Frequently Asked Questions

Find answers to common questions about Eatomate's features, billing, privacy, and more.

📱
Product & Features

How do I get help or support?

Look for the chat bubble in the bottom right corner of any page! Our support team is available through live chat to help you with any questions. You can also email us at support@eatomate.co.uk or visit our Support Center. The chat bubble is always there—just click it to start a conversation with our team.

How do I invite family members to my household?

Go to Settings → Household → "Invite Member". Enter their email address and they'll receive a magic link to join your household. They'll be able to track their own meals while sharing the same pantry and subscription. Each additional member costs £9.99/month. All household members can see shared meals and pantry inventory, but personal meals remain private.

Two people in my household signed up separately—how do we merge accounts?

This is a common situation! Go to Settings → Household → "Advanced Household Management" → "Merge Households". Enter the other founder's email address, and the system will check if the merge is possible. Both household founders will need to consent to the merge. Your accounts, pantry data, and meal history will be combined into one household. The source household (the one being absorbed) must be on Trial, Trial Expired, or have its subscription cancelled. If the source has an active paid subscription, cancel it first from Settings → Billing. The target household (the one you're merging into) must be in Trial or Active status.

Do all members of a household need an Eatomate account?

Yes, every household member must have their own Eatomate account under the same household. This is a fundamental requirement, not a business decision. Eatomate achieves 93% weekly accuracy through pantry reconciliation: when a pantry item is consumed, the system needs to know exactly who ate how much from the shared pantry. If even one person is eating from the pantry without tracking, the mass conservation equation breaks (consumed + remaining + waste must equal purchased) and accuracy drops for everyone. Every member needs to log their meals so the system can attribute consumption correctly and reconcile against the shared pantry inventory.

How do I leave a household?

Any non-founder member can leave a household at any time. Go to Dashboard → Household and click "Leave Household". You'll be moved to a new solo household with your personal meal history and calibration data. The shared pantry stays with the original household. If the original household is still on a free trial, your remaining trial period carries over to your new household. If the original household had a paid subscription, you'll need your own subscription to continue. If multiple members want to leave together, each person leaves individually and then uses the Merge Households feature to combine their new solo households into one. Note: Founders cannot leave directly—transfer the founder role to another member first, then leave.

Can children use Eatomate?

Yes! Children can absolutely use Eatomate through our helper relationship feature. Parents or guardians can track meals and nutrition on behalf of their children. Here's how it works: Add your child as a household member in the app, set up a parent-child helper relationship, and when scanning meals, use the account selector to choose who the meal is for. All meals are recorded to your child's account, giving them age-appropriate nutrition tracking with parental oversight. This is especially useful for children under 13 who may not have their own smartphone, or for parents who want to maintain visibility into their child's nutrition.

What if a household member doesn't have a smartphone?

Eatomate supports helper relationships for exactly this scenario! Any household member with the app can track meals on behalf of someone without a smartphone. Common use cases include: elderly parents (adult children can track meals for parents who don't use smartphones), young children (parents can track nutrition for kids without devices), caregivers (track meals for anyone you're caring for), and spouses (help each other with meal tracking). When you scan a barcode or log a meal, simply use the account selector dropdown to choose which household member the meal is for. All nutrition data remains private to the user, even when tracked by a helper. To set this up, add the person as a household member and establish a helper relationship.

Does Eatomate require you to take meal photos?

No. Eatomate does not require meal photos. We use a simple, accurate approach: weigh your meal on a kitchen scale (or scan the barcode for packaged items). That's it. No photos, no visual estimation, no AI guessing from images. Why? Because photos are fundamentally unreliable for nutrition tracking—camera angles, lighting, and plate size create optical illusions that make portion estimation wildly inaccurate (±30-50% error). A kitchen scale gives you ground truth in grams (±0.5% precision). Mass is what your body processes, not volume. We achieve 93% accuracy by measuring actual grams consumed, not by analyzing pixels.

How does Eatomate track my meals?

