The Accuracy Gap: Why Calculators Disagree
Our nutrient targets look different from the free calculators and AI chatbots you have tried. That is intentional. Many “calculator” pages use simple linear multipliers. Many AI tools summarize text without running a transparent, unit-consistent calculation. This page shows the assumptions we use and why energy-based scaling changes the result.
Generic Calculators
- Linear scaling (Simple weight × multiplier)
- It ignores the metabolic exponent (\(kg^{0.75}\))
- Provides zero breed-specific physiological adjustments
- Result: Can be off for toy and giant breeds
AI-Powered Tools
- How it works: It predicts the next word based on internet averages. It is a language model, not a clinical calculator.
- The Danger: It blends conflicting blogs into one confident-sounding mistake. It cannot execute a sequential, multi-step clinical algorithm with hard-coded safety ceilings.
- Missing: It lacks any hard-coded safety ceilings or metabolic guardrails.
Raw & Well Engine
- Allometric scaling (\(kg^{0.75}\))
- Applies a documented sequence of checks and constraints
- Calculates with breed-specific physiology models
- Result: Designed to be unit-consistent and reviewable
The Anatomy of a Calculation
Here is the exact 6-step path your dog's profile takes through the engine the moment you press 'Calculate'. The system runs sequential data processing.
This is how your 8-point input moves through a defined sequence of checks:
Step 1: The Biological Baseline (Intake)
We take your 8-point clinical profile and calculate the Resting Energy Requirement (RER) using Kleiber's Law: \(\text{RER} = 70 \times (\text{BW}_{\text{kg}})^{0.75}\). We then apply breed and activity multipliers to set the day-to-day Energy Requirement (DER).
Step 2: The Safety Gauntlet (Target Calibration)
Before generating your day-by-day goalposts, the profile goes through a sequence of safeguards. These include NRC Safe Upper Limits where they exist, and separate, labeled condition-specific policies when a veterinarian has confirmed a diagnosis.
Step 3: The Dry-Matter Matrix (Meal Planning)
As you build your bowl, the engine does not just add up raw weights. Raw food often contains a high proportion of water (often cited around 65-80%), so the engine converts every ingredient to its Dry Matter (DM) basis in real-time, preventing false deficiency flags. You can toggle between viewing nutrients as-fed (for meal prep) or on a dry matter basis (for clinical discussions) with a single click.
Step 4: The Bioavailability Audit
Before we check for gaps, we strip away what your dog may not absorb. Where the literature provides usable guidance, the engine can apply absorption/bioavailability adjustments. When data is uncertain, we treat the output as an estimate and prioritize transparency over false precision.
Step 5: The Clinical Feedback Loop
Every absorbed nutrient is cross-referenced against the NRC Recommended Allowances. If a gap exists, the engine consults the 200g Volumetric Ceiling. If fixing the gap with whole foods requires too much volume, it routes to a measured supplement recommendation.
Step 6: The Diagnostic Overlay (Symptom Correlation)
When you log an observation, the engine maps qualitative
severity tags (normal, warning, danger) against
the exact bonePct, meatPct, and organPct of that day's
meal plan to isolate clinical triggers without arbitrary scoring.
Step 1: The Biological Baseline (Intake)
Precision needs context. We cannot calculate a safe diet based on weight alone. First, we gather your dog's complete 8-point biological profile, then route it into the correct metabolic engine using Kleiber's Law: \(\text{RER} = 70 \times (\text{BW}_{\text{kg}})^{0.75}\).
The 8-Point Clinical Intake
To follow NRC 2006 metabolic scaling with care, our engine needs your dog's complete 8-point profile before it calculates a single calorie.
For \( kg^{0.75} \) Kleiber's Law scaling.
Triggers pediatric, adult, or senior growth/maintenance curves.
Can apply optional breed/context modifiers when relevant, without replacing energy-based scaling.
Accounts for significant hormonal metabolic shifts.
Adjusts baseline for safe, clinical weight management.
Calculates the final day-to-day Energy Requirement (DER).
Sets the architectural ratios (BARF, PMR, or Custom).
Activates safety overrides (IRIS staging, allergies, etc.).
Three Distinct Engines. One Platform.
A growing Great Dane puppy, an adult Dalmatian, and a senior Labrador with kidney issues require different clinical math. Raw & Well doesn't use a generic template; we route your dog through one of three specialized nutritional pipelines.
