The science-backed supplement claim has become a competitive necessity in the pet supplement space, but the path from "we should do a study" to a defensible, marketable body of data is littered with expensive mistakes.
At May’s National Animal Supplement Council Annual Conference, Ravi Sheth, co-founder of Kingdom Superculture, and Sara Phillips, vice president of global pet business at FoodScience, drew on more than a dozen combined in vivo pet clinical studies to map the terrain: where brands go wrong, what good design actually looks like, and how to build rigorous science without a massive budget.
5 pitfalls to avoid in pet supplement clinical research
Sheth framed clinical research as vulnerable to "death by a thousand cuts" — a process so complex that a single misstep in design, execution or stakeholder alignment can render an expensive study unusable. He and Phillips walked through five predictable failure points.
Pitfall 1: No clear definition of "win"
Phillips described an early FoodScience experience that cost the company six figures. The team invested in an impeccably executed study — double-blind, placebo-controlled, randomized crossover — comparing its flagship joint supplement against a top competitor. The science was sound. The problem was the outcome: the competitor performed equally well.
"The mistake we made was not truly defining what we wanted our outcome to be," Phillips said. "What does a win look like? And perhaps more importantly, what are the other possibilities for your research outcome?"
Her takeaway: treat clinical research as a lifecycle, not a one-time bet. "Think about a series of clinical studies starting small and growing over time into a cohesive data story."
Pitfall 2: Improperly designed studies
A well-designed study, Sheth said, is double-blind, placebo-controlled, with clearly defined primary and secondary endpoints and a pre-specified statistical analysis plan, all documented before data collection begins.
"There's a predefined statistical analysis plan, a predefined set of biomarkers that you're going to analyze," he said. "It's really easy once you've done a very expensive clinical study to then kind of look at the data and start doing statistics to try and find an effect, but from a technical and scientific perspective, it's really not rigorous."
He also cautioned against treating clinical research as a one-time exercise. On one Kingdom ingredient alone, what began as a single planned study has grown into five completed trials with three or four more in progress. "It’s important to think about as a pipeline and a competency rather than just a one-off clinical study."
Pitfall 3: Wrong study population or poor group stratification
If a joint supplement is intended for aging dogs with existing mobility issues, testing it on young, healthy dogs is likely to produce underwhelming results and render the data unusable for its intended commercial purpose.
"That population should be representative of the intended use of the product," Sheth said. He described how Kingdom enrolls dogs with measurably elevated breath compounds for its oral health studies, then distributes those animals into placebo and treatment groups stratified by baseline biomarker levels. Skipping that step, he explained, introduces a significant statistical confounder.
He also issued a regulatory caution that applies broadly across the pet supplement space: avoid studying animals with diagnosed medical conditions. "If you're marketing and working on a product in the pet supplement space, you've got to look at those exclusion criteria," he said. "If animals are on prescription medications or have a diagnosed medical condition, that study will not substantiate a health claim. It will substantiate treating and preventing a disease."
Pitfall 4: Leaving key stakeholders out of the process
"If you don't remember anything that I say today, please remember these three words: involve key stakeholders," Phillips said.
She shared a cautionary tale from earlier in FoodScience's history: a comprehensive body of research on a flagship joint product — placebo-controlled trial, in vitro studies, biomarker panels — was titled using osteoarthritis language. Years later, when the brand shifted toward consumer-facing claims, an NASC audit flagged the study title as an implied disease claim.
"It wasn't the end of the world," Phillips said. “But it's a really good example of how easy it can be to avoid these sorts of pitfalls when you just involve your colleagues and your teammates."
The stakeholders she identified as essential: the regulatory team, legal, marketing (for consumer insights), finance (to build and pressure-test the business case), and the commercial team.
"If your salespeople say they don't think their retail buyer is going to care about that claim, I guarantee they're not going to sell it."
Pitfall 5: Endpoints that don't resonate or substantiate anything
Measuring an impressive array of biomarkers is not the same as generating meaningful data. Sheth pointed to public examples from the human supplement space: studies with extensive analytical output that, on close examination, don't substantiate any consumer-noticeable benefit.
"If the study doesn't substantiate anything meaningful, it could be a liability," Phillips said. "People can see through it quickly."
Brands must also navigate the tension between transparency and protection. "You have to find that right balance between transparency and legal protection," Phillips said, noting the importance of working closely with the legal team before deciding what to publish or reference publicly.
A framework for best-in-class design
The alternative to these pitfalls, Sheth said, is a consumer-first design process that works backwards from the intended claim.
"Start with: what is the core unmet need? What is the core claim I want to make so consumers understand this product is differentiated?" he said. From there, map to clinical biomarkers that can credibly measure that benefit — ideally ones that correlate with changes pet owners will notice at home — and then translate those endpoints into a consumer story.
"Once you identify the unmet need or the set of claims you want to substantiate, the rest of the work becomes a lot easier," Sheth said. "If you haven't defined that up front, you have no constraints for all the work that comes after."
He used breath health as an illustration: primary endpoint is a validated instrumental measure of oral volatile compounds; secondary endpoint is a panel of human assessors evaluating the dog's breath directly. The combination of objective and sensory data builds a richer, more defensible story.
Building credible science on a limited budget
Neither Sheth nor Phillips argued that rigorous science requires a large budget — only that budget constraints must be matched with smart sequencing.
Phillips outlined two approaches. The first is scrutinizing ingredient supplier research before making a purchasing or formulation decision. "These folks should be willing and excited for your team to have rigorous discussions with them,” she said. “They should welcome your healthy and respectful skepticism on their research."
When evaluating a supplier's clinical data, she said, key questions include: How was the study designed? Is the population relevant to your target consumer? Was the validated dose close to what you intend to use in your product?
The second approach is the incremental study stack, as illustrated by FoodScience's Composure calming supplement. The product launched in the veterinary channel with a 25-animal internal field trial, the results printed on an unbranded flyer. That built internal confidence. A later, fully powered CRO study produced the claim "works within 20 minutes, lasts up to four hours." Subsequent consumer insight about longer duration led to a targeted follow-on study testing only the six- and eight-hour time points, yielding a new "lasts up to eight hours" claim — without having to replicate the earlier work.
"You can start small, gain confidence and grow from there," Phillips said.
Sheth agreed. "A survey is not a full clinical trial — but start somewhere," he said. "Not everything has to be there on day one. Starting small can be a smart strategy."
Execution: the right partner, locked documentation, regulatory alignment
Study design is only half the equation. Sheth closed with execution fundamentals drawn from Kingdom's work across more than a dozen in vivo studies.
Selecting a qualified contract research organization is the single most important operational decision, he said. "Being really comfortable with the facility, the approach, the technical experts on their team; finding the right partner and vetting them is really, really important, because that's where all the work is being done."
Documentation must be finalized before enrollment begins, not after data is in hand. For Kingdom's studies, that typically means a 40- to 60-page protocol covering all biomarkers, statistical analyses, adverse event procedures and logistics, including verification that study product is manufactured to label claim levels at the time of use.
Most critically: review the full protocol with the regulatory team before the study starts. "It is cheaper to adjust on paper than after a study has started," Sheth said. "Reviewing protocol with regulatory before you start is essential."
