Tax Court Denies Majority of Research Credits in Complex Poultry Farming Case, Upholds Penalties on Substantiation Grounds

The $4.5 Million Dispute: How a Poultry Empire’s Research Credits Went Awry

The stakes could not have been higher when the U.S. Tax Court delivered its February 3, 2026, ruling in Gary C. George and Robin A. George v. Commissioner. At issue was a staggering $4.47 million in claimed research credits under Section 41, the federal incentive for increasing research activities, along with $201,319 in accuracy-related penalties. The case pitted the IRS against Gary George, the sole shareholder of George’s of Missouri, Inc. (GOMI), an S corporation in the poultry business, over whether millions in claimed credits for broiler chicken research were valid. The court’s decision did more than resolve a tax dispute; it reasserted the Tax Court’s de novo authority to independently assess research credit claims, free from IRS determinations, and clarified the boundaries of qualified research in the agricultural sector.

The outcome was a partial victory for the Georges. While the court ultimately allowed 6% of substantiated qualified research expenses (QREs); a fraction of the claimed credits; it also rejected the IRS’s blanket disallowance, signaling that the Tax Court will not defer to the agency’s interpretations when the record is incomplete. Most notably, the Georges avoided penalties entirely, with the court finding they had reasonably relied on third-party consultants; a rare win in an era of heightened IRS scrutiny over research credits. The decision underscores the Tax Court’s willingness to exercise its full jurisdiction over S corporation shareholder-level adjustments, a power it has increasingly wielded in recent years to police aggressive credit claims. For taxpayers in research-intensive industries, the case serves as a cautionary tale: meticulous documentation and adherence to the four-part test under Section 41 are non-negotiable.

From Humble Grocery to Poultry Empire: The George Family’s Century-Long Rise

The George family’s poultry dynasty traces its roots to 1922, when C.L. George opened a small country grocery store in Bush Creek, Arkansas. The Great Depression forced him to abandon the struggling store, but rather than retreat, he pivoted to live poultry hauling; transporting chickens to open-air markets in Kansas City, St. Louis, and Chicago. His sons, Gene and Luther, joined the business as they came of age, transforming the operation from a simple hauling service into a vertically integrated poultry producer. By the 1950s, George’s had partnered with a processing plant in Springdale, Arkansas, marking the first step toward the fully integrated model that defines the company today.

The 1960s brought consolidation under Gene George’s leadership, who expanded the business into a commercial production complex encompassing a female hatchery, farms, a processing plant, and even a small egg production operation. His son, Gary George, entered the fold as a teenager, beginning his career in the hatchery before rising through the ranks. After a brief hiatus for college, Gary returned to the family business and, in 1980, assumed the presidency at age 30; becoming the third generation to lead George’s. His tenure was defined by strategic expansion, including the 2001 acquisition of a Virginia-based complex to break into the East Coast retail market. Gary also cultivated critical relationships with fast-food giants like Kentucky Fried Chicken, further cementing George’s as a dominant player in the poultry industry.

By 2012, Gary passed the torch to his twin sons, Carl and Charles, who became co-presidents on their 30th birthdays. Gary remained chairman, overseeing the family’s century-long legacy. Under his leadership, George’s had evolved from a regional live-hauling operation into one of the largest fully integrated poultry producers in the U.S., with operations spanning hatcheries, feed mills, farms, and processing plants. The company’s growth, however, came with complexity: a tangled web of entities; GOMI, George’s Farms, Inc., George’s Processing, Inc., and others; was designed by accountants and lawyers but had little practical impact on day-to-day operations. Employees often worked for one entity while being paid by another, and expenses were frequently shuffled between entities to reflect operational realities rather than legal formalities.

The poultry industry’s challenges shaped the Georges’ business strategy. Tight margins, disease outbreaks like lymphoid leukosis (LT) and coccidiosis, and the industry’s shift toward "no-antibiotic-ever" chicken forced constant innovation. GOMI, the S corporation owned by Gary George, managed the live production chain; from incubation and hatching to transportation for slaughter; while George’s Farms, Inc. handled sales to end customers, and George’s Processing, Inc. operated the processing plants. Despite the byzantine entity structure, the Georges’ focus remained singular: producing high-quality poultry efficiently in an increasingly competitive market.

A Data-Driven Industry: How GOMI Tracked Every Penny and Every Chicken

The poultry industry’s relentless drive for efficiency had long since abandoned guesswork in favor of granular data. GOMI, the Georges’ live production arm, epitomized this evolution, treating each broiler’s lifecycle as a controlled experiment where every metric; from feed conversion to mortality rates; was tracked, analyzed, and optimized. The company’s survival in an era of "no-antibiotic-ever" chicken and razor-thin margins depended on this precision.

The lifecycle of a broiler began in the hatchery, where eggs from GOMI’s breeder hens were incubated under strict protocols. Vaccines were administered in ovo; before hatching; to confer early immunity against diseases like Marek’s virus and Newcastle. Once hatched, chicks were transported to grow farms, where their growth was meticulously monitored. The farms, whether company-related or independent contractor-owned, were not passive hosts but active participants in GOMI’s data ecosystem. Growers recorded daily mortality rates, feed consumption, and weight gain, while GOMI’s service technicians conducted weekly inspections, collecting blood samples to test for disease immunity and performing necropsies on euthanized birds to detect subclinical infections.

Broilers were categorized by target weight: small broilers (four pounds, harvested at 35–37 days) and large broilers (seven to eight pounds, harvested at 60 days). The distinction mattered because feed recipes, housing conditions, and processing logistics varied accordingly. GOMI’s feed mill produced customized rations for each flock, adjusting protein and probiotic levels based on real-time performance data. The goal was to maximize feed conversion; the ratio of feed consumed to weight gained; while minimizing mortality, which could spike due to outbreaks of coccidiosis, necrotic enteritis, or other poultry pathogens.

Disease management was a constant battle. Vaccination schedules were tailored to regional risks, with breeders receiving booster shots to pass immunity to their progeny. Probiotics and acidifiers were routinely added to feed to suppress harmful gut bacteria, while litter management practices; such as leaving shavings between flocks to encourage microbial balance; were designed to reduce pathogen loads. When outbreaks occurred, GOMI’s veterinarians deployed emergency protocols, including targeted antibiotic treatments (though increasingly restricted) or adjustments to ventilation systems to curb heat stress.

