A top supplement recommendation is the single-ingredient card that sits at the top of Unfair's ranked feed for a specific goal and a specific user context — one compound, one product, one dose window, with a visible score, a stated confidence, a two-line rationale, and a safety read-out. It is the answer to "if I change one thing today, what should it be," and it is scored against the full user log rather than a population average. When the evidence for a single ingredient is cleaner than the evidence for any multi-compound combination, a top-supplement card outranks every top stack recommendation in the feed.
Why a single-ingredient card exists
Most real progress happens one compound at a time. Adding three things in one week makes any post-hoc attribution guesswork — if it worked, which one worked, and if a side effect appears, which one caused it. A top-supplement card is the app's way of routing users toward single-variable changes when the data supports them. It also serves users who have a foundation in place and want a clear next move rather than a wholesale rebuild — the "add one clean compound" path inside building your first supplement stack.
Anatomy of the card
Every field on the card is loaded deliberately. A recommendation without a rationale and a safety read is a shopping ad; the card below is the full contract.
| Field | What it shows | Example |
|---|---|---|
| Compound and product | Generic name plus a specific recommended product and form | Magnesium bisglycinate, 200 mg, Thorne |
| Goal match | Which user goal the card is scoring against | Sleep quality |
| Score | 0–100 composite from the deterministic scoring layer | 74 |
| Confidence | Low / Medium / High, tied to log density and evidence tier | Medium |
| Expected effect size | Qualitative plus rough timeline | "Small to moderate effect within 1–2 weeks" |
| Rationale snippet | Two-line plain explanation of why this rose to the top | "Matches your sleep-quality goal, your magnesium log shows a 4-day gap, and your trial history shows prior positive response." |
| Dose window and timing | Recommended dose and a specific window | 200 mg, 30–60 min before bed |
| Overlap flags | Detected duplication with products already in the stack | "No overlap detected" or "Already dosing magnesium oxide at 250 mg — consolidate" |
| Risk and contraindication notes | Absolute stoppers plus conditional cautions | "Reduce dose if loose stools appear; caution with severe renal impairment" |
| Primary action | The one button | "Add to stack" |
| Secondary actions | Context-preserving alternatives | "Queue for next cycle," "Compare alternatives," "Not now — why?" |
The primary action is always singular. The card is deliberately shaped to make adding a second top-ranked compound at the same time harder than adding one and reviewing.
How the score is built
The score on the card is a weighted combination of four inputs, each capped so no single dimension can monopolize the ranking.
- Evidence strength. The compound's evidence tier for the chosen goal, capped at 40 points.
- Personal history. Prior dose logs, prior tolerance, and prior subjective-proxy response for this user, capped at 30 points.
- Gap and readiness. Whether the user's foundational supplement stack already covers this role and whether the current cycle is an appropriate time to add, capped at 20 points.
- Logistics. Product availability, interaction checks against the current stack, and cost band, capped at 10 points.
A compound with strong trial evidence but heavy overlap with an existing product can cap out at roughly 70 because the logistics and gap-and-readiness components pull the composite down. The score is explainable by design — tapping the number opens a four-bar breakdown rather than a black box.
Confidence and when to trust the card
Confidence is separate from score on purpose. A high score with low confidence means the compound looks promising for this goal but the user has not logged enough to personalize the ranking. A medium score with high confidence means the compound has been tested against this user's data and is a reliable move even if it is not dramatic. The decision rule is simple — prefer high-confidence medium scores over low-confidence high scores when making a real change.
| Confidence | Log density required | Suggested action |
|---|---|---|
| Low | Under 14 days of cadence | Queue for next cycle, collect more data first |
| Medium | 14–28 days of cadence, no prior trial | Add at recommended dose with a 14-day reassessment |
| High | 28+ days with a prior trial of the same compound or close analog | Add with confidence; reassess on normal cycle |
A numeric example
A user with an eight-week log, a "sleep quality" goal, no current magnesium in the stack, and a prior positive trial of magnesium glycinate receives a top-supplement card for magnesium bisglycinate at 74/100 with Medium-High confidence. The breakdown reads: evidence 32, personal history 26, gap-and-readiness 12 (no current magnesium), logistics 4. The rationale lists the goal match, the prior positive trial, and the absence of overlap. The card proposes 200 mg, 30–60 minutes before bed, with a reassessment in 14 days. That specificity is the difference between a recommendation and a suggestion.
How this appears in Unfair
Top-supplement cards appear at the top of the stack feed when single-ingredient clarity outranks bundle certainty, which is often the case during the first 60 days of logging and during cycle transitions. They appear inline in recommendation ranking alongside stack cards, and the primary action stays a single "Add to stack" button so adding one compound at a time is the path of least resistance.
Clinical safety note
A top-supplement card is a suggestion, not a prescription. If any of the listed cautions apply, if a user is pregnant, pursuing pregnancy, breastfeeding, under active medical treatment, or taking prescription medication with known supplement interactions, the card should be treated as input for a clinician conversation. Adverse symptoms after adding a single compound — rash, palpitations, chest pain, mood collapse, GI bleeding, syncope — are a reason to stop the compound and seek care, not a reason to open the app for a ranking adjustment.