Week 1-2: Weigh each meal type 3 times so the system learns your typical portion and how much it naturally varies. Week 3+: The app predicts your top 5 likely meals at each meal time. 90% of the time your meal is in that list—just tap it (1 second). The calibrated portion is automatically applied. 9% of meals need a quick search (10 seconds), 1% need weighing (novel items, 30 seconds). Average: 2 seconds per meal. Physics-based reconciliation then corrects estimates to ground truth (~93% accuracy by Week 4) using mass conservation when pantry items expire.

Do I weigh my whole plate or individual items?

Weigh each recipe separately, even if you eat them together. For example:

  • Thai curry → weigh separately (300g)
  • Rice → weigh separately (150g)
  • Coke → barcode scan (no weighing needed)
This matters because the system tracks ingredients at the recipe level. When your coconut milk expires, mass conservation corrects your curry portions—not your rice portions.

When you save, the app generates a memorable name for the entire meal (e.g., "Weeknight Thai Feast"). Quick Log uses these meal names to predict your full plate combinations—tap once to log all items instantly.

How accurate is the nutrition tracking?

Weekly nutrition reports achieve 93%+ accuracy (2σ) after just 4 weeks of using Eatomate. This is post-reconciliation accuracy — last week's data after physics has run. The system starts at ~82.5% in Week 1 and improves as reconciliation accumulates pantry data. Post-reconciliation, portion error collapses to zero by mass conservation. The only remaining error is biological CV — natural variation in caloric density per gram — which is quantified per ingredient in our whitepaper. All measurements track weight in grams with statistical confidence intervals.

How does Eatomate achieve 93%+ accuracy? What are your accuracy assumptions?

Eatomate eliminates two of the three sources of caloric tracking error (portion weight and product identity) through barcode scanning and kitchen scales. The only remaining error is caloric density variation per gram — an irreducible biological fact. Our whitepaper quantifies this for every common food type.Single-ingredient barcoded items (biological CV only):• Manufactured & processed: ±1–8% at 2σ (sugar, oil, crisps, cereal, peanut butter)• Dairy, grains, legumes & eggs: ±5–12% at 2σ (milk, rice, oats, cheese, lentils, eggs)• Meat & fish (barcoded supermarket): ±8–18% at 2σ (tofu, beef mince, chicken breast, salmon, beef steak/lamb)• Fresh produce (USDA-matched): ±17–25% at 2σ (apple, banana, broccoli, avocado)Multi-ingredient barcoded products (ready meals):• UK/EU: ±15% at 2σ (FSA compliance midpoint; legal tolerance ±20%) • US/China: ±8% at 2σ post 1.1× correction (legal tolerance 100–120%). Not reconcilable — consumed directly.Restaurant meals:• Independent restaurant (unweighed): ±67.5% at 2σ (Urban et al., 2016, JAND — within-cuisine residual variance after correction)• Restaurant weighed (no barcode): ±62% at 2σ (removes portion error; residual is caloric density variation)• Chain restaurant (calorie-labelled): ±15% at 2σ (McDonald's, Wagamama, Pret — standardised recipes)How these combine to 93%+:Post-reconciliation, portion error collapses to zero by mass conservation. The remaining biological CVs combine via quadrature (root-sum-of-squares) across 7 independent food groups. Across three diet profiles (meat-heavy, balanced, plant-heavy) at 14,000 kcal/week, the whitepaper shows 93.0–93.5% weekly accuracy for home-only cooking (Scenario A) and 93.2–93.5% for modern households with ≤18% ready meals and ≤10% restaurant (Scenario B).What about the first two weeks?During calibration, expect around 82.5% accuracy as the system learns your recipes, portions, and cooking patterns. Accuracy improves to 93%+ once reconciliation has enough pantry data.What if I eat out a lot?Keep restaurant meals to ≤2 per week to maintain 93%+ accuracy. Restaurant-heavy diets (55%+ non-home calories, Scenario C) fall to ~91% — still significantly better than competitors but below the 93% threshold.Is 93% a target or a guarantee?It's a mathematical floor based on conservative assumptions. The whitepaper's worst-case profile (meat-heavy, Scenario A) achieves 93.0%. See our full whitepaper for the complete error propagation calculation.