The data collected at every stage was compiled into weekly grower reports, which ranked farms by performance metrics like average daily gain and feed efficiency. These reports were not just for monitoring but for compensation: growers were paid based on their flocks’ outcomes, creating a financial incentive to adhere to GOMI’s best practices. The system was ruthlessly efficient, but it left little room for error. A single outbreak of lymphoid leukosis (LT) or a miscalculation in feed ratios could erase weeks of meticulous data tracking, underscoring the industry’s dependence on innovation to stay profitable.

The Research Trials: A Desperate Search for Solutions in a High-Stakes Industry

The relentless pressure to innovate in poultry production pushed GOMI into uncharted territory. As the industry grappled with the loss of critical antibiotics, evolving diseases, and shifting consumer demands, the company launched a series of research trials to test feed additives, vaccines, and genetic lines. These weren’t mere tweaks to existing practices; they were desperate gambles to survive in an industry where a single misstep could mean financial ruin.

The trials were born from necessity. The poultry industry’s dependence on antibiotics was crumbling under regulatory and consumer scrutiny, leaving producers scrambling for alternatives. At the same time, diseases like coccidiosis and lymphoid leukosis (LT) were resurging, threatening flocks that had once thrived under the protection of 3-Nitro and other now-banned treatments. GOMI’s research wasn’t just about improving efficiency; it was about survival.

The first major hurdle was coccidiosis, a parasitic disease that devastates broiler flocks by attacking their digestive systems. The parasite spreads through contaminated feed and litter, creating a cycle of reinfection that GOMI had long struggled to break. The company’s initial focus was Salinomycin, a chemical coccidiostat that disrupts this transmission cycle by killing the parasite before it can be shed in waste. But GOMI’s trials in early 2012 revealed a harsh truth: Salinomycin’s effectiveness waned over time, and its benefits were inconsistent across farms. The company tested higher doses in combination with other coccidiostats like Robenz, but the results were underwhelming. The trials, spread across 12 contracts from January to June 2012, yielded no clear path forward. GOMI considered it a failure.

Undeterred, GOMI turned to HatchPak and Tylan, a vaccine-and-antibiotic combo designed to outmaneuver the parasite’s evolution. HatchPak, a genetically engineered coccidiosis vaccine, replaced the dominant strain in broiler houses with a vaccine strain susceptible to traditional treatments. Tylan, an antibiotic, was added to control necrotic enteritis; a known side effect of HatchPak. The theory was elegant: vaccinate in the hatchery, then administer Tylan in the field to prevent secondary infections. GOMI rolled out the combination in late 2012 and 2013, testing it on 14 contracts in 2012 with adjusted feed expenses of $5.1 million. The initial results were promising; broilers gained 0.1 pound more per bird, and feed conversion improved. Dr. Fussel, GOMI’s veterinarian, called it “worked like a charm.” But by 2013, the tables turned. The same combination, now tested on 20 contracts, produced worse outcomes: declining weight and poorer feed conversion. GOMI blamed the parasite’s evolution, but the trial’s failure left the company back at square one.

The search for alternatives led GOMI to probiotics, a controversial but increasingly necessary tool as antibiotics fell out of favor. The company tested three strains; Floramax, Calsporin, and Sporulin; each marketed as a way to improve gut health and reduce necrotic enteritis. Floramax, administered via water to 45 contracts in 2012 and 2013, showed no measurable difference in feed conversion, weight gain, or mortality compared to control groups. The probiotics were tested alone and in combination with other treatments, including Salinomycin and HatchPak, but the results were inconsistent. GOMI’s veterinarians performed necropsies to check for signs of necrotic enteritis, yet the data revealed no clear benefits. Sporulin and Calsporin fared no better. Despite promising lab results from vendors, the real-world trials failed to deliver. Dr. Greenwood, GOMI’s nutritionist, recommended discontinuing their use.

Another critical challenge was phosphorus digestion. Corn, a staple in broiler diets, contains phosphorus in a form broilers can’t break down. Traditionally, producers added synthetic phosphorus to feed, increasing costs and creating waste management issues. Phytase, an enzyme that unlocks phosphorus in corn, offered a solution; but only if dosed correctly. GOMI’s trials in late 2012 focused on adjusting phytase levels based on feed composition. Dr. Greenwood developed a formula to determine the optimal dosage, walking flocks to check for bone density issues and performing necropsies to look for green bone, a sign of phosphorus deficiency. The trials, conducted on 232 contracts, aimed to reduce production costs, but the results were inconclusive. The company ultimately abandoned phytase, unable to replicate the cost savings promised by vendors.

The existential threat of lymphoid leukosis (LT) loomed largest in the winter months, particularly west of I-49, where outbreaks could wipe out entire flocks. GOMI’s struggle with LT was a case study in trial and error. The company tested two vaccination methods: the industry-standard CEO vaccine, administered via spray, and an unorthodox water-based method that promised more uniform immunity. GOMI’s field technicians mixed the vaccine with water, turned off drinking lines to make broilers thirsty, and then delivered the mixture through water lines. The results were striking: broilers vaccinated via water had more uniform immunity and fewer side effects. GOMI expanded the trials in 2012 and 2013, testing the method on 27 and 111 contracts, respectively. The company also experimented with priming; administering a vector vaccine (HVT–LT) in the hatchery to reduce side effects from the CEO vaccine later. The trials showed promise, but the industry’s reliance on the CEO vaccine meant GOMI’s innovations remained on the fringes.

GOMI’s fight against infectious bursal disease (IBD) led to another round of trials centered on Vaxxitek, a viral vector vaccine designed to provide long-lasting immunity with fewer doses. Before the research years, GOMI had seen dramatic improvements in broiler uniformity after administering Vaxxitek at full dosage. But during the trials, the company questioned whether a lower dose could achieve the same results at a lower cost. GOMI tested Vaxxitek in various combinations; with Floramax, HatchPak, Tylan, and probiotics; across 26 contracts in 2013. The results were mixed. While Vaxxitek initially improved feed conversion and bursa gland size, its effectiveness declined as the IBD virus evolved. GOMI concluded that Vaxxitek worked best at full dosage for one cycle per year, but the trials left lingering doubts about its long-term viability.

