How I Evaluate a Pet Food Company

The site frames pet food in three eras. Pet Food 1.0 was mostly about nutrients. Pet Food 2.0 became mostly about ingredients. Pet Food 3.0 has to integrate both, plus processing and the individual animal. This page is the operational version of that idea: what evaluating Pet Food 3.0 actually looks like when you sit down with a brand.

Pet Food 3.0 looks through four lenses — the Four Ps: Profile, Parts, Process, and Pet.

That is the memory device for the larger framework I talk about across the longevity pages: nutrients, ingredients, processing, and personalization. The Four Ps are simply the owner-facing version of that idea.

For more context on why nutrients, ingredients, processing, and individual fit matter to healthy aging, see the canine and feline longevity pages.

All four of those lenses sit downstream of one larger question.

Who runs this company?

Most pet food is produced by a business. That is not a complaint; it is the structural fact most checklists ignore. When you evaluate a pet food, you should also be evaluating the company. The product is downstream of the company’s mission, its leadership, its capital structure, and the trade-offs those things force.

Four questions sit above the product.

What is the company’s mission, and does it resonate with you?

Most companies start with a wedge: disruption, value pricing, fresh delivery, a particular ingredient story, a particular processing story, veterinary credibility, convenience, sustainability, or science. Read interviews with the founders. Read what they have published. Some of that will be marketing. Some of it will be the real thing. You are looking for a coherent mission that survives the marketing layer.

Who is in charge, and how credible are they?

Founders and senior leaders set the tone, and that tone shapes everything downstream: formulation decisions, food safety culture, marketing claims, response to recalls. You will rarely meet these people, but most of them have a public footprint. Read it.

Who is the company responsive to?

A publicly traded company has thousands or hundreds of thousands of shareholders to report to and quarterly margins to hit. A privately held company may be answering to a single family. A venture-backed startup may be building toward an acquisition or a growth milestone. None of these structures is automatically good or bad. But they produce different trajectories, and the same brand at one stage of capital structure can be a meaningfully different company at another.

Who is actually responsible for the recipe today, not just at launch?

This is where many existing checklists fall short. “Is there a veterinary nutritionist on staff?” is no longer a complete question. It is a question we now need to get past and redefine. By now, almost any company that wants to can claim a credentialed nutritionist somewhere in the history of the product or somewhere on the org chart. That tells you less than most owners think.

The questions that actually matter are different.

Who formulated the diet? Are they still at the company? Are they responsible for the diet today, or did they hand off a recipe years ago and move on? When the recipe needs to change for cost, ingredient availability, supply disruption, or processing reasons, who signs off? Is it still the same person who put the original formula on paper? Does that person have real authority, or are they advisory?

I have seen recipes, including recipes I originally wrote, modified later with ingredients or premixes I would not have approved, while still being described publicly as nutritionist-formulated. The recipe is a snapshot. The team is the ongoing answer.

It also takes more than a nutritionist to build a good pet food. A serious team may include a formulator with experience in the specific format, a dedicated food safety lead, a production expert, quality and regulatory expertise, and ideally a scientific or educational team member who can talk about the product in a way that is not marketing-flavored. Kibble, canned, fresh, freeze-dried, raw, and air-dried foods each have their own processing and formulation problems.

If a company has one credentialed nutritionist who is a consultant and considers the box checked, it may not have the right team, depending on the portfolio and products being sold.

A few company-level red flags worth naming: the company will not tell you who formulated the diet or who is responsible for it now; the “nutritionist on staff” line has no name attached or names a person who is no longer involved; food safety has no clear internal owner; marketing material is not reviewed by anyone with scientific, veterinary, or nutrition expertise before it goes out.

None of these is automatically disqualifying. All of them are conversations.

Nutrients are still the first technical question, and often the easiest to answer.

Before the conversation turned to ingredients, we argued about whether dogs and cats were getting the nutrients they actually needed. That argument produced the modern nutrient standards and the foods designed primarily to meet them. Those foods sometimes used cheap, common, or highly available raw materials, but they also solved a real problem: preventing nutrient deficiency and providing digestible, complete diets at scale.

Pet Food 1.0 solved a real problem and created the conditions for the next one.

Nutrients still matter. Here is how I look at them.

Evaluate on a caloric basis.

Nutrient claims on a pet food label are usually expressed as percentages: by weight, “as fed,” or on a dry-matter basis. None of those, by itself, tells you what the animal is actually getting per calorie consumed. The unit that often matters clinically is per calorie, usually expressed as grams per thousand kilocalories, milligrams per hundred kilocalories, or percentage of calories from protein, fat, and carbohydrate.