What does the accuracy percentage in my dashboard represent?

The accuracy percentage shows how accurate your total weekly nutrition intake is, not individual meals. Post-reconciliation, portion error is eliminated by mass conservation — the only remaining error is biological CV (natural variation in caloric density per gram). Independent errors from different food groups partially cancel via quadrature (root-sum-of-squares) across 7 food groups. Restaurant meals cannot be reconciled and carry higher error (±30-35% at 2σ). The system aggregates over 1 week to give you an honest, physics-verified accuracy figure — typically 93%+ for home-cook and modern household diets.

Why are my barcoded meals showing higher calories than the package label?

Multi-ingredient barcode products (ready meals, protein bars, frozen dinners, etc.) may show corrected calorie values depending on your region. US & China: FDA regulation (21 CFR 101.9) allows manufacturers to understate calories by up to 20%. Eatomate applies a 1.1× correction factor (midpoint of the 0–20% tolerance) to remove this systematic understatement. A ready meal labelled 400 kcal becomes 440 kcal. UK & EU: Regulation 1169/2011 allows ±20% bidirectional tolerance with no systematic bias, so no multiplier correction is applied. Important: Single-ingredient products (chicken breast, rice, milk, etc.) are not corrected because their caloric uncertainty comes from biological variation within the SKU, not label compliance. The correction applies only to multi-ingredient products where the recipe composition is opaque.

What makes Eatomate different from other nutrition apps?

Traditional apps require 15-23 minutes per day of manual logging (Harvey et al. 2019, Obesity)—that's 2+ hours/week. Eatomate achieves 2 seconds per meal average (2 min/week total)—that's 60× faster. How? Week 1-2: Calibrate each meal type with 3 weigh-ins to learn your typical portions. Week 3+: Single-tap from predicted top 5 list (90% of meals). Physics-based reconciliation uses mass conservation (consumed = purchased - remaining - waste) to learn your recipes (not generic) and correct estimates to ground truth. Achieves ~93% accuracy by Week 4. Other apps guess portions; Eatomate measures ground truth.

Do I need to scan my groceries?

Yes, for best accuracy. You can scan grocery receipts or scan individual barcodes to track your pantry. Your first receipt requires manual barcode matching (~100% initially, but as our user base grows, network effects will reduce this to ~10% for new users). Week 4: 95% auto-matched as intelligent caching learns your specific shopping patterns (fuzzy trie with OCR-aware edit distance). Week 8: 100% auto-matched—zero manual scans needed for your regular groceries.

How do I add non-barcoded items efficiently?

For items without barcodes, search for the item by name in our database and select the best match from the results. Then weigh the item on your kitchen scale and enter the weight in grams. The system matches against 50K+ recipes and 100K+ alternative names to find accurate nutrition data. This works for fresh produce, bulk items, and anything else without a barcode.

What's the difference between 'Setup Pantry', 'Scan Receipt', and 'Log Meal'?

These are three separate features in Eatomate:

  • Setup Pantry: A one-time onboarding flow where you scan your existing pantry staples (rice, pasta, oils, spices, etc.) so the system knows what you already have. Only done once during setup.
  • Scan Receipt: Used every time you shop—photograph your grocery receipt and the system extracts line items automatically to add purchased items to your pantry.
  • Log Meal: Track what you eat—search for the meal, weigh it on a scale (or use calibrated portions after 3 weigh-ins), and log it to track consumption.

Think of it as: Setup Pantry (once) → Scan Receipt (after shopping) → Log Meal (when eating).

What happens if I scan the same receipt twice by mistake?

Eatomate automatically detects duplicate receipts by comparing store name, receipt number, purchase date, total amount, and line items. If you try to scan a receipt that may have already been scanned, the system will show a warning popup asking "This receipt may have already been scanned. Do you want to scan again?" You can choose to proceed if it's actually a different receipt, or cancel to avoid adding duplicate items to your pantry. This prevents accidental double-counting of groceries and maintains accurate pantry inventory.

Can I track restaurant meals?