The most consequential trial, however, was the Ross 708 genetic line experiment. GOMI’s standard broiler, the Cobb 500, was optimized for small birds but struggled to meet demand for larger cuts. The Ross 708, a genetic line designed for rapid growth, promised higher yields and more breast meat; critical for fast-food chains. In 2014, GOMI conducted a head-to-head trial at the Littrell Broiler Farm, comparing Ross 708 to Cobb 500 under identical conditions. The results were transformative: Ross 708 broilers processed at 8.5 pounds showed an increased margin of $0.021 per pound, translating to over $4 million in annual gains. The trial also revealed trade-offs; higher mortality and slightly worse feed conversion; but the superior yield and quality outweighed the drawbacks. By September 2014, GOMI had switched its large broiler production to the Ross 708.

These trials were more than academic exercises. They were a high-stakes gamble in an industry where margins were measured in fractions of a penny per pound. Each failure meant lost revenue, wasted feed, and compromised flocks. But each success could mean the difference between profit and ruin. GOMI’s research wasn’t just about innovation; it was about survival.

The Research Credit Study: How a Third-Party Consultant Turned Routine Data into Millions in Credits

The trials had paid off; GOMI’s broiler production was now dominated by the Ross 708, a genetic line that promised superior yield and efficiency. But the company’s leadership knew the real work had only just begun. In an industry where even a fraction of a cent per pound could mean the difference between profit and loss, innovation wasn’t optional; it was survival. And survival required more than just better genetics. It required data. Lots of it.

By 2014, GOMI had spent decades refining its operations, tracking everything from feed conversions to flock mortality. But the company’s financial team, led by CFO Gini Driskell; a former Frost PLLC accountant; saw an opportunity beyond the spreadsheets. Frost PLLC, GOMI’s longtime accounting firm, had a suggestion: research credits. Section 41 of the Internal Revenue Code allows businesses to claim a tax credit for increasing research activities, provided the expenses meet a strict four-part test. The credit, worth up to 20% of qualified research expenses (QREs), could offset millions in tax liability. But GOMI had never claimed it before. That’s where alliantgroup came in.

alliantgroup, a tax consulting and lobbying firm with over a thousand employees, specialized in structuring and substantiating research credit claims. The firm had deep experience in the agriculture sector, having worked with poultry producers, grain processors, and livestock operations across the country. Its team included former IRS commissioners, Senate Finance Committee tax counsel, and ex-members of Congress; lobbyists who understood both the tax code and the political landscape. All new hires underwent rigorous training on the nuances of Section 41, and the firm hosted annual updates to keep its consultants current on IRS guidance and court rulings.

GOMI’s leadership, including CEO Gary George, had never heard of alliantgroup before Frost PLLC’s recommendation. But given the firm’s long-standing relationship with the company; Frost PLLC had prepared GOMI’s annual returns for decades and even had a former employee in a key financial role; Gary Greenlit the study without hesitation. The task was delegated to his sons, the co-CEOs of GOMI, who signed an engagement letter with alliantgroup on August 30, 2014. The lead consultant assigned to the project was Jeremy Troutman, an alliantgroup Associate Director with 16 years of experience and roughly 300 research credit studies to his name. Troutman’s specialty? Agriculture.

The study proceeded in three distinct phases. The first was discovery. Troutman and his team conducted in-depth interviews with GOMI’s technical, accounting, and finance employees to identify any activities that might qualify for the credit. Not all clients made it past this stage; some had no qualifying research at all. But GOMI was different. Troutman made two site visits to GOMI’s facilities, speaking with live production managers, accountants, veterinarians, and quality assurance executives. He asked pointed questions: What problems were you trying to solve? How did you test solutions? Did you document failures? The answers would determine whether GOMI’s routine data collection qualified as research under Section 41.

Troutman also requested documents; anything that could shed light on GOMI’s research activities. Kyle Avey, a GOMI service technician and broiler manager, compiled the materials: spreadsheets tracking feed ingredient changes, vaccination records, settlement data from processing plants, and internal reports on flock performance. These documents would form the backbone of the study.

Phase two was substantiation. Troutman’s team zeroed in on the projects that had emerged from the interviews. For 2012 and 2013, the qualifying research included trials on Calsporin (a probiotic supplement), Floramax (a growth promoter), HatchPak (a hatchery management system), LT (a feed additive), Salinomycin (an anticoccidial drug), Sporulin (a vaccine), Tylan (an antibiotic), and Vaxxitek (a viral vaccine). For 2014, the list narrowed to LT, Vaxxitek, and the Ross 708 genetic line. To calculate the credit, Troutman needed to connect each flock to a specific research trial and the associated expenses.

The feed costs were particularly tricky. GOMI’s settlement data included shrink adjustments; the difference between the feed delivered to the farm and the feed actually consumed by the birds. Shrink was a routine accounting adjustment, but under Section 41, only the feed actually consumed in research trials qualified as a QRE. Troutman calculated the average shrink adjustment per year across GOMI’s Virginia farms and removed it from the qualified supply expenses. He also excluded estimated manufacturing overhead costs that GOMI added to its feed expenses. To estimate overhead, he divided the total tons of feed per flock by the cost of overhead at each mill, then removed the resulting expense from the qualified supply calculations.

The numbers added up. Troutman determined that GOMI’s qualified supply expenses were $16,450,745 in 2012, $29,478,367 in 2013, and $17,025,243 in 2014. He did not calculate any qualified service expenses; wages for employees directly involved in research; because the effort required to allocate those wages wasn’t worth the potential credit. Instead, he focused on supply expenses, which were easier to trace to specific trials.

To calculate the base amount; the average research expenses from 2009 to 2011; Troutman used a ratio. He found that qualified supply expenses averaged 10.23% of total expenses during the research years. He applied this ratio to the total supply expenses from 2009 to 2011 to estimate the base amount for 2012. For 2013, he used the actual qualified supply expenses from 2012. For 2014, he used the actual expenses from 2012 and 2013. The calculations were checked by two additional alliantgroup employees, and Frost PLLC verified the financial data.

By the end of phase two, alliantgroup had enough information to prepare pro forma Forms 6765, the IRS form used to claim the research credit. These forms would allow GOMI to report the credits on its tax returns.

The final phase was documentation. alliantgroup memorialized its findings in two undated written reports, detailing the research trials, the methodology for calculating qualified supply expenses, and the connection between each flock and the corresponding trial. Troutman spent between 700 and 800 hours compiling the reports, ensuring every claim was backed by GOMI’s internal data. The documents were designed to withstand IRS scrutiny; a critical consideration, given the agency’s growing skepticism of research credit claims, particularly in industries like agriculture where routine activities could easily be misclassified as research.

