It’s Time to Make Ag Data Great Again

How farmers can use quality agriculture data to cut through marketing claims and find the truth

Every season, farmers are inundated with bold claims from agricultural input companies promising bigger yields, better soil and higher profits. If every one of these claims were true, our nation’s fields would be producing yields so impossibly high they’d defy the laws of farming. Yet, year after year, growers face the same challenges: shrinking margins, rising input costs and mounting pressure to make every acre count.

The reality? Many of these promises lack the foundation of high-quality ag data. Even when data is provided, its credibility often raises questions. Was the study statistically sound? Were conditions realistic to your farm? Did the data account for variables like soil type, crop rotation or weather patterns? Far too often, farmers and distributors are left wondering if the ROI will match the marketing.

With so many options on the table, making decisions that deliver real results requires more than flashy brochures and big promises—it requires great data.

Read on to learn the answers to:

• 3 Primary Types of Agriculture Data
• What Makes Ag Data Great?
• What to Look for in the Quality of Ag Product Data

3 Primary Types of Agriculture Data

Great agricultural data depends on the source that it comes from. There are typically three primary sources of trial data for products like biologicals:

1. Trials conducted directly by the agricultural company: These trials are often designed to highlight the product’s performance only under ideal conditions, and may lack the independence needed for unbiased evaluations.
2. Field demos conducted by farmers in a certain area: These real-world demonstrations can give localized insights into product performance under specific conditions. However, they don’t have the same rigor as controlled trials, leading to variability in results.
3. Trials conducted by third-party contract research organizations (CROs) or universities: Independent CROs bring a professional, unbiased approach, with state-of-the-art tools and procedures that ensure reliable results. The trials are randomized, replicated and conducted independently of any agricultural company, minimizing bias. They also provide a detailed analysis, which gives a deeper understanding of how the product performs.

What Makes Ag Data Great?

There are three main factors that impact the quality of agricultural trial data:

1. Data source
2. Trial protocols
3. Statistical significance

Data Source: Independent, Third-Party Trials

Many consider independent, third-party trials as the gold standard for agricultural data. When picking between inputs, farmers and distributors should prioritize products supported by credible CRO data for several reasons:

CRO use professional tools and procedures to ensure accurate, unbiased and reliable ag data.

Contract Research Organizations (CROs) are staffed with professional researchers who specialize in conducting agricultural trials daily. These researchers follow rigorous protocols using state-of-the-art tools and procedures to ensure that every trial is executed with precision and reliability. Their expertise minimizes the likelihood of errors that could compromise the accuracy of trial outcomes, such as improper application methods or inconsistent field conditions. They also operate independently of agricultural companies, meaning results are free from bias.

Application rates and strain selections are rigorously tested and validated in CRO ag data.

When it comes to agricultural inputs, like biologicals, a higher application rate does not always mean a better product. Some biological products promote high application rates without data to support it. The truth is that application rates depend on the Colony Forming Units (CFU) and strain count. Products with higher CFU counts can be applied at lower rates with better performance.

CRO trials identify the right application rates by crop to ensure the product delivers consistent results without overuse or waste. They also evaluate strain selection to confirm that the specific microbes included are optimized for the target crop, soil type and growing conditions.

CRO ag data is consistent across different soil types and conditions.

One of the key strengths of CRO data is its consistency across diverse soil types and environmental conditions. Agricultural fields vary widely—from sandy soils in dry climates to heavy clay in wetter regions. A product that performs well in one location might not deliver the same results in another without rigorous testing.

CRO trials are conducted across multiple locations and soil types to ensure the data reflects real-world farming scenarios. By testing under varied conditions, CROs can demonstrate whether a product consistently delivers results, no matter the soil type, weather patterns, or regional challenges. This level of testing gives farmers confidence that the product will perform well on their specific fields, providing reliable results and predictable outcomes, regardless of the variability in their conditions.

With CROs, every variable is carefully controlled, and every step is meticulously documented, delivering data that farmers can confidently rely on to make informed decisions.