Yes! When adding a meal, uncheck the 'from your home pantry?' checkbox (located at the top of the meal entry screen), then search for the dish in our database of 50K+ recipes and restaurant dishes. If you weigh it (takeaway container on scale), accuracy is ±62% at 2σ. If you don't weigh it, accuracy is ±67.5% at 2σ (Urban et al., 2016, <em>JAND</em>). Chain restaurants with published calorie data (McDonald's, Wagamama, Pret) achieve ±15% at 2σ. Restaurant meals cannot be reconciled (no pantry tracking), so keep them to ≤2 per week for 93%+ weekly accuracy.

How many restaurant meals can I eat and still get 93%+ weekly accuracy?

To maintain 93%+ weekly accuracy (2σ), keep restaurant meals to ≤2 per week. Our whitepaper models three scenarios: Scenario A (home-only) achieves 93.0-93.5%, Scenario B (modern household: ≤18% ready meals, ≤10% restaurant) achieves 93.2-93.5%, and Scenario C (restaurant-heavy: 55% non-home calories) falls to ~91%. Chain restaurants with published calorie data (±15% at 2σ) have less impact than independent restaurants (±67.5% at 2σ).

Why does Eatomate focus on weekly accuracy instead of per-meal accuracy?

Because that's how humans actually make diet decisions. Nobody adjusts their next meal based on a single calorie reading — real dietary decisions happen in weekly blocks. 'I've eaten well this week, I can enjoy that dinner out' or 'I overate Wednesday, I'll be lighter Thursday and Friday.' Per-meal accuracy is also statistically misleading. A restaurant meal might have ±67.5% uncertainty — not because our system is imprecise, but because that uncertainty is real and no app can eliminate it. What matters is that over a full week, independent errors from different food groups combine via quadrature (root-sum-of-squares), and post-reconciliation portion error collapses to zero by mass conservation. Your weekly total converges to 93%+ accuracy. Other apps show you 487.3 calories per meal with false precision. We show you honest uncertainty ranges that converge to research-grade weekly accuracy — because that's the number that actually drives results.

Does this work for meal prep?

Yes! The meal planner suggests mathematically-optimized meals based on your pantry inventory and preferences. It prioritizes ingredients expiring soon to minimize waste. For batch-cooked meals, you log each portion when eaten by searching and entering the weight. Reconciliation automatically validates that all logged portions sum correctly to the actual pantry depletion from when you cooked the batch.

What happens if I don't have a pantry item scanned?

The system will still match your search to our database of 50K+ recipes and common ingredients, but accuracy may be lower (~75% vs 93% weekly) without pantry reconciliation. You can manually add items or the system will prompt you when it detects missing ingredients during reconciliation.

Can I use Eatomate without scanning receipts?

Yes, but accuracy will be significantly lower (~65-75% vs 93% weekly). The reconciliation engine requires pantry data to work. You can manually add items, but receipt and barcode scanning is the fastest way to build your pantry.

What if I eat at someone else's house?

When adding the meal, uncheck the 'from your home pantry?' checkbox (located at the top of the meal entry screen), then search for the meal in our database. If you can weigh it (e.g., takeaway container on a scale), accuracy is ±62% at 2σ. If you can't weigh it, accuracy is ±67.5% at 2σ. Lower than home cooking (93%+ via pantry reconciliation) since the system doesn't have access to their pantry, but still useful for tracking.

Does this work for snacks and drinks?

Yes! Scan snack barcodes when you buy them. For individual items without barcodes (fresh fruit, coffee, etc.), search for them in the database, weigh, and enter the weight. The system tracks everything from full meals to single items.

How does reconciliation work with shared meals?

Week 1-2 (Calibration): Each person weighs their own portion 3 times so the system learns their individual portion sizes. Week 3+: Each person single-taps their meal—the system auto-applies their calibrated portion (e.g., Dad: 350g curry, Mom: 250g curry, Kid: 150g curry). Reconciliation ensures all portions sum to the actual total consumed through pantry depletion tracking.

What does "Improve accuracy" mean in my pantry?