The Dispute: IRS vs. Petitioners on What Counts as Qualified Research

The George Family’s poultry empire, GOMI, claimed nearly $4.5 million in research credits under Section 41 over six tax years, arguing that its systematic trials to improve broiler health, feed efficiency, and gut health qualified as qualified research. The IRS, however, challenged the legitimacy of these claims, setting the stage for a fundamental disagreement over what constitutes qualified research under the tax code.

The dispute hinged on whether GOMI’s activities met the four-part statutory test under Section 41(d), which requires research to satisfy: (1) a business component test, (2) a technological information test, (3) a Section 174 test, and (4) a process of experimentation test. The IRS contended that GOMI’s trials were either routine data collection or adaptations of existing business components, while the petitioners argued that the trials involved systematic experimentation to resolve technological uncertainty in poultry production.

At the heart of the disagreement was Section 41(d)(1)(A), which defines qualified research as research "with respect to which expenditures may be treated as expenses under Section 174." Section 174, in turn, allows deductions for research and experimental expenditures that are incident to the development or improvement of a product; but only if the research is experimental in nature and undertaken to eliminate uncertainty. The IRS took the position that GOMI’s trials; such as testing different feed formulations or monitoring flock health metrics; were ordinary production activities rather than experimental research, pointing to Treasury Regulation § 1.174-2(a)(1), which excludes routine quality control or data collection from qualifying expenditures.

The petitioners countered that their trials were aimed at discovering technological information under Section 41(d)(1)(B)(i), which requires research to be "undertaken for the purpose of discovering information which is technological in nature." They argued that their work relied on principles of poultry nutrition, immunology, and biostatistics; fields that fall within the ambit of hard sciences as defined by Treasury Regulation § 1.41-4(a)(4). The IRS dismissed this claim, asserting that the trials were merely adaptations of existing business components, such as refining feed formulas that were already in use, and thus failed the business component test under Section 41(d)(2)(B).

A critical sub-issue involved feed expenses, which GOMI claimed as qualified supply expenses under Section 41(b)(2)(A)(ii). The petitioners argued that these expenses were pilot model costs under Section 174, deductible because they were incurred to evaluate and resolve uncertainty in feed formulations. The IRS, however, maintained that the feed expenses were ordinary production costs under Section 263A(c)(2), not experimental expenditures, and thus ineligible for deduction under Section 174 or inclusion in the research credit calculation.

The IRS also challenged the substantiation of GOMI’s research credit study, arguing that the third-party consultant’s report was post hoc; compiled years after the trials were conducted; and lacked contemporaneous documentation of the scientific method. The petitioners countered that the report, while compiled later, was based on GOMI’s internal data and reflected the systematic methodology used in the trials, including controlled flock studies, feed trials, and health monitoring.

The dispute over qualified research thus boiled down to competing interpretations of Section 41’s four-part test, the scope of Section 174’s experimental expenditures, and the adequacy of documentation to substantiate the claims. The IRS’s position reflected a broader skepticism toward research credit claims in industries like agriculture, where routine activities can easily be misclassified as research. The petitioners, meanwhile, argued that their trials were far from routine, involving systematic experimentation to address technological uncertainty in poultry production; a claim that would hinge on whether the Tax Court viewed their activities as qualified research under the tax code’s stringent standards.

The Court’s Analysis: Why Most of GOMI’s Research Trials Failed the Four-Part Test

The Tax Court’s ruling hinged on whether GOMI’s research trials met the four-part statutory test under § 41(d)(1); a framework designed to separate genuine innovation from routine business activities. Before dissecting each trial, the court first clarified the legal standards at play, ensuring taxpayers understood the bar they had to clear.

Section 41(d)(1) requires research to satisfy four distinct tests: (1) the business component test, (2) the technological information test, (3) the § 174 test (permitted purpose), and (4) the process of experimentation test. The court emphasized that these tests are not cumulative; each must be satisfied independently. Failure in any one category disqualifies the research, regardless of how well the others are met.

The court also addressed § 174, which governs the deductibility of research expenditures. Under Treas. Reg. § 1.174-2(a)(1), research must be experimental in nature and intended to resolve technological uncertainty. The 2014 amendments to § 174 expanded the definition of a pilot model; a full-scale or small-scale prototype used to test design concepts under real-world conditions. The court applied this definition rigorously, rejecting claims where research was merely routine testing or adaptation of existing products.

Finally, the court scrutinized the shrinking-back rule under Treas. Reg. § 1.41-4(b)(2), which allows taxpayers to exclude non-qualifying portions of a research project while retaining the credit for qualifying components. However, the court made clear that this rule requires meaningful substantiation; taxpayers cannot rely on vague assertions to carve out portions of their research.

---

Salinomycin: A Case of Unsubstantiated Investigatory Activities

The Salinomycin trials failed the § 174 test because GOMI could not demonstrate that it undertook investigatory activities to resolve technological uncertainty. Petitioners argued that GOMI was uncertain whether higher dosages of Salinomycin, combined with chemical coccidiostats, would improve broiler health by controlling coccidiosis. However, the court found that GOMI’s contemporaneous feed records contradicted this claim.

The feed recipes showed that Salinomycin was administered at the same dosage before, during, and after the alleged trials, with no evidence of dose variation or addition of chemical coccidiostats to the feed. The court noted that Robenz (a chemical coccidiostat) was only added to one feed recipe dated April 3, 2012; too late to be relevant for most of the experimental flocks. Without corroborating documentation, the court could not conclude that GOMI conducted investigative activities to resolve the alleged uncertainty.

The court also rejected GOMI’s argument for applying the shrinking-back rule, stating that petitioners provided no meaningful basis for dissecting the trials into qualifying and non-qualifying components. The Salinomycin trials were denied in their entirety.

---

HatchPak and Tylan: Qualified for 2012, Not for 2013

The HatchPak and Tylan trials met the four-part test for 2012 but failed for 2013, highlighting the court’s strict interpretation of the § 174 uncertainty requirement.

Business Component Test

The court agreed with petitioners that the business component was an improved poultry product; specifically, broilers with better gut health and resistance to coccidiosis. The court rejected the IRS’s argument that the trials were merely process-related, finding that GOMI’s goal was to enhance the quality of its standard broiler, not just reduce costs.

