An email arrived in February from a patient I have corresponded with several times. She had purchased the same tincture — same brand, same label, same dispensary — three months apart. The first bottle helped her sleep reliably within forty-five minutes. The second bottle, she wrote, “does almost nothing, and I'm wondering if they changed the formula or if something is wrong with me.”
Nothing was wrong with her. Something was different about the product. And the explanation, while unsatisfying, is important to understand: cannabis is not ibuprofen. Not yet, anyway.
Why the Same Product Can Be Different
Pharmaceutical manufacturing operates on the principle of batch-to-batch consistency. When I worked in quality control, our acceptable variance for active ingredient concentration was typically plus or minus 5% of label claim, verified by validated analytical methods before any lot shipped. That level of consistency is achievable because pharmaceutical ingredients are synthesized or purified to exacting specifications under controlled conditions.
Cannabis is a plant. Plants are, by their nature, variable. Even under controlled indoor cultivation, the chemical profile of a cannabis harvest is influenced by factors that are difficult to standardize completely:
Genetics and phenotypic expression. Even clones from the same mother plant can express slightly different cannabinoid and terpene profiles depending on their position in the grow room, light exposure, and microenvironment. A 2020 study from Wageningen University (published in Phytochemistry) documented measurable chemical variation among genetically identical cannabis plants grown under ostensibly identical conditions.
Growing conditions. Temperature fluctuations, humidity, light spectrum, nutrient concentration, and irrigation timing all influence cannabinoid and terpene biosynthesis. A grow room running two degrees warmer during flowering may produce flower with a different terpene ratio than the previous cycle. The plant is responsive to its environment in ways that a chemical synthesis is not.
Harvest timing. The cannabinoid profile shifts during the final weeks of flowering. Harvest a week early and THCa may be lower while precursor cannabinoids remain elevated. Harvest a week late and some THC may have degraded to CBN. The window matters, and it is not always hit precisely.
Post-harvest processing. Drying speed, curing duration, storage temperature, and light exposure all affect the final product. Terpenes are volatile — they evaporate. A batch dried quickly at higher temperatures may lose monoterpenes that a slower-cured batch retains. For extracts, the extraction parameters, solvent ratios, and purification steps introduce additional variables.
What Is Normal Variation vs. a Quality Problem
This is the question that matters, and it requires some honesty about where the cannabis industry currently stands.
Normal variation means a product's THC content shifts from, say, 21.3% in one batch to 19.8% in the next, with a similar but not identical terpene profile. The experience may be subtly different. This is the biological reality of a plant-derived product and, within reason, it is expected.
A quality problem looks different. If a product labeled at 20% THC tests at 12% in the next batch, that is not normal variation — that suggests a significant change in cultivation, processing, or possibly a testing irregularity. If the terpene profile shifts entirely — dominant myrcene in one batch, dominant limonene in the next — you are functionally purchasing a different product under the same name.
The honest answer is that the cannabis industry's tolerance for batch variation is wider than most patients realize, and wider than pharmaceutical standards would permit. This is not necessarily negligence. It reflects the difficulty of standardizing a complex botanical product. But it does mean patients should calibrate their expectations.
How to Use COAs to Compare
If you experience a noticeable difference between batches, the certificate of analysis is your diagnostic tool. Request the COA for your current batch and, if possible, for the previous one. Compare:
Total THC and CBD concentrations. A shift of two to three percentage points in flower is common. More than that warrants a question to the dispensary.
Terpene profiles, if reported. Look for changes in the dominant terpenes or significant drops in total terpene content.
Batch and harvest dates. A product sitting on a shelf for months may have degraded, particularly in terpene content. This is not a manufacturing defect — it is chemistry. Terpenes evaporate. THC converts slowly to CBN. Time is a variable.
If COAs are not available for comparison, that itself is useful information. It means you cannot verify what changed, and the dispensary cannot explain it with data.
What I Tell Patients
Expect some variation. Document what works — product name, batch number if available, dosage, effects. When something changes, ask for the COA before assuming the product is defective or that your body has changed. Sometimes the plant is simply being a plant.
Consistency in cannabis will improve as cultivation, processing, and testing mature. But it will never reach pharmaceutical precision, because we are not dealing with a single synthesized molecule. We are dealing with a living organism's chemistry, captured at one moment in time. Understanding that does not make the inconsistency less frustrating. It does make it less mysterious.