Pediatric & Orthopedic Development
Puppies do not grow in neat \"age buckets.\" They follow a continuous metabolic curve. Small errors in calcium or phosphorus during this window lead to irreversible Developmental Orthopedic Disease (DOD).
Protocol: Exponential Maturity Model
Raw Multiplier = 3.2 × (e^(-0.87p) - 0.1)
Final Multiplier = Math.max(Raw Multiplier, 1.4) // The Maturity Bridge
Practical Implications: Instead of you guessing your puppy's growth rate, our app uses math to throttle calcium intake during growth-sensitive windows. The Maturity Bridge smooths transitions so targets do not swing fast as a puppy approaches adulthood. (Model coefficients are an internal policy layer; NRC tables remain the reference for nutrient targets.)
Adult Maintenance & Genetic Safeguards
A 20kg Bulldog and a 20kg Border Collie burn energy at very different rates. Our engine uses Kleiber's Law, then applies labeled modifiers (activity, body condition, neuter status, and other context). Breed is optional context, not a substitute for measured energy intake.
Protocol: Allometric Scaling + Modifiers
RER = 70 × (Body Weight in kg)^0.75
DER = RER × [Activity × Breed × Neuter × BCS × POMC]
Practical Implications: Two dogs with the same body weight can still have different energy needs. We use allometric scaling as a baseline, then apply transparent context modifiers you can review and adjust.
Geriatric Preservation & Renal Health
As dogs age, they lose muscle mass. Unless they have a confirmed kidney diagnosis, senior dogs often need MORE high-quality protein. Our engine separates healthy aging from clinical renal failure.
Protocol: Sarcopenia Defense
Target = Base Protein Requirement × 1.3 (Sarcopenia Boost Factor)
Practical Implications: We don't starve your senior dog of protein just because they are old. We boost amino acids to prevent muscle wasting (sarcopenia), unless their kidneys require restriction.
Step 2: The Safety Gauntlet (Target Calibration)
Once we know your dog's baseline, the engine applies hard constraints before we suggest a single ingredient. We lock in strict NRC safety ceilings to prevent mineral toxicity and ensure structural health.
The Clinical Safeguard Matrix
We don't just calculate minimums; our engine executes a sequential data-processing pipeline that runs your dog's profile through active clinical checkpoints before you ever see a recipe.
Dynamic Metabolic Multipliers
Applies energy and context adjustments based on the dog’s profile (body size, activity, and life stage). When we use breed-specific considerations, we treat them as separate, reviewable assumptions - not as NRC values.
Orthopedic Absolute Anchors & Ratio Locks
Puppy growth is sensitive to calcium and phosphorus balance. The engine highlights Ca:P directionality and encourages life-stage-appropriate targets instead of “set-and-forget” ratios. Numeric targets should be pulled from the relevant NRC growth tables.
Toxicity Safe Upper Limits (SUL)
When NRC provides Safe Upper Limits (SUL) for a nutrient, the engine can treat them as “do not exceed” reference ceilings in the same units as the target (often per 1,000 kcal ME).
Renal & Hepatic Overrides
If a veterinarian has confirmed CKD staging, the engine can apply separate, labeled renal policies (for example, IRIS-aligned guidance). These are not NRC table values and should be reviewed with your veterinary team.
Verified Baseline Mandate & Ghost Data
Data-quality flagging: If an ingredient lacks usable nutrient data for an essential nutrient, the engine flags the uncertainty so you can decide whether to replace the ingredient or add a measured source. We scope profile data to your user session and clear it on logout to reduce residual data risk.
Step 3: The Dry-Matter Matrix (Meal Planning)
As you build your bowl, the engine avoids comparing wet weights to NRC standards. Because raw food often contains a high proportion of water (often cited around 65-80%), doing so produces systematic false deficiency flags. The engine converts every ingredient to its Dry Matter (DM) basis in real-time before any nutrient check runs.
Protocol: DM Conversion
DM Nutrient (mg) = Raw Nutrient (mg/100g) × [100 / (100 − Moisture%)] × Ingredient Weight (g) / 100
Practical Implications: Raw chicken breast at 74% moisture has only 26g of dry matter per 100g fed. If the engine compared the wet-weight zinc content against the NRC day-to-day target, it would flag a zinc deficiency in many cases even when the dog is closer to target. DM conversion reduces that “wet vs dry” mismatch.
Why This Matters in Practice
- A raw diet reporting 65% moisture means only 35% of each gram fed is actual nutritional content.