Technological Information Test

The court found that GOMI relied on biological sciences (health monitoring, necropsies) to study broiler performance, satisfying the technological information test.

§ 174 Test: Uncertainty Resolved by 2013

The court held that 2012 trials met the § 174 test because GOMI faced objective uncertainty about whether HatchPak and Tylan would effectively control coccidiosis under its standard production process. The court relied on expert testimony from Dr. Bobeck and Dr. Johnson, who agreed that vendor research (conducted in sterile, small-scale conditions) does not translate to commercial-scale poultry production. GOMI’s 2012 trials on 14 flocks provided definitive data that the combination worked in its standard production process, resolving the uncertainty.

However, the 2013 trials failed the § 174 test because the uncertainty was resolved by the end of 2012. The court cited Siemer Milling Co. v. Commissioner (2019), which held that subsequent failures do not revive uncertainty if prior tests provided objective data. Since GOMI’s 2012 trials established that HatchPak and Tylan worked, the 2013 trials were not research; they were routine testing of a known solution.

Process of Experimentation Test

The court found that GOMI’s 2012 trials followed a systematic process:

1. Hypothesis formation: HatchPak + Tylan would improve gut health.
2. Testing: Administered to 14 flocks, with data collection on mortality and illness.
3. Analysis: Compared results to historic data, confirming improved outcomes.

The court rejected the IRS’s argument that the lack of a contemporaneous control group disqualified the trials. Instead, the court noted that GOMI’s historic performance data served as a de facto control group, and the industry’s data-driven nature justified this approach.

Exclusions Under § 41(d)(4)

The court rejected the IRS’s argument that the trials were routine quality control or adaptation of an existing product. The court distinguished the trials from Treas. Reg. § 1.41-4(a)(8) (Example 2), where a paint manufacturer tested a nozzle after selecting a new paint color. Unlike that example, GOMI’s uncertainty was not resolved by vendor research; it required applied testing under real-world conditions.

---

Probiotics: Qualified Trials, But Floramax Substantiation Failed

The probiotic trials met the four-part test, but the court disallowed credits for Floramax due to lack of substantiation.

Business Component & Technological Information Tests

The court agreed that the business component was an improved poultry product (healthier broilers with better gut health and reduced antibiotic reliance). The trials relied on biological sciences (health monitoring, necropsies), satisfying the technological information test.

§ 174 Test: Uncertainty Resolved Through Applied Testing

The court found that GOMI faced objective uncertainty about whether probiotics would improve broiler health under its standard production process. Unlike HatchPak and Tylan, the probiotics were not commercially proven in GOMI’s specific environment. The court cited Union Carbide Corp. v. Commissioner (2009), holding that applied research is required when vendor data does not translate to real-world conditions.

Process of Experimentation Test

GOMI’s trials followed a scientific method:

1. Hypothesis: Floramax would improve gut health.
2. Testing: Administered to flocks, collected health data.
3. Analysis: No noticeable improvement → revised hypothesis to test Sporulin.
4. Repeat: Tested Sporulin, then Calsporin, with no success.

The court rejected the IRS’s argument that the trials were merely evaluating available products. Unlike Siemer Milling Co. (2019), where the taxpayer failed to document a process of experimentation, GOMI’s systematic trial-and-error met the standard.

Exclusions & Substantiation Failures

The court rejected the routine testing exclusion, finding that GOMI’s additional data collection and analysis went beyond its standard production monitoring. However, the court disallowed Floramax credits because petitioners failed to identify the experimental flocks. Unlike the HatchPak and Tylan trials, there was no special feed label or contemporaneous documentation linking Floramax to specific flocks. The court refused to estimate the flocks, stating that speculation is not sufficient.

---

Phytase: No Evidence of Investigatory Activities

The Phytase trials failed the § 174 test because GOMI’s contemporaneous feed records showed no variation in dosage during the alleged trials. The feed logs revealed that Phyzyme TPT 2500 was added at a constant 0.4 pound per ton from 2010 onward, contradicting petitioners’ claim that Dr. Greenwood varied dosages to determine the most effective amount.

The court also noted that March 2012 lab tests on phosphorus levels could not be connected to any investigatory activities, as the dosages did not change. Without evidence of experimentation, the Phytase trials were denied entirely.

---

LT (Method of Administration): Unsubstantiated Timeline

The LT method of administration trials (2012–2013) were denied due to lack of substantiation. Witness testimony could only narrow the timeline to 2012 or 2013, with no contemporaneous documentation to support the trials. The court stated that vague assertions about when the trials occurred cannot substitute for records, leaving the court unable to apply the four-part test.

---

LT Priming (2014): Qualified, But Limited Substantiation

The LT priming trials (2014) met the four-part test, but the court limited the credit to six flocks due to inadequate substantiation.

Business Component & Technological Information Tests

The court agreed that the business component was an improved poultry product (healthier broilers with fewer side effects from the CEO vaccine). The trials relied on biological sciences (vitals monitoring, field surveys), satisfying the technological information test.

§ 174 Test: Uncertainty Resolved Through Priming

GOMI faced objective uncertainty about whether priming broilers with HVT-LT vaccine would reduce the harsh side effects of the CEO vaccine. The court rejected the IRS’s argument that prior priming of breeders resolved the uncertainty, noting that broilers and breeders are drastically different in age, purpose, and vaccine response.

Process of Experimentation Test

GOMI’s trials followed a systematic process:

1. Hypothesis: Priming would reduce CEO vaccine side effects.
2. Testing: Administered HVT-LT in ovo, then CEO vaccine in the field.
3. Analysis: Compared side effects to historic data, confirming reduced reactions.

The court found that the substantially all test was met, as all activities related to the trials were part of the process of experimentation.

Substantiation Limitations

The court limited the credit to six flocks because:

• Only one contemporaneous document (a vaccine protocol dated September 2, 2014) identified the experimental flocks.
• The protocol applied to big broilers west of I-49 and north of I-40, but settlement data showed that only six flocks met this criteria.
• The remaining 18 flocks were small broilers, with no documentation linking them to the trials.

---

Vaxxitek: No Evidence of Dosage Variation

The Vaxxitek trials were denied because petitioners failed to substantiate:

1. When the trials occurred (witnesses provided three different timeframes: 2012, 2014, or 2012–2014).
2. Which flocks were experimental (no special feed labels, vaccination records, or contemporaneous data linked Vaxxitek to specific flocks).
3. Whether dosages varied (invoices referenced full-dose administration, not experimental variations).