Staples (rice, dal, flour, pasta, oils, eggs, canned goods, etc.) don't expire quickly, so they need periodic rescanning to maintain accuracy. The system prompts you to rescan staples after 7+ days if they've been consumed by >20% and have >7 days until expiry. This quick rescan updates the remaining quantity and improves accuracy for meal predictions. You'll see an unobtrusive "Improve accuracy" banner showing how many staples are ready to rescan—tap it to weigh them again. This takes seconds and unlocks easy accuracy wins without the hassle of tracking every meal ingredient.

What happens if I throw food away or it goes bad?

Most food waste is handled automatically. The system knows from its food taxonomy that apple cores aren't eaten, chicken breast has a trim factor, and lentils have near-zero waste — this is applied silently without any input from you. For pantry ingredients that go bad (e.g., milk expiring), each item has a simple per-item waste slider to log what was discarded — this takes two seconds and happens a few times a year per household. For cooked meals or leftovers being thrown out, you log the wasted item just like a regular meal but weigh the portion being discarded. This keeps the mass balance accurate so reconciliation can do its job. Small amounts of unlogged waste will not significantly impact your weekly accuracy, but consistent unlogged waste will degrade reconciliation over time. Over time, if you regularly throw out leftovers, the system learns your typical leftover waste percentages per meal type — so even waste tracking gets easier.

How quickly does receipt scanning reach 99% auto-match?

Most households buy the same 50-80 items regularly. Each time you or any user confirms a receipt mapping (e.g., "JS S/SKIM MLK" → Tesco Semi-Skimmed Milk 2L), that mapping is saved permanently to our shared database and benefits everyone. Two forces work together: your own shopping repetition means most of your items get mapped within a few weekly shops, and network effects mean other users are constantly mapping items you haven't seen yet. By month 2-3, virtually every item on your receipt auto-matches without any input — typically reaching 99%+ auto-match.

How accurate is body composition tracking?

Eatomate decomposes your weight changes into fat, muscle, glycogen, and water using a three-tier Bayesian fusion model. Energy balance (your physics-verified intake at ±5%) is the primary constraint — fat change is derived as the energy residual, not from the Navy tape measurement. Exercise-type priors from the research literature estimate muscle change, and the Navy tape provides a tertiary signal that becomes powerful for large waist changes (9 cm = 13× signal-to-noise). Fat loss accuracy is ±22% at Week 4, tightening to ±10% by Week 12. The system shows honest 95% confidence intervals: "You lost 6.5 kg of fat [5.8–7.1 kg]." For resistance trainers, it tracks muscle preservation: "Your RT is preserving muscle: +0.9 kg." Known limitation: significant core/compound strength training can confound waist measurements — for these users, we recommend supplementing with periodic DEXA scans.

Do I have to weigh every meal forever?

No! You only weigh 3 portions per meal type during the first 1-2 weeks to calibrate your portion sizes. After that, 90% of your meals are single-tap selections from a predicted top 5 list. The system learns your typical 202g chicken breast or 300g curry portion and applies it automatically. You only weigh again when trying a completely new meal or when intentionally recalibrating portion sizes for behavior change.

How much time does tracking actually take per week?

About 2.2 minutes per week total (131 seconds): Meal logging: 56 seconds (28 meals × 2 sec average), Receipt scan: 15 seconds (Mistral AI auto-maps items), Pantry maintenance: 60 seconds (reweigh 3-10 staples/week, varies by household). Compare to competitors: 15-23 minutes per day / 2+ hours per week (Harvey et al. 2019, <em>Obesity</em>). Eatomate is 60× faster because of single-tap meal logging, expiry-based auto-reconciliation, and network-effect receipt mapping.

How does single-tap meal logging work?

After calibrating your portion sizes (3 weigh-ins per meal type), the app predicts your top 5 likely meals based on time of day, eating history, pantry inventory, and preferences. 90% of the time, your meal is in that list—just tap it. 9% of the time you search manually (10 seconds). 1% of the time you weigh a novel item (30 seconds). Weighted average: 2 seconds per meal.

Do I have to manually map receipt items to pantry items?

