What does the coverage percentage mean?

The share of today's calories that came from foods carrying a micronutrient profile. If it reads 61%, then 39% of what you ate contributed nothing to the totals below — not zero nutrients, but unknown ones. Every micronutrient figure in Base is therefore a floor rather than a measurement, and the banner sits above the numbers so you read it first.

What is the difference between the Build and Baseline tiers?

Build nutrients are the ones where a shortfall costs you training adaptation — magnesium, zinc, iron and the rest. Baseline nutrients are the ones where a shortfall costs you health rather than gains. The screen puts Build first because that is the tier a training app is uniquely positioned to help with.

Why do my micronutrients read zero when I logged a full day?

Because Quick Add entries carry no nutrient profile at all. A day logged mostly through Quick Add produces correct calories and near-zero micronutrient coverage. The seeded food directory carries full profiles, so logging from search rather than by hand is what fills this screen.

Do supplements count toward these totals?

Yes. A supplement's payload flows straight into the micronutrient totals without appearing on the food timeline and without carrying calories. That separation is what keeps magnesium out of the way when you are searching for chicken.

Track Micronutrients and Read Your Coverage

Q: Why does Base track eighteen micronutrients instead of forty?

Each one has to clear one of two bars: being short of it demonstrably costs training adaptation, or being short of it costs health and a meaningful share of adults actually are short. Thiamin, riboflavin, niacin, B6, biotin, pantothenic acid, phosphorus, copper, manganese, chromium and molybdenum are deliberately absent because population inadequacy for them is close to nil in anyone eating enough calories. A row that is always green teaches you nothing and costs a row of attention.

Coverage first

The coverage banner sits above the numbers rather than below them. "Built from 61% of today's calories" is the context that decides whether a shortfall underneath is real or an artefact of how the day was logged.

Read the banner before the rows. It answers one question: how much of what you ate does this screen actually know about.

logged today          2,840 kcal
from profiled foods   1,730 kcal
coverage              1730 ÷ 2840 = 61%

At 61%, a magnesium reading of 210 mg means "at least 210 mg". The remaining 39% of your calories contributed an unknown amount, which could be nothing or could be most of your day's requirement.

This is why the screen never renders a red shortfall as a verdict. A low number with low coverage is a logging artefact; a low number with high coverage is a finding.

Raising coverage

Coverage is a property of how you log, not what you eat.

Logging method	Carries a nutrient profile
Foods from the seeded directory	Yes, full profiles
Barcode products	Sometimes, depending on the entry
Saved meals built from directory foods	Yes, inherited
AI plate scan	No
Quick Add	No

So the fix for low coverage is to log staples from search rather than by hand. A day logged mostly through Quick Add has correct calories and almost no micronutrient signal.

Why eighteen, and not forty

Every nutrient on the screen had to clear one of two bars:

Build — being short of it demonstrably costs you training adaptation. A shortfall here is muscle left on the table.
Baseline — being short of it costs you health, and a meaningful share of adults genuinely are short of it.

Thiamin, riboflavin, niacin, B6, biotin, pantothenic acid, phosphorus, copper, manganese, chromium, molybdenum and chloride are deliberately absent. Not because they do not matter biologically, but because population inadequacy for them is close to nil in anyone eating enough calories with any protein in the diet.

A row that is always green teaches you nothing and costs you a row of attention. Eighteen fits on one screen. Forty is a spreadsheet you stop opening.

The two tiers

Build comes first, because that is where a training app has something specific to say. Baseline second.

Each row expands into its own explanation — what the nutrient does, what the target is, and where the target came from. That reasoning is always reachable and never on screen uninvited.

The weekly view

A single day of micronutrient data is close to meaningless. Intake for most of these is genuinely lumpy — one portion of liver or a handful of Brazil nuts moves a weekly average on its own.

So the screen also offers an average across the days you logged in the week, and it is explicit that a blank day is unknown intake rather than zero. Averaging a blank day in would invent a shortfall that did not happen, which is the same class of error as rendering a missing reading as zero.

Where supplements fit

Supplements contribute to these totals without appearing on the food timeline and without calories. See logging your supplement stack.

That separation is a deliberate data-model decision: a supplement is not a food. Treating it as one would put magnesium in your search results when you are looking for chicken, and would put a 0-calorie entry on your timeline every morning.

Common mistakes

Reading a shortfall without reading the coverage. At 40% coverage almost everything looks short. The banner is above the numbers for this reason.

Chasing a single day. Weekly is the right window for nearly all of these.

Logging staples with Quick Add. It is the fastest way to zero out this screen while your calories look perfect.

Treating the targets as thresholds to exceed. They are adequacy references. More is not better for several of these, and the per-row explainer says so where it applies.

Assuming a saved meal carries complete micronutrients. It sums whatever its components had. One unprofiled ingredient makes the meal's totals a floor too, and Base carries that caveat forward rather than hiding it.

Expecting the AI scanner to fill this in. It estimates macros from a photograph. It has no nutrient profile to contribute.

What to do next

Open the micronutrient screen and read only the coverage figure. If it is under about 70%, the fix is not a supplement — it is logging your three most frequent foods from search once and saving them, so they carry profiles every time thereafter.