The court stated that guesswork cannot replace documentation, leaving the Vaxxitek trials completely disallowed.

---

Ross 708: The One Trial That Qualified

The Ross 708 genetic line trial was the only trial fully substantiated by GOMI. The court found that:

• The business component was an improved poultry product (healthier, more uniform broilers).
• The trials relied on biological sciences (health monitoring, performance tracking).
• GOMI faced objective uncertainty about whether the Ross 708 genetic line would perform better under its standard production process.
• The trials followed a systematic process of experimentation, with hypothesis testing, data collection, and analysis.

The court allowed the credit in full, noting that GOMI’s contemporaneous documentation (feed records, performance data, and expert testimony) clearly substantiated the trial.

---

The Court’s Power Over the IRS: A Message to Taxpayers

This ruling underscores the Tax Court’s willingness to reject IRS positions when taxpayers fail to meet their burden of proof. The court rejected the IRS’s argument that GOMI was the wrong entity to claim the credits, finding that GOMI bore the economic burden of the research. It also rejected the IRS’s attempt to link supply expenses to wage expenses, clarifying that § 41(b)(2)(A) does not require wage expenses to be claimed to qualify supply expenses.

However, the court exercised its authority to deny credits where documentation was lacking, even for trials that arguably met the four-part test. This sends a clear message: Taxpayers must document every step of their research; from hypothesis formation to data analysis; or risk losing the credit entirely.

For future taxpayers, this case serves as a cautionary tale: Routine activities cannot be rebranded as research, and third-party consultants cannot substitute for contemporaneous records. The Tax Court’s scrutiny of agricultural research credits is intensifying, and only those with rigorous documentation will prevail.

Pilot Models and Feed Expenses: How the Court Expanded § 174 for Agricultural Research

The Tax Court’s ruling in George’s of Missouri, Inc. v. Commissioner marks a rare and consequential expansion of Section 174, the provision governing research and experimental expenditures. While most § 174 disputes focus on laboratory settings or high-tech industries, this case squarely addressed agricultural research; specifically, whether pilot model expenses and feed costs for experimental broilers qualified under the statute. The court’s reasoning not only validated these expenses but also asserted broad interpretive authority over § 174, signaling to taxpayers and the IRS alike that agricultural innovation can; and should; be treated with the same rigor as traditional R&D.

The dispute centered on seven research trials conducted by GOMI between 2012 and 2014, all aimed at developing an “improved poultry product.” The IRS argued that these trials were routine data collection masquerading as research, while GOMI claimed they qualified for the research credit under Section 41 and Section 174. The court’s analysis hinged on whether the trials met the experimental requirements of § 174, particularly the treatment of pilot models and feed expenses as research expenditures.

Pilot Models: A Novel Interpretation of § 174

The court’s most significant departure from prior precedent came in its treatment of pilot models under Treasury Regulation § 1.174-2(a)(4), which defines a pilot model as a full-scale or small-scale model used to test design concepts or evaluate performance under real-world conditions. GOMI argued that its broiler flocks; raised under controlled conditions to test new feed formulations, vaccination strategies, and housing designs; constituted pilot models. The IRS countered that these were ordinary production activities, not experimental.

The court sided with GOMI, holding that broiler flocks raised for experimental purposes qualified as pilot models under § 174. This interpretation broadened the definition of a pilot model to include agricultural products, a departure from traditional applications in manufacturing or software development. The court emphasized that Treasury Regulation § 1.174-2(a)(11) (Example 7), which allows for the inclusion of feed expenses in pilot model testing, directly supported GOMI’s position. In that example, the IRS explicitly permits costs associated with feeding experimental animals as part of a pilot model’s development, provided the animals are used to test nutritional hypotheses or refine feed formulations.

This ruling exercises the Tax Court’s interpretive power in a way that expands § 174’s reach into agriculture, a sector often overlooked in R&D discussions. By treating broiler flocks as pilot models, the court effectively elevated routine poultry farming into the realm of qualified research, a move that could have far-reaching implications for farmers, feed manufacturers, and agricultural innovators.

Feed Expenses: The Court’s § 174 Green Light

The court’s treatment of feed expenses as research expenditures under § 174 was equally groundbreaking. GOMI incurred significant costs feeding broilers in experimental trials, where different feed formulations were tested to optimize growth rates and disease resistance. The IRS argued that these expenses were ordinary production costs, not research expenditures, because the broilers were ultimately sold for commercial purposes.

The court rejected this argument, relying on Treasury Regulation § 1.174-2(a)(1), which permits the deduction of research and experimental expenditures even if the resulting product is later sold. The court held that feed used in experimental trials qualified as a research expense because it was directly tied to the testing of a new feed formulation; a process designed to eliminate uncertainty about nutritional efficacy. This reasoning distinguished pilot model expenses from ordinary production costs, clarifying that experimental feed costs are deductible under § 174 even when the broilers themselves are sold commercially.

This distinction is critical for agricultural taxpayers. The court made clear that § 174 does not require the abandonment of experimental products; rather, it allows deductions for costs incurred during the research phase, regardless of whether the final product enters the market. This interpretation aligns with the statute’s purpose; to incentivize innovation; by ensuring that agricultural research is not penalized simply because it produces a marketable commodity.

Implications for Future Taxpayers

The court’s ruling in George’s of Missouri signals a new era for agricultural research credits, one where pilot models and feed expenses are no longer automatically dismissed as routine production. For taxpayers in agriculture and other industries where physical products are developed, this case provides clear guidance:

1. Pilot models need not be confined to laboratories; agricultural flocks, crop trials, or livestock herds can qualify if they are used to test design concepts or refine products.
2. Feed expenses (or similar input costs) can be research expenditures if they are part of an experimental process, even if the final product is sold commercially.
3. Documentation remains paramount; taxpayers must demonstrate the experimental nature of their activities, including hypotheses, testing methods, and results, to withstand IRS scrutiny.

The Tax Court’s assertive interpretation of § 174 in this case challenges the IRS’s traditional skepticism toward agricultural research credits. By expanding the statute’s scope, the court has empowered farmers, feed producers, and agricultural innovators to claim credits for activities that were once considered outside the realm of qualified research. However, the ruling also serves as a cautionary tale: rigorous documentation is non-negotiable. Taxpayers who fail to distinguish experimental activities from routine production risk seeing their claims disallowed; just as GOMI nearly did.

