What are Health Scores and Score Factors

A step count tells you how many steps someone took. It doesn’t tell you whether they sat for 11 hours between two short walks. A sleep duration of 8 hours sounds healthy — until you learn it happened 3 hours later than usual after a week of mounting sleep debt. Single metrics are easy to collect but often misleading in isolation.

Health scores synthesize multiple related metrics into a single measure of wellness for a specific dimension of health — activity, sleep, readiness, wellbeing, or mental wellbeing. Each score is a value from 0 to 100, but what makes it actionable isn’t the number — it’s the factors underneath it. Every score breaks down into independently measured and scored factors, so you always know why the score is what it is and which specific area to focus on.

Scores are not clinical. They don’t diagnose conditions or replace medical assessment. They measure daily wellness patterns — the kind of everyday behaviors that compound over time and shape how a person feels, performs, and recovers. The goal is to make those patterns visible and actionable.

How Scores Work

Each score has two layers: the score itself and its contributing factors.

The score is the top-level output — a value from 0 to 100 with a state label that translates the number into plain language:

Factors are the individual health dimensions that feed into the score. Each factor returns a value (what was measured), a sub-score (0 to 100), a state, and a goal (a static, research-backed reference target). The sub-score is the most useful piece for a product — by comparing factor sub-scores, you can immediately see which dimension is weakest and suggest the right habit to build.

Here’s a concrete example. The Activity Score has six factors, all available from phone data alone:

The overall score is 65 (medium) — despite 8,200 steps. The factor breakdown reveals why: extended inactivity is minimal at 35 (9 hours of sitting against a 4-hour goal) and active hours is low at 45 (only 4 of a target 12 hours had any movement). This person walks to work and back but sits all day in between. A product can surface this as: “You’re getting plenty of steps, but they’re packed into a short window. You sat for 9 hours today. Try a short walk every couple of hours — distributing movement matters as much as total volume.”

That’s the difference between showing a number and guiding a habit. The person doesn’t need to move more — they need to move differently. The factor breakdown surfaces that distinction. Goals give context (8,200 steps against a target of 10,000), but the sub-score is what drives the recommendation — it tells you which factor to focus on and, by extension, which small daily habit to suggest.

Available Scores

Activity Score

Measures daily movement across 6 factors: steps, active hours, extended inactivity, active calories, high-intensity activity duration, and floors climbed. All 6 work from phone data alone.

What makes Activity useful is that it captures the pattern of movement, not just the volume. Someone who runs 5k in the morning but sits for 12 hours afterward may score lower than someone who walks regularly and takes the stairs throughout the day. Research consistently shows that breaking up prolonged inactivity matters as much as total exercise — and the factor breakdown makes that visible.

Sleep Score

Measures sleep health across 7 factors: duration, regularity, continuity, sleep debt, circadian alignment, physical recovery (deep sleep), and mental recovery (REM sleep). 4 factors work from phone data; the remaining 3 (sleep stages and recovery) require wearable data.

Sleep regularity — going to bed and waking at consistent times — is one of the strongest predictors of sleep health, sometimes more impactful than duration itself. The Sleep Score captures this alongside accumulated sleep debt, which tracks whether a user has been under-sleeping across multiple nights. A single good night doesn’t erase a week of deficit, and the score reflects that.

Readiness Score

Measures recovery and preparedness across 8 factors, comparing today’s data against a shifting 30-day baseline that becomes more tuned to the individual over time. 4 factors work from phone data; wearable data adds cardiovascular signals like HRV and resting heart rate.

Readiness factors in accumulated strain — not just last night’s recovery, but a week of poor sleep or several consecutive days of high exertion. This is why it can feel decoupled from a single good night’s sleep; it’s designed to flag hidden fatigue that hasn’t cleared the system yet. Think of it less as a performance report and more as a personal weather forecast — it tells you whether to push harder or play it safe before you feel the crash.

Wellbeing Score

The broadest score — 13 factors spanning both activity and sleep, producing a single daily wellness measure. 10 of 13 factors work from phone data.

Wellbeing is useful when a product needs one number for “how is this person doing overall” without asking the user to interpret multiple separate scores. Because it draws on factors from both movement and sleep, it catches patterns that neither score would flag alone — a user who sleeps well but is completely sedentary, or one who moves plenty but has erratic, poor-quality sleep. The combined view reflects how these dimensions actually interact: poor sleep reduces daytime energy, reduced activity disrupts the next night’s sleep, and the cycle compounds.