A dry food, canned food, and fresh food can look very different on an as-fed label simply because moisture and calorie density change the math. Labels are changing, and new labels should make some of this easier by showing the percentage of calories from protein, fat, and carbohydrate. But owners will still need to know where to look and what those numbers mean.

The first question is simple: can the company give me the nutrient profile on a caloric basis?

If they cannot, that itself is a signal. They may not be evaluating their own product in the way a veterinary nutritionist would.

Matching nutrients to the animal is where a veterinarian or board-certified veterinary nutritionist can be especially helpful. A senior dog may need more protein, not less. A cat with kidney disease may need lower phosphorus, and increasingly, attention to the form of phosphorus, not just the amount. A dog with chronic enteropathy may need higher digestibility across protein and fat. These are personalization decisions that depend on the nutrient profile being expressed in units a clinician can actually use.

Beyond the guaranteed analysis.

The guaranteed analysis on most pet food labels was never built to support cross-product comparison. Protein and fat are listed as minimums, not typical amounts. Fiber and moisture are listed as maximums. The values are usually percentages by weight, as fed. That means moisture, fiber, and calorie density can make two foods look similar or different in ways that do not reflect what the animal actually receives.

Pets eat calories, not label percentages. So the more useful question is how many grams of protein, fat, carbohydrate, phosphorus, sodium, EPA/DHA, or other nutrients are delivered per unit of energy.

For many nutrients, that means grams, milligrams, or international units per 1000 kcal. For macronutrients, percent of calories from protein, fat, and carbohydrate can also be useful, especially as newer labels begin to show calorie distribution more clearly.

The table below gives broad comparison ranges for generally healthy dogs and cats. These are not medical targets. They are a way to understand whether a food is relatively low, moderate, or high in protein, fat, or carbohydrate before asking the more important question: is that profile right for this animal?

Percent-calorie values below use modified Atwater factors (3.5 kcal/g for protein, 8.5 kcal/g for fat, 3.5 kcal/g for carbohydrate), which have historically been used in pet food. This may underestimate the actual calorie contribution from different macronutrients in highly digestible diets like home-prepared or raw diets. Pet food labels usually do not list carbohydrate directly; it is usually estimated by subtraction from the other proximate values. Pet food energy calculations may use metabolizable-energy estimates that differ by product and method, so these should be treated as practical ranges rather than exact biological values.

There is a meaningful gap between “formulated to meet” a nutrient profile and “tested in finished product to verify.” Most companies formulate to a target. Fewer measure the finished food in a lab to confirm the formula came out the way it was designed. Fewer still do this on a meaningful schedule. A full typical analysis of a finished diet costs a few thousand dollars. It is a real but bounded investment.

Measurement matters more, not less, as you move away from extruded kibble. Extruded pet food often uses ingredient streams designed to hit consistent protein and fat values across batches. Fresh, gently cooked, freeze-dried, and human-grade raw materials can be more variable batch to batch, paradoxically, because the supply chain for human-grade meat and produce is often built around consumer-facing consistency rather than tight nutrient consistency. A “high-protein fresh diet” can swing more than its kibble equivalent if no one is checking the finished food.

So when a company markets less-processed food, ask harder questions about typical analysis.

Do you measure the finished food? What do you measure? How often? What do you do when a result comes back outside specification?

On the evidence hierarchy.

A few common credibility signals are worth more discussion than they usually get.

Formulated to meet AAFCO or FEDIAF profiles is the basic standard. Most reputable companies do this. It tells you what the food was designed to provide, not necessarily what is present in every finished batch.

Typical nutrient analysis on the finished product is meaningfully better. It can catch formulation errors, ingredient substitutions, processing losses, and real-world variation.

Digestibility testing is more revealing than people realize. Crude protein on a label tells you how much protein is in the food. Digestibility helps tell you how much of that protein the animal can actually use. Two diets with identical protein numbers can perform differently. Functional high-fiber diets may deliberately score lower on overall digestibility, which may be appropriate, but outside that intentional design choice, digestibility is one of the best indirect markers for ingredient quality and formulation discipline.

AAFCO feeding trials are evidence, but they answer a narrower question than many people assume. Adult maintenance feeding trials are more likely to detect obvious inadequacy than to prove optimal long-term nutrition. Growth and lactation trials can catch more because the demands are higher. Feeding trials are useful, but they are not the whole answer. There is also a structural disadvantage for smaller companies: trials require either a colony of research dogs or cats, or access to a facility that maintains one, which is a meaningful cost. The absence of feeding trials at a smaller brand is not, on its own, a damning signal.

