Your aisle idea first, because you're right

You said a shopping app should remember what's in each aisle of your stores and in what order you'll walk them. That's exactly right, and it's a bigger deal than I'd scoped. What we shipped is a fixed keyword dictionary in one global order. It groups produce together, but it has no idea that at your Walmart produce is the first thing you hit and dairy is the far back wall. A list sorted in an order that doesn't match the store you're standing in is barely better than unsorted.

The design: multiple stores per user, each with its own ordered aisle list (the sequence you actually walk). Setup has to be nearly free or nobody does it: a new store seeds from our default aisles, then you drag them into your order once. Pick your store at the top of the list and it re-sorts instantly. And when you move an item to a different aisle at that store, we remember it forever — that learning is exactly what makes AnyList's version good and ours (currently) not.

Now the trackers: what they actually ship

Logging (where they compete hardest)

Barcode scanning, AI photo logging (snap the plate, 90%+ claimed accuracy), voice and natural-text logging ("I had two eggs and toast"), manual entry, custom foods, a recipe importer (URL or paste → parsed ingredients), custom recipes with per-serving math, a meal diary by time of day, copy-yesterday and quick-add favorites, and restaurant menus.

The food database (their real moat, and the fault line)

Two philosophies, and the split matters:

• MyFitnessPal: 20 million+ items, crowdsourced, unverified. Enormous coverage, conflicting entries for the same food, constant accuracy complaints.
• Cronometer: USDA + NCCDB, lab-verified, 84 nutrients, staff-verified submissions. Smaller but far more accurate — and that accuracy is its entire market position.

Targets, extras, and insight

Targets: calories from age/weight/activity/goal, macros in grams or percent, micronutrient targets vs RDA (Cronometer's signature), adaptive targets (MacroFactor reads your weigh-ins and moves your goals automatically — it's why it charges the most), and protein-forward presets.

Beyond food: weight and measurements with trend smoothing, water, exercise and steps from wearables, intermittent fasting timers, CGM/glucose, and sleep-vs-intake correlation.

Insight: daily nutrition analysis, the famous micronutrient-vs-RDA report, weekly trends, AI pattern-spotting ("your energy dips when you skip breakfast"), and cycle-phase awareness (more protein in the luteal phase).

Integrations: Apple Health, Google Fit, Fitbit, Garmin, Withings, Oura, plus CGMs.

What we already have (more than you'd think)

We have the hard half (accurate nutrition on real recipes) and almost none of the easy half (logging, goals, trends).

What they charge (useful for your pricing decision)

Trackers charge $40 to $80 a year for nutrition data alone. Our fake door is $48 to $96/yr. We are not underpriced for this neighborhood.

The call: do NOT become MyFitnessPal

Straight talk, because this is the important part:

1. Their moat is the packaged-food database, and we can't cross it. 20M crowdsourced items, every cereal box, every restaurant. That took MyFitnessPal fifteen years and an Under Armour acquisition.
2. It's off-wedge. We sell a beautiful, permanent, shareable family cookbook. "Log everything you put in your mouth" is a different emotional product for a different person. Your Family Archivist is not shopping for MyFitnessPal.
3. Tracking retention is brutal. Most people quit in two weeks. Don't build our stickiness on a behavior that reliably fails.

The sweet spot: track what you cook, not everything you eat

And here's the small, lovely part: we already log that you cooked something (with dates), and we already know its full nutrition. So "what did my cooking actually give me this week, protein-wise?" is nearly free. It's a join we simply aren't doing yet. No new logging burden, no calorie policing.

That lands squarely on your own audiences: protein-first, menopause and HRT, APOE4-aware, SmartStrongAlive. Those readers care about protein and micronutrient density in real meals, not about logging a granola bar.

What I'd take

The honest summary

We accidentally built the expensive half of a nutrition tracker. Cronometer's whole market position is "USDA-accurate, micronutrients, verified," and we have USDA-derived macros and 10 micros on your own 1,500 recipes — something Cronometer literally cannot produce without hours of hand entry.

But the tracker business is a database-coverage business and a daily-habit business, and we shouldn't enter it. Take the five nearly-free Tier 1 features that turn nutrition we already computed into insight, keep logging recipe-shaped, and consider writing to Apple Health so we complement her tracker instead of fighting it.