- Comparing wet-weight values to NRC DM standards overstates deficiencies by up to 3× for high-moisture ingredients like liver and muscle meat.
- Every ingredient in our database carries a verified moisture value. If moisture data is missing, the engine flags it as unverified and refuses to count the nutrient yield.
User Control: The Dry Matter Toggle
While the engine always calculates using dry matter in the background, we give you control over how you view the results. The Nutrient Analyzer includes a "Show Dry Matter Basis" toggle with two display modes:
- As-Fed Display (Default): Shows nutrient values per 100g of food as you'd feed it. This matches how ingredients are sold and makes meal prep intuitive. Example: "Chicken breast provides 23g protein per 100g as-fed."
- Dry Matter Display: Shows nutrient values per 100g of dry matter, removing water weight from the equation. This is the veterinary standard used in clinical research and allows direct comparison to NRC 2006 guidelines. Example: "Chicken breast provides 88g protein per 100g DM."
Why this matters: When discussing your dog's nutrition with a veterinarian, they'll reference dry matter values. When shopping for ingredients or prepping meals, you'll work with as-fed weights. Our toggle lets you switch between both perspectives without recalculating anything by hand.
Step 4: The Bioavailability Audit
Dry Matter weight is what's in the food. Bioavailability is what your dog can use. The engine applies NRC Bioavailability Coefficients to calculate the Effective Yield - the fraction of each nutrient that enters the bloodstream.
Zinc (Raw Meat)
~35% Bioavailability CoefficientCalcium (Bone)
Varies Bioavailability CoefficientPhosphorus (Bone)
Varies Bioavailability CoefficientIron (Muscle Meat)
Varies Bioavailability CoefficientManganese (Plant Sources)
Varies Bioavailability CoefficientProtocol: Effective Yield Calculation
Effective Yield (mg) = Estimated absorbed amount (mg)
Practical Implications: Absorption can differ by nutrient form, ingredient matrix, and the dog. When you treat every label value as absorbed in full, you can end up with false precision. We treat “effective intake” as an estimate and keep the assumptions visible.
Where home-prepared diets often go wrong
- Zinc: Trace minerals are a common weak spot in home-prepared rations unless recipes are measured and balanced (see Dillitzer et al., 2011).
- Ca:P ratio: Bone-heavy or bone-light recipes can swing calcium and phosphorus in opposite directions; the safest approach is to calculate and check against a life-stage-appropriate reference range rather than “eyeballing.”
- Manganese: Many common raw ingredients are low in manganese; if you do not add manganese-rich foods or supplements on purpose, it can end up low in the final ration.
Clinical Bioavailability Registry
Bioavailability is not a single fixed number for “raw.” It varies by nutrient form, ingredient processing, interactions, and the dog. This table shows the kinds of assumptions calculators sometimes use; Raw & Well treats these as configurable inputs, not universal constants.
| Nutrient Category | Ingredient Source | Abs. Coefficient |
|---|---|---|
| Minerals | Calcium (Bone) | Varies (context-dependent) |
| Phosphorus (Bone) | Varies (context-dependent) | |
| Zinc (Muscle Meat) | Varies (context-dependent) | |
| Vitamins | Vitamin A (Liver) | Varies (context-dependent) |
| Vitamin D (Fish Oil) | Varies (context-dependent) | |
| Vitamin E (Plant Oils) | Varies (context-dependent) | |
| Macronutrients | Protein (Muscle Meat) | Varies (context-dependent) |
| Protein (Bone/Cartilage) | Varies (context-dependent) | |
| Fat (Animal Source) | Varies (context-dependent) |
Note: Values represent FRACTIONAL absorption (e.g., 0.35 = 35% of the raw mineral enters the bloodstream). NRC requirements are expressed as absorbable amounts.
Step 5: The Clinical Feedback Loop
Every absorbed nutrient is cross-referenced against the NRC Recommended Allowances. When a gap exists, the engine follows a Food-First, Supplement-Supported decision matrix that closes deficiencies without overloading the bowl or the dog's caloric budget.
How the Engine Decides: Food vs. Supplementation
Our \"Food-First, Supplement-Supported\" algorithm follows a 3-step decision matrix to ensure safety without overstuffing the bowl.
1. The Volumetric Ceiling
Our engine has hard-coded physical limits. If fixing a Zinc gap requires an \"impractical\" amount of food that would cause digestive distress, the engine triggers a supplement recommendation instead.