Peer-reviewed research is one of the clearest signals that a company is willing to test claims beyond marketing. It is especially useful for smaller, founder-led, or venture-backed brands that have not yet had decades to build a reputation. It is also easily overextended. Foundational research is often done years before the current formulation is sold. Products that “have research behind them” sometimes do, and sometimes have research behind a different formulation that has since changed.

Pet Food 2.0 was the era of ingredient transparency, ingredient anxiety, and ingredient marketing. Some of that conversation produced real improvements. Some of it produced a vocabulary that owners now use with confidence but that does not always mean what they think it means.

A few ideas are worth getting right.

Human-grade and feed-grade are categories about regulatory status, not automatic categories about quality.

A human-grade ingredient meets the regulatory standards for human consumption at every stage of processing and handling. A feed-grade ingredient does not. That distinction is meaningful, especially in a food-safety world that has changed substantially over the last two decades.

But “human-grade” is not synonymous with “high-quality.” Plenty of human-grade cuts of meat are lower quality from a nutritional standpoint: high in connective tissue, variable in fat, or intended for processed foods rather than direct consumption. A high-quality feed-grade ingredient stream with good supplier audits can be a better input than a low-quality human-grade one.

The label is the start of the question, not the answer.

Ask, too, whether “human-grade” applies to every ingredient: the meats, the produce, the additives, the carrier ingredients in the vitamin and mineral premix, the palatants, the oils, the fibers, and the processing aids.

Ingredient names tell you less than people want them to.

“Meat byproducts” and “meat” are nonspecific terms that hide real variation. The percentage of meat in a formula is mildly informative, but the full ingredient panel does not tell you about digestibility, supplier quality, batch-to-batch consistency, or how the ingredient was actually handled. Those are under the company’s control and rarely disclosed publicly.

Mechanically separated meat is a useful example. It is produced by physically separating residual meat from bones using high pressure. It can be cost-effective, but the process may incorporate more collagen and connective tissue than hand-deboned meat, which changes the amino acid profile and potentially the digestibility of the resulting protein. Most companies do not disclose whether they use mechanically separated meat, in part because the label may not make it obvious. It is the kind of detail you only get by asking.

Digestibility, again.

The single best indirect marker of ingredient quality is often digestibility testing. Has it been done? On this product, or on a previous formula? In which species? What were the results? Companies that take ingredients seriously usually have these numbers and are willing to share them, at least directionally.

Phytonutrients are the next frontier and are almost never measured.

Carbohydrate sources and fiber sources can be meaningful sources of phytonutrients: bioactive plant compounds that may affect the gut microbiome, inflammatory tone, and other systems. The carbohydrate-or-no-carbohydrate debate has obscured this. We know fiber affects the microbiome. We know specific plant compounds have measurable biological effects in other species. The problem is that phytonutrients are difficult and expensive to measure, and they are not part of any standard pet food testing panel.

This is likely to become a larger conversation in pet nutrition. The companies that will lead it are not the ones with the longest ingredient list. They are the ones building the analytical capacity to understand what their ingredients actually deliver.

Supplier transparency is a fair question.

Where do the major ingredients come from? Does the company audit its suppliers? How often? What happens when an audit fails? Most companies will not tell you the names of their suppliers because those are competitive relationships, but they should be able to talk about their supplier standards and audit program.

Also worth saying clearly: quality control matters more than the country an ingredient is sourced from. For some ingredients, sourcing within the United States is not realistic. You cannot source kangaroo domestically. What matters is whether the company has a defensible quality-control process around the supplier, not simply where on a map the ingredient was harvested.

Recalls are not always evidence of bad practice.

This deserves emphasis because the public conversation around recalls is often reactive. Pathogens, particularly Listeria, which thrives in refrigerated and damp environments, are a fact of commercial food production. Producing food at scale exposes ingredients to many processes and many environments, and even rigorous food safety programs occasionally identify contamination that triggers a recall.

A company that recalls promptly, communicates clearly, and identifies the root cause may be showing signs of a functioning food-safety system. A company that has never had a recall may be excellent, lucky, small, or not testing aggressively enough to know.

The right question is not “have you ever recalled a product?”

It is “how do you handle a recall when it happens, and what did you change after the last one?”

Processing is the tier most owners think about least and where some of the most genuine uncertainty lives.

Facility type.

The first question is whether the food is made in a pet food facility, a human-grade food facility, or a contract manufacturer that produces for multiple brands. None of these is automatically better.