No! Mistral AI automatically maps receipt line items to our ingredient taxonomy. The first time ANY user scans a receipt item (e.g., 'Tesco Whole Milk 2L'), Mistral maps it once and saves it to our database. All future users get instant mapping via network effects. By month 12, 95%+ of common items are pre-mapped, requiring zero manual work.

What are Gaussian portion models?

When you weigh the same meal 3 times (e.g., chicken breast: 195g, 210g, 200g), the system learns your typical portion (202g) and how much it naturally varies (±8g). This lets it estimate future portions automatically. But here's the key: These are just initial estimates. When pantry items expire (e.g., chicken package finishes), reconciliation uses physics (mass conservation) to calculate the TRUE consumption and retroactively corrects all historical meals with ground truth. Portion models enable fast logging; reconciliation delivers research-grade accuracy (93%).

What happens to my weight loss in the first 1-2 weeks?

The first 1-3 weeks of any dietary change involve significant glycogen and water shifts — not just fat loss. If the energy balance predicts 0.45 kg of fat loss in Week 1 but the scale shows −1.5 kg, the 'extra' 1.05 kg is glycogen and water depletion. Eatomate's body composition model identifies this immediately, pinning glycogen state from the very first week. By Week 4, glycogen has fully settled and the model retroactively decomposes all earlier weight changes into fat, muscle, glycogen, and water — with 95% confidence intervals. This is why Week 2's apparent 'plateau' isn't a plateau at all: your fat loss rate was consistent, the glycogen adjustment just finished.

💳
Billing & Subscription

What's included in the free trial?

All features are included during your 30-day free trial. No credit card required. Cancel anytime with no penalties.

Can I switch plans later?

Yes! Add or remove household members anytime. Changes take effect immediately.

How do I switch between monthly and annual billing?

To switch between monthly and annual billing, go to Settings > Subscription and cancel your current subscription. Your current plan remains active until the end of your billing period. Once it expires, refresh the page and you'll see the subscription ended screen where you can choose your preferred billing cycle (monthly or annual) and resubscribe. Your data is retained throughout this process.

What payment methods do you accept?

We accept all major credit and debit cards (Visa, Mastercard, Amex, Diners Club), plus Link, Amazon Pay, Revolut Pay, and Klarna—all through Stripe. We do not store card details—Stripe handles all payment processing securely.

Is there a long-term commitment?

No. All plans are month-to-month or annual (prepaid). Cancel anytime with no penalties. Annual plans can be refunded within 14 days of the annual subscription payment.

Do you offer refunds?

Refunds are provided on a case-by-case basis within 14 days of payment. Contact support@eatomate.co.uk with your reason and we'll work with you.

What happens if I cancel my subscription?

Your subscription remains active until the end of your billing period. After that, you lose access to the Service but can re-subscribe anytime to restore your data (we retain your records for 90 days). This is different from deleting your account—canceling just stops payment while keeping your data safe for 90 days.

Can I pause my subscription?

We don't offer subscription pausing. You can cancel your subscription anytime and re-subscribe later to restore your data (kept for 90 days after subscription ends).

Do you offer student or family discounts?

Not currently. However, our household plan (£9.99/month per additional member) is already significantly discounted. Annual plans save 50% vs monthly.

What happens to my data if I don't pay?

Your subscription remains active until the end of your billing period. After that, you lose access but we retain your data for 90 days. Re-subscribe within 90 days to restore everything. This is subscription cancellation (90-day retention), not account deletion (30-day grace period).

🔒
Privacy & Data

What data does Eatomate collect?

Receipt scans (stored for 30 days then automatically deleted), barcode data, pantry inventory, meal logs, nutrition history, exercise logs, and body composition measurements. We also collect account info (email, name, height/weight/age/gender for calculations). Your city and country are stored in your profile; store locality is derived from receipts (raw addresses are immediately discarded). See our Privacy Policy for complete details.

Do you sell my data?

