TMS
TransportManagement
- Logistics
- B2C SaaS
- Enterprise
- Web
- Product Design
- AI Content
Vitaloop
Symptoms In, Direction Out
- Health-tech
- B2C
- Startup
- Web & Mobile
- User Research
- Design System
- Product Design
Vitaloop turns a symptom check-in and lab results into a clear next step — which tests to take, which specialist to see, what to track. Not a diagnosis, a direction.
View Live Project (opens in a new tab)
- Role
- Product Designer
- Researcher
- Team
- Founder
- Product Designer
- Tools
- Figma
- Google Forms
Overview
About Vitaloop
Vitaloop is a health-tech service. Users complete onboarding, fill in a symptom check-in, optionally upload lab results — and get prioritized concerns, test suggestions, supplement recommendations, and a prompt to see a doctor.
The product originally worked only with existing lab results: upload a PDF, get a breakdown. I proposed extending it with a symptom check-in — an entry point for people who don’t have results yet and don’t know where to start.
Problem
People regularly feel unwell, can’t identify the cause, and don’t know which specialist to go to. The team assumed the product’s core value was supplement recommendations. That assumption had never been tested.
Solution
A full validation cycle: quantitative survey → qualitative usability testing → findings split by applicability → a prioritized change plan.
How I Validated It
Measure
Quantitative survey — 19 questions across 5 blocks, 38 responses from the target market. Tested whether the problem exists and where users actually see value.
Observe
1-hour think-aloud usability session. Full path: onboarding → symptom check-in → upload → results. No moderator prompts.
Separate
Split findings by market dependency — interface mechanics that reproduce anywhere vs. conclusions tied to US healthcare economics.
Prioritize
Change plan sorted by impact vs. cost, with product decisions flagged separately from UX fixes.
Deliverables
- Quantitative survey — 19 questions, 5 blocks, 38 responses
- Usability test — 1-hour think-aloud session, unmoderated prompts
- Hypothesis framework — 4 hypotheses defined before testing, resolved after
- User flows for onboarding, symptom check-in, results, and paywall
- Design system and hi-fi screens for web and mobile
- Prioritized change plan — impact vs. cost
Impact
The core product hypothesis was disproved. The team believed it was selling supplement recommendations; target-market data showed users value navigation — which doctor, which test, what to track. Supplements ranked last of four.
- Face the problem regularly
- 76%
- Value supplements — last of four
- 45%
- To first result, behind a paywall
- 15-20 min
Research
I ran a two-step validation: a survey on the target market, then a think-aloud usability test.
01. Survey
Google Forms, 19 questions across 5 blocks: profile → current behavior → problem framing → concept test → interview opt-in. 38 responses from a Ukrainian-speaking audience via Facebook, Telegram, and personal network — matching the product’s target market.
What users value, ranked:
- Which doctor to see — 68%
- Symptom tracking over time — 61%
- Which tests to take — 55%
- Supplement recommendations — 45%
Limits: 38 responses give direction, not statistical significance. The “biohacker” segment couldn’t be recruited — specialized communities are closed or paid.
Ukraine
02. Usability Testing
One participant, ~1 hour, think-aloud, no moderator prompts. Task: sign in, complete onboarding, run a symptom check-in, upload results, read the output.
The participant was based in the US — outside the target market after the pivot to Ukraine. Rather than discard the session, I split the findings by whether they depend on market context.
Findings
03. Hypotheses
Four hypotheses were defined before testing. Two of them did not survive.
| Hypothesis | Method | Status |
|---|---|---|
| The problem exists | Survey, target market | Confirmed: 76% feel unwell without knowing which doctor to see |
| Supplements are the core value | Survey, target market | Disproved: ranked last of four (45%) |
| Biohackers are the target segment | Survey | Not tested: segment could not be recruited |
| The flow completes without friction | Usability test | Disproved: 15-20 min to the first result, then a paywall |
04. Market-independent findings
These are interaction problems. They reproduce on any audience.
Getting to the result
- Value arrives after payment. 15-20 minutes of input, zero output, then a paywall. The main drop-off point.
- Data is requested twice. Fields collected at onboarding are asked again in symptom check.
- Too much choice, no default. The store shows many options; the user wanted one clear pick.
- No personalized dosage. Height and weight are already collected but unused. A safety gap.
Reading the result
- No summary. Nothing to read in 5 seconds.
- Wall of clinical text. No hierarchy between critical and minor.
- Urgency doesn't read. Warning in pale yellow blends in.
- “See a specialist.” No specialty, reads as a brush-off.
- The best part is hidden. Supplements, the only positive reaction in the session, sit at the very bottom.
05. Findings requiring revalidation
Tied to US healthcare economics — flagged, not discarded.
- The competitor is ChatGPT, not other health apps. The participant never once compared Vitaloop to a health service. In Ukraine the economics differ, but with mixed willingness to pay in the survey, free ChatGPT may still be a strong competitor.
- Subscription doesn't match the need. The most common survey answer (about 40%) was to check in only when something feels wrong, not on a schedule.
- Intake-form familiarity. Familiar to a US user from clinic forms; for a Ukrainian user that anchor doesn't exist.
Design
06. Flows, system, screens
I mapped user flows for onboarding, symptom check-in, results, and purchase. The map made the 15-20 minute gap between input and value visible as a structural problem, not a copy problem.
I built the design system and hi-fi screens for web and mobile: components, states, and a clearer hierarchy on the results pages.
Decisions
07. Change plan, prioritized by impact vs. cost
Quick wins
High impact, low cost
- Fix the Complete check-in bug. Nothing else matters until it works.
- Add a summary at the top of results: the problem and the next action.
- Remove duplicate data requests.
Next
High impact, medium cost
- Show a partial result before the paywall.
- One default supplement instead of a catalog, with dosage calculated from data already collected.
- Name the specialty and strengthen the urgency signal.
Product decisions
Not UX fixes
- Rework monetization for an event-driven need: one-off, bundle, or freemium.
- Explore a chat format as an answer to the ChatGPT comparison.
Reflection
08. A disproved hypothesis was the main result
The team was building toward supplement recommendations. The data said users want direction: which doctor, which test, what to track. That is a shift in product understanding, not a visual fix.
There is no “after” metric yet. The test also ran with a user outside the current target market, so part of the findings needs revalidation. I split them instead of presenting everything as equally solid: interface mechanics reproduce anywhere, while conclusions about competition and pricing depend on healthcare economics.
Next: repeat qualitative testing with 3-5 Ukrainian users. The survey’s interview opt-in already gives a recruiting pool.
Work
Next case