Extruded kibble is generally not made in human-grade facilities, because the equipment, ingredient streams, and production demands usually differ from human food production. Air-dried, freeze-dried, canned, and fresh products can be made in human-grade facilities, and many are.

Contract manufacturers, meaning facilities that produce for many brands, have a mixed reputation that is often undeserved. A good contract manufacturer is sometimes the best option because production is their whole business and they may have invested in compliance and quality control that smaller in-house operations cannot match. Several large and well-known fresh pet food brands use contract manufacturers for exactly this reason. The trade-off is reduced direct control over the specifics of the process.

A company that operates its own facility has the most control in theory. In practice, owning your own facility is no guarantee. Small in-house facilities can be run with serious food-safety discipline, or they can be operated in ways that would make me uncomfortable. The relevant question is not “do you own your facility?” It is “what are your food-safety practices, who is responsible for them, and what does your most recent audit look like?”

Processing parameters.

Within every category, including kibble, canned, fresh, raw, gently cooked, home-prepared, freeze-dried, and air-dried, processing varies enormously. Two diets in the same category, sold with the same descriptive language, can be produced very differently.

The toast analogy is the simplest version. You can take a slice of sourdough and lightly warm it, or you can burn it black, and both are still “toasted bread.” Pet food works the same way. Two “gently cooked” diets can be processed at very different temperatures, pressures, and times, producing very different chemical end states. The label does not tell you which one you have.

Two specific outputs of processing are worth knowing about.

The first is gut microbiome impact. Short-term studies in dogs suggest that the same formula extruded under different conditions can produce different microbiome composition, different short-chain fatty acid output, and different immune responses. Same recipe, different processing, different downstream effects.

The second is advanced glycation end products, or AGEs: compounds formed when sugars and amino acids react under heat. Higher-temperature processing can generate more of them. Their long-term impact on canine and feline health is still being characterized. Counter to most consumer intuition, some commercial canned diets, which are often perceived as less processed because they look more like meat, have produced high circulating AGE responses in dogs in the studies done so far. Whether that matters clinically is exactly the question the field has not yet answered.

The point is not that one format is good and another bad. The point is that “processed” is too blunt a word, and the category alone does not tell you what happened to your food.

Processing aids and palatants.

This is one of the least-addressed transparency problems in pet food. A range of compounds may be used during production: processing aids that help the food flow through equipment, palatants that improve taste, additives that influence texture or manufacturing, and ingredients that may sit under broad terms like “natural flavors.”

Some may not appear on the label in the way owners would expect, depending on their function, level, and labeling category.

This matters more in some products than others. Soluble phosphate additives in cat foods are a meaningful example because phosphate sources can differ in solubility and absorption, and may produce different post-meal phosphorus responses than phosphorus naturally bound in animal tissue or plant matter. That may be relevant for older cats and cats at risk for kidney disease.

The market does not currently provide the level of processing-aid transparency I would like. You can ask, and the act of asking pushes the industry.

Food safety governance.

Food safety is not a slogan. It is a system.

One strong signal that a company takes food safety seriously is whether someone inside the company clearly owns the food safety plan.

External consultants can be useful. Contract manufacturers can be highly capable. But a company should still know who is responsible for food-safety decisions, who reviews deviations, who controls recall procedures, who communicates with the manufacturer, and who has authority to stop production or hold product when something is wrong.

Most companies will not show you their food safety plan. They should be able to describe its existence, name the people responsible for it at a general level, and explain how often it is reviewed or audited.

It is worth understanding that food safety programs typically test for what they are required to test for and what their hazard analysis identifies as most relevant. They do not test for every theoretical hazard. Comprehensive testing for pesticide residues, heavy metals, environmental contaminants, toxins, and adulterants is not standard for every product or every batch.

That does not mean testing is unimportant. It means testing has to be meaningful, validated, and tied to decisions.

The better question is not only “do you test?”

It is: what do you test for, why do you test for it, how often do you test, are the methods validated, and what happens if a result fails?

Raw food and pathogen control.

Raw foods deserve special attention because untreated raw meat can carry pathogens that matter for pets and people.

Commercial raw foods vary widely. Some use high-pressure processing. Some use bacteriophages. Some use fermentation or other hurdles. Some test finished products. Some rely on sourcing and freezing. Some have robust pathogen-control systems. Some do not.

Regulatory and market pressure have pushed many commercial raw companies toward pathogen-reduction steps such as high-pressure processing or bacteriophage treatment. A commercially produced raw food with validated pathogen-reduction steps is not the same as meat from a grocery store or slaughterhouse chain prepared at home without a kill step.

That does not make raw food risk-free. It means the risk depends heavily on the system.

