In the modern world of farming and food production, companies generate vast amounts of digital records. Every time a sensor measures the temperature of a milk tank, a computer logs the weight of a harvested crop, or a system tracks the movement of a shipment, a digital entry is created. For decades, the assumption has been that these digital records are the same thing as proof. If a computer says a cow was fed, the logic went, then the cow was fed. However, a new study challenges this simple equation. It argues that having a digital record is not the same as having a verifiable claim. Just because a number exists on a screen does not mean an independent observer can look at the original events, follow the steps taken to calculate that number, and confirm that the result is true. This distinction matters deeply for anyone who relies on food companies to be honest about their quality, their sustainability, or their efficiency. If a company claims its digital systems have made its livestock healthier or its processes greener, that claim needs to be backed by a chain of evidence that can be rebuilt from scratch, not just a final number on a dashboard.

Researchers Oksana Makovoz and Sergii Lysenko set out to investigate exactly how well this chain of evidence holds up in the real world. They focused on the agricultural and food industries, a sector where physical events like animal growth or crop yields are increasingly managed by complex digital systems. The team examined twelve different companies, five based in Ukraine and seven international firms, looking at their public reports, sustainability disclosures, and operational data. Their goal was not to count how many computers these companies owned, but to see if the specific claims made in their reports could be independently reconstructed. They asked a simple but difficult question: if an outsider wanted to prove that a specific business claim was true, could they trace it back to the original event, identify who was responsible for it, and understand exactly how the final number was calculated?

The researchers found a clear pattern in the evidence they gathered. In every single case they studied, the companies successfully reported their final results. Whether it was the total amount of milk produced, the number of eggs processed, or the daily weight gain of livestock, these operational outcomes were always visible. However, the story changed when the researchers looked for the steps taken to get there. In three out of four cases, the companies provided some detail about the digital mechanisms or controls they used. But when it came to the most critical piece of the puzzle—clear evidence of who was responsible for the data and whether the records could be audited—the picture became much less clear. In more than half of the cases, the public reports did not contain enough specific information to link a final number directly back to a responsible person or system. While companies often described their general governance and oversight processes, they rarely provided the specific, step-by-step trail needed to rebuild a single claim from the ground up.

The study used a method called sensitivity analysis to test how much of this missing information was a true gap versus just a matter of detail. When the researchers allowed for partial information—such as general descriptions of responsibility rather than specific names—the number of cases with sufficient evidence rose significantly. This suggests that most companies do have governance systems in place, but they often keep them at a high level that does not connect directly to the specific numbers they publish. The difference is subtle but vital: a report might say "we have strict controls," which is a broad statement of intent, but it does not show the specific log of who entered the data and how it was changed to arrive at the final figure. Without that specific link, the claim remains difficult to verify independently.

To understand the practical impact of these findings, the researchers looked at specific examples of how companies reported changes in their performance. One large Ukrainian company, Astarta, reported a thirty-five-point-five percent increase in milk production between 2020 and 2025. The company also documented that it had implemented new digital livestock management systems during this time. While the growth in production was real and measurable, the public records did not provide enough detail to separate the effect of the digital systems from other factors, such as the company simply buying more cows. The researchers concluded that while the digital systems existed and the production grew, it was impossible to prove that the digital systems were the cause of the growth based on the available evidence.

Another case involved a company called KSG Agro, which reported an eight-point-one-nine percent improvement in the daily weight gain of its livestock. This is a positive result that might suggest a successful digital intervention. However, the company's own documentation explicitly stated that the improvement was due to better genetics, not digital technology. In this instance, the researchers found that the evidence chain was strong enough to reject the idea that digital systems were responsible, even though the outcome was positive. This highlights a key finding of the study: a good result does not automatically prove that a digital system caused it. If the evidence chain points to a different cause, such as genetics or weather, that cause must be accepted, regardless of how advanced the company's technology might be.

A third example came from a company named Ovostar, which saw a significant shift in how many of its eggs were processed versus how many were produced. The data showed a clear change in the flow of products, but the public reports did not identify a specific digital mechanism that explained this shift. The researchers found that while the operational change was real, the link to a digital cause was missing from the documentation. In a fourth case, a company called YASENSVIT had a well-documented digital control system for managing water in its poultry houses. The system was clearly designed and the architecture was visible. However, the public records did not include the specific timestamps of when alarms were acknowledged or how quickly the system responded. Without these details, it was impossible to verify if the system actually improved the speed or quality of the response, even though the system itself was real.

These examples illustrate the core problem the study addresses. The researchers argue that organizations often treat digital transformation as a technology adoption problem, assuming that buying the right software will automatically create trust. Instead, they propose that trust depends on a specific organizational capability they call evidence verifiability. This capability is not about having the most advanced technology, but about designing processes that preserve the story of how a number was created. It requires keeping a record of who did what, how the data was changed, and why a specific decision was made. If any part of this chain is missing, the final claim cannot be fully verified, no matter how sophisticated the underlying technology appears.

The study suggests that for companies to truly benefit from digital transformation, they need to change how they think about their records. Instead of just collecting data to produce a final report, they should design their systems to ensure that every claim can be traced back to its source. This means keeping detailed logs of corrections, identifying the specific people or systems responsible for each step, and ensuring that the rules used to calculate results are clear and accessible. The researchers found that when companies do this, they reduce the effort required for outsiders to verify their claims. When they do not, they create a gap where trust must be given without proof.

The findings have important implications for how we view sustainability and quality in the food supply chain. As more companies make claims about their environmental impact or animal welfare, the ability to verify those claims becomes crucial. The study shows that having a digital system is not enough to prove a claim is true. The evidence must be complete, from the original event to the final statement. If a company cannot show the full chain of evidence, its claims remain unverified, regardless of how positive the results look. This does not mean the companies are lying; it simply means that the public record does not contain enough information to prove the story behind the numbers.

Ultimately, the research points to a need for a new approach to managing digital evidence. It suggests that the value of digital records lies not in their volume, but in their ability to tell a complete and reconstructable story. For the companies in the food industry, this means that the next step in their digital journey is not just to gather more data, but to organize it in a way that allows anyone to follow the path from the farm to the final report. By doing so, they can turn simple digital records into defensible evidence that builds genuine trust with consumers, regulators, and partners. The study concludes that without this evidence-by-design approach, the promise of digital transparency will remain just that—a promise, rather than a reality.