Mental Wellbeing Score

Measures behavioral patterns linked to mental health across 6 factors — all from phone data alone, no wearable required.

Mental Wellbeing doesn’t measure mood and doesn’t diagnose conditions. Instead, it tracks the behavioral patterns that research consistently links to mental wellbeing — things like daily routine consistency, sleep timing regularity, and activity patterns. These behavioral markers are observable, objective, and modifiable. When they shift — routines become irregular, sleep timing drifts, activity drops — it often precedes or accompanies changes in mental wellbeing. A product can surface these patterns without ever asking the user “how are you feeling?”

Some factors appear in multiple scores — sleep duration contributes to Sleep, Wellbeing, and Readiness — but the weight and context differ based on what each score is measuring. Every score works with smartphone data alone. Wearable data adds additional factors (particularly sleep stages, heart rate, and HRV) but is never required for a meaningful result.

Key Features

Real-time updates. Scores recalculate within one minute of new data arriving. When a user finishes a sleep session or completes a workout, their scores reflect it almost immediately — not at the end of the day or on a batch schedule.

Retroactive generation. When a user first connects, scores generate retroactively for the previous 14 days. This matters because scores like Readiness rely on a baseline built from recent history — without backfill, the first two weeks would produce incomplete or unreliable results. Retroactive generation eliminates the cold-start gap entirely.

Phone-first design. Every score produces meaningful results from smartphone data alone. Two scores — Activity and Mental Wellbeing — have 100% factor coverage from phone data. The remaining scores gain additional factors from wearable data (sleep stages, heart rate, HRV), but no score requires hardware beyond the phone the user already carries.

Explainable by design. Every score breaks down into independently scored factors. There are no hidden weights or black-box aggregations — the factor breakdown is the explanation. This transparency matters for user trust: people are more likely to act on a recommendation when they can see exactly what drove it.

Graceful handling of missing data. Not every factor is always available — a user without a wearable won’t have sleep stage data, and some days may have gaps in tracking. When a factor can’t be calculated, it returns null and the score adjusts around the remaining factors. Each score defines a minimum factor threshold below which the score itself returns null rather than producing a misleading result.

Research-backed methodology. Every factor in every score is grounded in peer-reviewed research. The thresholds, goals, and factor selections aren’t arbitrary — they reflect established health science. The science guides for each score detail the specific evidence behind every contributing factor — see the Further Reading section below.

When to Use Scores

Scores are the right choice when your product needs to tell a user how they’re doing in a specific health dimension — and why. They’re designed for features like daily health summaries, coaching nudges, progress tracking, and any experience where an explainable, actionable health measure is more useful than raw data.

If you need something different, other Sahha products may be a better fit:

Use Biomarkers when you need raw data. Biomarkers are individual measurements — step count, sleep duration, resting heart rate, active calories. They answer narrow questions: how many steps today? Scores synthesize multiple biomarkers into a broader measure: how healthy is this person’s overall activity pattern? If your feature needs a specific data point rather than an interpreted health measure, use biomarkers directly.

Use Insights when you need engagement signals. Insights surface meaningful moments — trends (e.g., sleep quality improving over two weeks) and comparisons (e.g., activity higher than usual baseline). They’re designed to drive user engagement by highlighting what’s changed and what’s worth paying attention to. Scores measure the current state; Insights highlight what’s noteworthy about how that state is evolving.

Use Archetypes when you need to personalize recommendations. Archetypes classify users by sustained behavioral patterns — types like “early riser” or “cardio oriented.” They’re useful for tailoring product experiences, content, and recommendations based on who a user is, not just how they’re doing today. Archetypes include an ordinality parameter so your product can detect when a user’s behavioral pattern shifts over time.

Further Reading

Score Guides

Each score has its own guide covering how it works, what factors drive it, and how to use it in your product:

• Activity Score Explained
• Sleep Score Explained
• Readiness Score Explained
• Wellbeing Score Explained
• Mental Wellbeing Score Explained

Science Guides

Each score has a companion science guide detailing the peer-reviewed evidence behind every contributing factor:

• The Science Behind the Activity Score
• The Science Behind the Sleep Score
• The Science Behind the Readiness Score
• The Science Behind the Wellbeing Score
• The Science Behind the Mental Wellbeing Score

Documentation

For technical integration — API reference, schemas, webhooks, permissions, and SDK setup:

• Scores Overview — Docs