For now, the poultry industry; and agriculture more broadly; has a new tool for innovation funding. But the IRS may not be pleased with the court’s broad reading of § 174, and future disputes are likely. Taxpayers would be wise to lean on this precedent while ensuring their own research activities meet the experimental standards the court has now codified.

Substantiation Failures: Why the Court Rejected GOMI’s Post Hoc Research Credit Study

The Tax Court’s skepticism of GOMI’s research credit claims was rooted in a single, unassailable principle: post hoc substantiation is no substitute for contemporaneous records. Section 6001 requires taxpayers to maintain records sufficient to establish the amount of deductions claimed, and the court made clear that this obligation is not satisfied by retrospective studies assembled years after the fact. Treas. Reg. § 1.6001-1(a). For research credits under § 41, the Treasury Regulation further mandates that taxpayers retain records "in sufficiently usable form and detail to substantiate that the expenditures claimed are eligible for the credit." Treas. Reg. § 1.41-4(d). The court has repeatedly applied the Cohan rule; allowing reasonable estimates when records are incomplete; but it has drawn a hard line: no estimate is permissible when the taxpayer provides no evidence at all permitting an informed calculation. Reinke v. Commissioner, 46 F.3d 760 (8th Cir. 1995); Shami v. Commissioner, 741 F.3d 560 (5th Cir. 2014).

GOMI’s research credit study, prepared by alliantgroup years after the alleged research activities, fell far short of these standards. The court rejected the study’s estimates for base year qualified research expenses (QREs) from 2009 to 2011, finding that GOMI had failed to introduce any contemporaneous documentation to substantiate the claimed activities. The study attempted to reconstruct QREs by applying a 10.23% ratio of feed expenses from later years, but the court dismissed this as speculative. The court emphasized that even under the Cohan rule, there must be "sufficient evidence in the record to provide a basis upon which an estimate may be made." Mendes v. Commissioner, 121 T.C. 308 (2003). GOMI’s reliance on vague board minutes and employee recollections; such as references to genetic line trials or ventilation improvements; was insufficient to establish that these activities met the four-part statutory test for qualified research. The court refused to "wing it" with an estimate ungrounded in the record, concluding that GOMI had no QREs for 2009, 2010, and 2011. This ruling underscores the court’s insistence that taxpayers must maintain contemporaneous records documenting the nature, purpose, and experimental process of their research activities.

The court’s rejection extended to GOMI’s attempts to substantiate QREs for specific research trials, such as the Salinomycin, Vaxxitek, and LT method of administration projects. For the Salinomycin trials, the court found that GOMI’s contemporaneous feed recipe records contradicted the claim that experimental dosages were tested. The records showed that GOMI continued to use the same dosage of Salinomycin before, during, and after the alleged trials, and the court refused to rely on employee testimony that conflicted with the documentary evidence. Similarly, for the Vaxxitek trials, the court noted the absence of any contemporaneous records linking specific flocks to varying dosages of the vaccine. Witness testimony provided three conflicting timeframes for the trials, and the court could not identify which flocks were experimental or whether varying dosages were administered. The court concluded that GOMI had failed to substantiate these trials, leaving no basis for an estimate under the Cohan rule.

The LT method of administration trials presented another example of post hoc reconstruction failing to meet the court’s standards. GOMI claimed research credits for trials conducted in 2012 and 2013, but the court found no contemporaneous documentation to establish when the trials occurred, which flocks were involved, or whether the trials were part of a systematic process of experimentation. The court noted that the lack of a timeline or documentation made it impossible to apply the four-part statutory test, as required by § 41. Without contemporaneous records, the court could not determine whether the trials were conducted to resolve uncertainty or were merely routine adjustments to existing vaccination protocols.

The court’s refusal to accept alliantgroup’s estimates for base year QREs and its rejection of GOMI’s post hoc substantiation of specific trials demonstrate the Tax Court’s growing skepticism of third-party consultants’ retrospective studies. While the court has acknowledged that taxpayers need not maintain records in any particular form, Fudim v. Commissioner, T.C. Memo. 1994-235, it has consistently held that the records must be contemporaneous and sufficiently detailed to substantiate the claimed activities. The court’s application of the Cohan rule is not a license for speculation; it is a limited exception reserved for cases where the taxpayer provides a reasonable basis for an estimate. GOMI’s failure to meet this standard resulted in the disallowance of millions in claimed research credits, a cautionary tale for taxpayers relying on consultants to reconstruct research activities years after the fact.

The Penalties: How Reliance on Third-Party Consultants Saved the Day

The George family’s poultry empire nearly lost its $4.5 million research credit claim to penalties, not just because of the IRS’s disallowance of the credits themselves, but because of the agency’s insistence that the underpayments stemmed from negligence or substantial understatement. The IRS argued that the petitioners’ reliance on third-party consultants; alliantgroup and Frost PLLC; was insufficient to avoid penalties under Section 6662(a), which imposes a 20% accuracy-related penalty on underpayments attributable to negligence or disregard of rules or regulations. The court, however, rejected the IRS’s position and instead sided with the petitioners, finding that their reliance on professional advice provided reasonable cause and good faith under Section 6664(c)(1).

The IRS’s argument hinged on the fact that GOMI’s research credit claims failed to meet the stringent substantiation requirements outlined in prior sections of the opinion. The agency contended that the petitioners’ reliance on alliantgroup’s study was misplaced, as the study itself was flawed and did not adequately document the scientific method required for qualified research under Section 41. The IRS further argued that the petitioners’ failure to maintain contemporaneous records; particularly given the post hoc nature of the research credit study; demonstrated negligence, as defined in Section 6662(c) and Treas. Reg. § 1.6662-3(b)(1), which includes “any failure to make a reasonable attempt to comply” with the Code or to “keep adequate books and records or to substantiate items properly.”