We never sell your personal information. De-identified pricing intelligence (individual product prices with no user ID) and anonymized aggregate statistics (population-level health and nutrition trends with no individual records) may be used for commercial research, public health intelligence, and data licensing. No aggregate statistic from user data is ever based on fewer than 5 users (k-anonymity). Your exact weight, height, age, and body composition are never included in aggregate datasets. See our Privacy Policy for full details.

Where is my data stored?

Data is stored on secure servers in the EU (GDPR compliant). All data is encrypted in transit (TLS) and at rest (AES-256). Receipt scans are deleted after 30 days.

Can I delete my account?

Yes. Settings → Account → "Delete Account". You'll have a 30-day grace period to cancel the deletion. After 30 days, your personal data is permanently deleted per GDPR requirements.

What's the difference between canceling my subscription and deleting my account?

Canceling Subscription: You stop paying but your account remains. We keep your data for 90 days so you can re-subscribe and restore everything. Deleting Account: You want to permanently leave. Your account enters a 30-day grace period, then all personal data is permanently deleted per GDPR. Cancel = pause with 90-day data retention. Delete = permanent with 30-day grace period.

Can other household members see my meals?

No. All meals are private. Other household members cannot see your logged meals. Each member has their own accuracy calibration.

How long do you keep my receipt scans?

30 days maximum. Receipt scans are automatically deleted after 30 days for privacy and storage efficiency.

⚙️
Technical

I can't see the chat bubble for support. How do I enable it?

The chat bubble (Intercom) requires functional cookies to be enabled. If you're not logged in, scroll to the bottom of any page and click "Manage Cookie Preferences" in the footer. In the modal that appears, toggle ON the "Functional Cookies" option, then click "Save Preferences". If you're logged in, go to Settings → Privacy & Cookies, find the "Cookie Preferences" section, and enable "Functional Cookies", then click "Manage Cookie Preferences" to save. The chat bubble should appear immediately in the bottom right corner. If you still don't see it after enabling functional cookies, try refreshing the page.

What phones are supported?

iOS: iPhone 6S or newer with iOS 11+. Android: Most devices from 2018+ with Android 7.0+. Camera required for barcode and receipt scanning. A kitchen weighing scale is recommended for precise meal logging. Web dashboard works on all browsers.

Can I download my data?

Yes! All plans include CSV data download. Settings → Account → "Download My Data". You own your data and can take it anywhere.

Is this a medical device?

No. Eatomate is not a medical device and does not provide medical advice. Always consult healthcare professionals before making dietary changes.

Can I integrate with other apps?

We do not offer API access. However, you can download your data manually (Settings → Account → Download My Data) for use in other applications or spreadsheets.

Does Eatomate require an internet connection?

Yes. Eatomate requires an active internet connection for all features including meal logging, barcode scanning, receipt scanning, and nutrition tracking. All processing happens on our servers to ensure accuracy.

Can I use this on a tablet or computer?

The web dashboard works on any device for viewing analytics, managing settings, and exporting data. The mobile app (iOS 11+ or Android 7.0+) is recommended for meal logging, barcode scanning, and receipt scanning.

What equipment do I need to use Eatomate?

Eatomate has different equipment requirements depending on which features you want to use:

1. Kitchen weighing scale (mandatory for all users): Any digital kitchen scale that displays weight in grams will work. This provides exact portion sizes for meal logging, which is how Eatomate achieves research-grade 93% nutrition accuracy. Cost: ~£10-20.

For Exercise & Body Composition features (optional):
2. Bathroom scale (required): A standard digital bathroom scale for measuring your body weight. Used for weekly body composition check-ins and validating energy balance predictions against actual weight changes. Cost: ~£15-30.

3. Flexible measuring tape (required): A soft, flexible measuring tape for taking body circumference measurements (neck, waist, hip) used in the Navy Method body fat calculation. A sewing tape measure or dedicated body measuring tape works perfectly. Cost: ~£3-5.

Summary: Kitchen scale is mandatory for nutrition tracking (~£10-20). If you want to use exercise tracking and body composition features for complete energy balance validation, you'll also need a bathroom scale and measuring tape (additional ~£20-35). Total for full system: approximately £30-50.

How much storage does the app use?