Trial Protocols: Randomized and Replicated

Regardless of the source, great agricultural data starts by setting up a strong foundation during the protocol stage. This happens months before the product trial begins and is typically set in motion by a company’s agronomy department.

As an example, let’s take a look at trial protocols for Locus Agriculture (Locus AG), a biological input company known for its high-quality CRO data. Lead agronomist Dave Dyson describes the process:

“To confirm that data produced during the trial is accurate, Locus AG’s agronomy team designs each trial to be replicated four times and randomized. This ensures the results we achieve are true and not just a “one off” due to an anomalous area in the field.

This is an example of a randomized and replicated Locus AG corn plot design from a CRO in Wisconsin.

When designing the trial protocols for different crops, we perform two types of product testing:

1. A program with several Locus AG biological products
2. Each individual Locus AG biological products separately

This approach allows us to see if only one product is working in the program or if there is a cumulative effect when multiple products are applied during a season.

This protocol reduces trial variability. The less variables there are during a trial, the more accurate the ag data will be.”

Statistical Significance Provides Performance Predictions

Statistics are a powerful way for researchers to make sense of data and predict future outcomes. Statistics is the science of collecting and analyzing data in large quantities for the purpose of making inferences on a whole based on the results from a representative sample. In agriculture, statistics from trial data predict product performance in real-world conditions.

Dave Dyson explains how Locus AG uses statistics to make confident predictions about how our products will perform on a larger scale:

“At Locus AG we use small plot trials and statistics to predict how the product will do in a larger field. There are two important statistical values we look at:

1. P-value is the probability of obtaining results at least as extreme as the observed results. Researchers use different p-values to make statistical significance claims for products that are being tested. At Locus AG we use a p-value of 0.05 when we calculate statistical significance—meaning there is a 95% chance that the outcome of the trial will be repeated.
2. Statistical significance refers to the claim that a result from data generated by testing or experimentation is likely to be attributable to a specific cause. When a result is statistically significant, it means the likelihood of it happening by random chance is very low—less than 5% in Locus AG trials that use a p-value of 0.05 or less. Statistical significance shows that the results of a trial are real and not just due to chance. In agriculture, it means farmers and distributors can trust that a product, like a biological input, consistently improves yield or performance.

Without statistical significance, it’s difficult to trust that the claims a product makes will hold true in real-world applications. Locus AG trials give farmers 95% or higher confidence that the observed benefits, such as increased yield, are consistent and repeatable under similar conditions.

In short, the biological product will deliver as promised.

As Sherlock Holmes declared in A Study in Scarlett, ‘It is a capital mistake to theorize before one has data.'”

Examples of Statistically Significant Agriculture Trials

Locus AG has partnered with CROs to conduct product trails in 20 states across the US, providing a nation-wide view of product performance across many soil types and environmental conditions.

Locus AG has partnered with CROs in 20 states, spread out across the country. This gives a wide view of product performance.

These trial results includes statistically significant yield improvements in eight crops: alfalfa, canola, corn, cotton, potatoes, rice, soybeans and wheat. With this high-quality ag data, farmers can rest assured that the biological products they choose are scientifically proven (with 95% confidence) to deliver the desired results.

Statistical significance means that Locus AG has seen very repeatable results across many trials. This provides a 95% confidence level that Locus AG trial results can be expected by our growers.

What to Look for In Quality Ag Product Data

In conclusion, the old saying “junk data in, junk data out” is as true today as ever. There is so much bad agriculture data available, and it can be a challenge to sort out what is true and what is false.

Three questions to ask to help determine the quality of product data are:

1. What is the source for the ag data? Make sure it was conducted through a professional and unbiased research organization.
2. Was the trial replicated and randomized? It should be. Dig deep into understanding how the trial was put together.
3. Is the data statistically significant, and what p-value was used? Find what p-value the trial is using to calculate their statistical significance. Some trials use all the way up to 0.3, which is much too high (resulting in only 70% confidence).

When we start with great data, we can feel confident in the investments we make on our farm. – Dave Dyson

Have questions about data quality? Want to see more Locus AG biological data?

Contact us to be put in touch with Dave Dyson or one of our sales team members.