The court, however, exercised its authority to distinguish this case from prior precedents where reliance on consultants was deemed unreasonable. The IRS pointed to Betz v. Commissioner, a 2023 Tax Court Memorandum decision, where the court found that reliance on alliantgroup was inconsistent with ordinary business care and prudence. The court in Betz noted in a footnote that the taxpayers’ reliance on alliantgroup’s advice did not constitute reasonable cause. But the George family’s case, the court held, was different. It emphasized that Section 6664(c)(1) does not impose a blanket rule against reliance on consultants; rather, it requires a case-by-case analysis of whether the taxpayer acted with reasonable cause and in good faith. The court’s willingness to carve out an exception here; despite the substantiation failures; demonstrates its power to interpret the reasonable cause exception flexibly, particularly when the taxpayer demonstrates a genuine effort to comply with the law.

The petitioners’ reliance on alliantgroup and Frost PLLC met the three-prong test for reasonable cause under Treas. Reg. § 1.6664-4(c)(1). First, alliantgroup was a competent professional with sufficient expertise to justify reliance. The record showed that alliantgroup had over a decade of experience in conducting research credit studies, including for clients in the agriculture industry, and employed professionals with tax policy and legal backgrounds. Second, the petitioners provided alliantgroup with necessary and accurate information. George’s, the entity with the relevant records, granted alliantgroup open access to its books, data, and employees, including C-suite executives and field service technicians. The data provided extended beyond the alleged research trials, including settlement data and feed expenses for all flocks in the research years. Third, the petitioners actually relied in good faith on alliantgroup’s advice. The study spanned three years, produced nearly 100 pages of analysis, and culminated in pro forma Forms 6765 that were filed with the amended returns. The court rejected the IRS’s argument that the timing of the amended returns undermined reliance, noting that alliantgroup was in constant contact with Frost PLLC and George’s, providing qualification information in advance of the final reports.

The court further distinguished this case from Betz by highlighting that the petitioners’ reliance was not merely on a consultant’s report but on a collaborative effort that included Frost PLLC, a firm with deep knowledge of George’s business operations. The court also noted that Betz was decided after the petitioners had already relied on alliantgroup, whereas Suder v. Commissioner, a 2014 Tax Court Memorandum decision, had found reliance on alliantgroup to be reasonable and in good faith. The court’s willingness to apply Suder retroactively underscores its authority to shape the reasonable cause exception based on the specific facts of each case, rather than adhering to a rigid rule.

The IRS’s attempt to introduce expert testimony to challenge the good faith of the petitioners’ reliance was also rejected. The court found Dr. Bobeck’s opinion; that the research credit studies lacked sufficient detail to show a process of experimentation; irrelevant to the reasonable cause analysis. The court emphasized that the research credit is one of the most complex provisions in the Code, and Gary George’s lack of expertise in tax and business law justified his reliance on professionals. The court’s refusal to defer to the IRS’s expert further demonstrates its independence in evaluating the facts and law, even when the IRS presents seemingly compelling evidence.

Ultimately, the court’s decision to waive penalties under Section 6662(a) was not a concession to the petitioners’ substantiation failures but a recognition that their reliance on competent professionals; despite the IRS’s objections; met the reasonable cause exception. The court’s exercise of authority here is notable: it refused to penalize the petitioners for the IRS’s preferred interpretation of the substantiation requirements, instead prioritizing the taxpayer’s good faith reliance on expert advice. This ruling sends a clear message to taxpayers and the IRS alike: while the substantiation bar for research credits remains high, the reasonable cause exception under Section 6664(c)(1) can serve as a safety net for those who seek and rely on professional guidance in good faith.

The Bottom Line: What This Means for Taxpayers Claiming Research Credits

The Tax Court’s ruling in George Family Poultry, Inc. v. Commissioner delivers a mixed verdict that simultaneously tightens the screws on research credit substantiation while carving out a pragmatic path for taxpayers who play by the rules. Petitioners secured reduced research credits; just 6% of substantiated Qualified Research Expenses (QREs) for tax years 2012 through 2014; but avoided penalties entirely, a result that underscores the court’s willingness to apply the reasonable cause exception under Section 6664(c)(1) when taxpayers demonstrate good faith reliance on professional advice. This outcome signals a critical shift: while the IRS continues to aggressively challenge research credit claims, the court is not reflexively imposing penalties when taxpayers act prudently.

The decision reaffirms the primacy of the four-part statutory test under Section 41(d)(1); a framework that demands research relate to a new or improved business component, rely on hard sciences, eliminate technological uncertainty, and follow a process of experimentation. The court’s rejection of GOMI’s post hoc research credit study; prepared years after the fact; serves as a stark warning: contemporaneous documentation is non-negotiable. Taxpayers cannot retroactively construct qualifying research from routine data collection; the IRS and the court will scrutinize whether experiments were designed, conducted, and documented in real time. This principle extends to pilot models under Section 174, where the court embraced a broad interpretation of qualifying expenditures in agricultural research, but only when tied to experimental testing of new designs or methodologies. For poultry producers and other agricultural businesses, this means that pilot barns, feed trials, or vaccine strain tests may qualify as Section 174 expenses; but only if they are part of a systematic effort to resolve technological uncertainty, not routine production adjustments.

The ruling also highlights the dual role of third-party consultants. While the court refused to penalize the petitioners for relying on expert guidance; even when the IRS disputed the methodology; the decision makes clear that consultant reports alone are insufficient without underlying primary evidence. Taxpayers who outsource their research credit claims must ensure their consultants’ work is grounded in contemporaneous records, lab notes, prototypes, and failure reports. The reasonable cause exception under Section 6664(c)(1) may shield taxpayers from penalties, but it does not immunize them from disallowed credits.

For practitioners and taxpayers, the takeaways are actionable and urgent. First, institutionalize contemporaneous documentation: every hypothesis, test, and result must be recorded as it happens, not reconstructed later. Second, separate qualifying research from routine activities using the shrinking-back rule, ensuring that only the experimental components of a project are claimed. Third, engage qualified consultants early, but treat their work as a supplement to; not a substitute for; your own records. Finally, prepare for IRS scrutiny: the court’s willingness to uphold penalties in other cases (where taxpayers lacked documentation or relied on flawed methodologies) means that aggressive claims will face heightened risk.

In an era where the IRS has launched compliance campaigns targeting research credits; particularly in agriculture, software, and manufacturing; the George Family Poultry decision offers both a caution and a lifeline. Taxpayers who meet the statutory requirements with ironclad documentation can still prevail, but those who cut corners will find the court’s patience; and the IRS’s appetite for enforcement; nonexistent. The message is clear: innovation deserves incentives, but sloppiness invites penalties.