The app is approximately 11MB on iOS and 54MB on Android. Cached data (meal photos, pantry items) may use additional storage over time depending on usage.

How accurate is body composition tracking?

The raw Navy method has ±3-6% absolute error versus clinical DEXA scans, with a proportional bias pattern (overestimates lean, underestimates higher body fat). Eatomate applies proportional bias corrections based on published validation data (Potter et al. 2022, <em>Frontiers in Physiology</em>; Combest et al. 2025, <em>Military Medicine</em>) to reduce this systematic error. After correction, the remaining error is smaller and more random — making week-over-week change tracking significantly more reliable. Controlled measurement conditions (morning, post-toilet, before meals, waist at navel, neck below Adam's apple) maximise repeatability. Our glycogen decomposition model also separates water weight from true fat loss. Note: if you do heavy core or compound strength training, simultaneous muscle gain in the trunk can confound waist measurements — in this case we recommend supplementing with periodic DEXA scans. Even with correction, the Navy method is a progress tracking tool, not a clinical instrument.

What do the error bars (± values) mean in my nutrition tracking?

Error bars show the uncertainty in your nutrition measurements using 2σ (two sigma) ranges. For example, '250g ± 5g' means we're 95% confident the true value is between 245g and 255g. Different food types have different biological CVs: manufactured items are tightest (±1-8%), dairy/grains/legumes moderate (±5-12%), meat and fish wider (±8-18%), and fresh produce widest (±17-25%). Restaurant meals carry ±62-67.5% uncertainty (±62% weighed, ±67.5% unweighed). Post-reconciliation, portion error collapses to zero — only the biological CV remains. This research-grade statistical approach honestly reflects real measurement limitations rather than showing false precision.

Why do you show error bars for calories and macros but not micronutrients?

We show confidence intervals for calories and macros — the numbers you make decisions on. Micronutrients are tracked but displayed without error bars, as natural food variation makes precise uncertainty ranges less meaningful at the individual meal level. Your weekly macro totals benefit from CLT-justified Gaussian error modeling — each recipe's total error is the sum of many independent ingredient deviations, which CLT justifies as Gaussian. Quadrature then combines independent recipe-group errors into a tighter 2σ confidence interval. Micronutrient bioavailability varies too much between foods to make meal-level error bars actionable.

How does Gaussian error cancellation (quadrature) work in nutrition tracking?

The Central Limit Theorem is a fundamental statistical principle that underpins Eatomate's accuracy model. Here's the precise logic:

Step 1 — CLT justifies the Gaussian error model at the recipe level. Your real spaghetti bolognese differs from the database version in many small, independent ways: slightly different pasta portion, a different brand of sauce, more or less oil, a heavier hand with cheese. Each ingredient-level deviation is a small, independent random error. CLT justifies that the sum of many independent errors converges to a Gaussian (normal) distribution — regardless of each individual error's shape. This is what justifies modeling the total nutritional error per recipe group as Gaussian (e.g., ±15% σ, or ±30% at 2σ).

Step 2 — Quadrature combines independent recipe-group errors into a provable 2σ confidence interval. Once we know each recipe group's error is Gaussian, we combine independent groups via root-sum-of-squares (quadrature): σ_daily = √(σ₁² + σ₂² + ... + σ_N²). For N groups with similar σ, this simplifies to σ / √N. With 6 independent recipe groups at ±15% σ each: daily error = 15% / √6 ≈ 6.1% (1σ), or ≈18% at 2σ — giving 82% accuracy as baseline.

Key distinction: CLT doesn't just say "errors average out over time" — it specifically justifies the Gaussian shape of the total error when many independent deviations contribute. This Gaussian shape is what makes quadrature mathematically valid and the 2σ confidence interval provable.

Important: CLT and quadrature only handle random errors from independent ingredient deviations, not systematic errors (like consistently using more oil than the database assumes). That's why we pair CLT (models random errors as Gaussian) with pantry reconciliation (corrects systematic errors through mass conservation physics) to achieve research-grade 93%+ accuracy.

Still have questions?

Can't find what you're looking for? Our support team is here to help.