And in the spirit of transparency, dried pet foods are recalled too for pathogens. Many of these occur in processes post-cooking, for example when palatants are sprayed onto the food for flavor.

Again, the category does not answer the question. The process does.

The fourth lens is where the framework meets the animal. None of the previous three tiers matters if the resulting food does not fit the dog or cat in front of us.

For animals with medical conditions, personalization is high-stakes and concrete.

A dog with chronic enteropathy may need very high-digestibility protein and fat, with carefully chosen fiber. A cat with kidney disease needs the right phosphorus level, and increasingly, the right form of phosphorus, not just the right amount. A senior dog losing muscle may benefit from more protein, not less. A dog or cat with a true food allergy needs not only the right protein source but a manufacturing context with no cross-contact from other proteins.

These are personalization questions where the answer can be specific and the consequences are clinical.

This is where a veterinarian or board-certified veterinary nutritionist earns the role. The label cannot tell you what your animal needs. A veterinarian or veterinary nutritionist can help match the food to the animal.

For healthy animals, personalization is more honest, but less concrete.

The harder truth, and one many of my colleagues will not state directly, is this: outside of medical conditions, we very rarely know what the optimal nutrient profile is for a specific healthy dog or cat.

“Complete and balanced” is a floor. It is not personalization. Modern pet food can keep almost any healthy animal alive and avoid catastrophic deficiencies. What it cannot yet do is tell you the optimal protein, fat, fiber, and micronutrient profile for this animal at this age in this household with this activity level.

That will eventually change. The work moving in this direction is blood-based, microbiome-based, and genetic-based. We are not there yet. In the meantime, the most honest answer to “what is the optimal food for my healthy dog or cat?” is some version of: we do not fully know, and humility is appropriate.

Why I think about rotation, given that uncertainty.

Here is a thought experiment I have used in lectures. Suppose I told you that you had to choose one specific food today, with all the information you currently have, and eat that exact food for the rest of your life. No variation. No phytonutrient differences. No protein source changes. No seasonal adjustment. You would, I assume, find that requirement absurd, because no single human food is optimized for the full range of nutrients and bioactives a person benefits from across years.

We ask animals to do that constantly.

For healthy animals, assuming the foods meet basic nutrient requirements, I think thoughtful rotation across complete and balanced diets is the rational response to our incomplete knowledge. You change premixes. You change ingredient panels. You change processing philosophies. You change, to some extent, the formulators behind the recipe. If we do not know what the single optimal answer is, varying across credible options is a hedge against having locked into the wrong one.

For animals with medical conditions, the calculus is different. Stability of the diet is often the medicine, and rotation should not be applied reflexively.

What you can actually act on today.

Match the diet to the animal, re-evaluate when the animal changes, and use rotation only when medical stability does not matter more. For animals with real medical needs, a veterinarian or board-certified veterinary nutritionist can help determine which nutrient profile, ingredient pattern, processing format, and feeding strategy actually fit.

A page like this should be honest about its own limits.

You will not get every answer.

• You may not know all suppliers.
• You may not know premix carriers.
• You may not know processing aids.
• You may not know palatants hidden under natural flavors.
• You may not know exact cooking parameters.
• You may not know batch-to-batch variation.
• You may not know how much authority the nutritionist really has.
• You may not know whether a published study reflects the current formula.
• You may not know whether ingredients have changed in ways that do not trigger obvious label changes.
• You may not know what happens inside the plant.

That does not mean every company is hiding something dangerous.

Often, the information is not public because the market has not demanded it, competitors could use it, formulations are proprietary, suppliers are protected, manufacturing details are sensitive, or the company is following industry norms.

Pet food companies are businesses. They may protect information for reasons that are understandable, even when those reasons are frustrating.

This is why I do not think the answer is paranoia.

The answer is pressure.

Informed, constructive, market pressure.

The reason I publish this framework is not to give owners a tool to interrogate brands one at a time, although that is one fair use. It is because the entire conversation about pet food transparency only moves when consumers demand more.

Industry conferences will not do it alone. Veterinary nutritionists talking to each other will not do it alone. We already have strong opinions about what should be disclosed, and those opinions have not produced enough disclosure. Regulatory pressure will help in some areas and lag in others. The real lever is purchasing.

Companies respond to demand.

If owners begin asking the questions above, and shift purchasing toward companies that answer them, the industry will move toward greater transparency. If owners do not, the current asymmetry will persist: companies know far more about what is in the food, how it is made, and how it changes over time than owners do.

Better questions will not make pet food simple.

They can make it more honest.