2. The Caloric Anchor
If a dog is already at their day-to-day calorie limit, adding more food to fix a gap causes weight gain. The engine \"anchors\" the calories and uses a zero-calorie supplement bridge to protect the dog's Body Condition Score.
3. Decimal-Point Precision
Whole foods vary. Supplements allow the engine to fill a measured gap with more control than whole-food-only tweaks, in particular when you’re working in mg or µg ranges. Outputs are treated as estimates unless you have lab data; the goal is repeatability and transparency, not false certainty.
The Clinical Toolkit
From complex math to peace of mind.
Micronutrient audit (beyond macros)
Don't wait for symptoms of deficiency to show up. As you build a recipe, our real-time Deficiency Meter audits your ingredients against NRC reference targets. See where a recipe may be low in zinc, copper, or manganese, and get one-click whole-food booster suggestions.
Automated NRC Meal Planner
Transitioning from kibble can feel overwhelming. Our planner translates your dog's targets into a practical shopping list for the week. You'll know how many grams of meat, bone, and organ to prep - and what assumptions those targets rely on.
Step 6: The Diagnostic Overlay (Symptom Correlation)
When you log an observation, the engine maps qualitative severity tags against that calendar day's macronutrient ratios to isolate clinical triggers - without arbitrary scoring. Built for the kibble-to-raw transition and long-term gut stability monitoring.
The Math of Gut Stability
Built to manage the 'Kibble-to-Raw' transition for beginners and long-term stability for pros, this matrix eliminates trial-and-error feeding.
| Severity Tag | Observation | Engine Action |
|---|---|---|
normal |
Ideal / Firm stool, stable energy | Day marked stable. Macro ratios logged as baseline. |
info |
Minor softness, mild itching | Flag recorded. Overlay checks for elevated organPct that day. |
warning |
Unformed stool, repeated itching | Correlation check: bonePct, meatPct, new protein
introduced. |
danger |
Liquid / diarrhea, lethargy | Traceable trigger identified against that day's organPct and ingredient
list. |
The Correlation Overlay: You Report, We Map
There is no arbitrary numeric score. The engine uses a date-based overlay
instead. Each symptom log is matched to that calendar day's macronutrient breakdown - the
bonePct, meatPct, and organPct from your meal plan. If a
danger flag for liquid stool coincides with a high-organ day or the introduction of a
new protein, the engine surfaces a traceable trigger, not a guess.
You provide the observations; the app provides the clinical map.
Built for Accuracy, Validated by Experts
This page explains the calculation framing and the boundaries of what a calculator can and cannot guarantee. Use it as a transparency layer, and discuss diet changes with your veterinarian - in puppies and in dogs with chronic disease.
Related Resources
- New to NRC 2006? Start with our simple explainer →
- Want the deep-dive science? Explore our technical guide →
- For veterinary professionals: Request our Technical White Paper → (Available upon verified request)
Still Have Questions?
Is this aligned with NRC 2006?
We design our nutrient checks to follow NRC 2006 framing: requirement classes (MR, AI, RA) and Safe Upper Limits (SUL), often expressed per energy intake (for example, per 1,000 kcal ME). NRC does not approve individual recipes, so we treat it as a reference for targets and ceilings rather than a blanket guarantee.
Why are your numbers different from other calculators?
Most calculators use linear scaling (weight × multiplier). We use metabolic scaling (\(kg^{0.75}\)) based on Kleiber's Law, which reflects how energy works in living organisms. Small dogs need more per kg; large dogs need less.
Can I trust this over my veterinarian's advice?
This tool is designed to complement veterinary care, not replace it. We encourage you to share our calculations with your vet. Many vets appreciate having precise nutritional data to work with. Our engine is built for collaboration, not conflict.
What's the difference between this and the NRC explainer?
The NRC explainer answers 'What is NRC 2006?' This page answers 'How does Raw & Well use NRC science?' They work together: start with the explainer, then review the math here.
Is this appropriate for puppies and seniors?
The engine supports life-stage context (growth vs adult maintenance) and can add extra, labeled clinical policies when a veterinarian has confirmed a condition (for example CKD staging guidance). These policies are not NRC table values; they are applied as separate constraints and should be reviewed with your veterinarian.
Where can I see the full technical documentation?
Our complete clinical specification (including formulas, constants, breed multipliers, and safety caps) is available in our Technical White Paper for veterinary professionals and technical users.
Ready to review the math?
Let the engine generate an NRC-aligned starting point you can review, adjust, and share with your veterinarian.
Start Your 7-Day Free Trial