADD: any signal — stakeholder reaction, waitlist conversion, user comments]
PROPOSED How I'd measure it:
Metric
Definition
How
Goal (not a result)
Limitation
Perceived credibility
% describing the brand as "trustworthy/medical" in a 5-second test
5-sec exposure, n≈20, open response
≥60% mention trust or health before "cute"
Perception, not behavior
Waitlist conversion
Visits → signups on the marketing site
Site analytics + form events
Baseline first; direction over absolute
Traffic source dominates early numbers
Kit-return intent
% of signups saying they'd mail a sample this cycle
One-question follow-up email
Establish baseline
Stated intent overstates behavior
CONSTRAINTS & SELF-CRITIQUE
Two months, three surfaces, one designer, and priorities that changed weekly. Two things I'd fix first with more time:
Contrast. The biomarker chips carry white text. I audited the shipped mockups after the fact: every chip fails, and every one passes with dark text.
MEASURED WCAG 2.1 contrast, sampled from the actual screen files (threshold 4.5:1):
Chip (sampled color)
White text
Dark text (#1A1A1A)
Fix
Sleep marigold rgb(247,204,78)
1.53 — fail
11.36 — pass
flip to dark text
Activities mint rgb(105,226,183)
1.60 — fail
10.91 — pass
flip to dark text
Blood coral rgb(253,138,147)
2.27 — fail
7.65 — pass
flip to dark text
E2 periwinkle rgb(140,166,255)
2.33 — fail
7.47 — pass
flip to dark text
Mock data as real content. The screens measure blood volume in mg in one place and ml in another; mock health data deserves the same QA as shipping copy.
STATUS & WHAT I TOOK AWAY
Next I'd validate whether the warm direction survives the skeptical half of the audience.
Brand & sitemine, end to end
App UIvisual system + nav restructure over the team's lo-fi
