> ## Content Index > Fetch the complete content index at: https://compounded.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # The Same Stool Sample, Seven Companies, Seven Different Answers - URL: https://compounded.ghost.io/the-same-stool-sample-seven-companies-seven-different-answers/ - Published: 2026-11-03T12:30:00.000Z - Updated: 2026-11-03T12:29:59.000Z - Author: Connor Hayes Researchers at the National Institute of Standards and Technology did something the direct-to-consumer microbiome industry had not done to itself. They took a single homogenized human stool sample, split it into identical portions, and sent them in triplicate to seven commercial gut microbiome testing services. Twenty-one tests, one biological sample, no variation to explain except the methods. The reports came back over two to eight weeks. Across all of them, 1,208 distinct microbial taxa were identified. Exactly three genera were detected in every single sample. Setting aside one outlier, seventeen genera were common to all datasets, which is under 2 percent of everything found. Portions of one person's stool, analyzed by different companies, came back as different from each other as samples from entirely different people run through the same method. ### From the Lab to the Ledger The gut microbiome is real biology and the research behind it is serious. The trillions of bacteria in the intestine influence digestion, immune signalling and metabolism, and differences in their composition have been associated with conditions ranging from inflammatory bowel disease to metabolic syndrome. None of that is in question here. What this study measured is whether a consumer test can tell you what your microbiome actually contains, and the answer turns on how these tests are built. There are two broad sequencing approaches. One reads a single marker gene, the 16S ribosomal RNA gene, which is cheap and identifies bacteria to roughly the genus level. The other sequences everything present, which is more expensive and resolves down to species. The seven companies split between the two. Beyond that, almost every methodological choice differed. Sequencing depth ranged from under 20,000 reads to over 10 million, a span of more than two orders of magnitude. The number of species reported ranged from fewer than 100 to more than 1,700\. Sample collection and handling differed. Each company compared results against its own reference population and its own definition of a healthy gut. The consequences showed up in the specifics. Abundances of five common and clinically discussed genera varied widely between companies. Clostridioides difficile, a pathogen with genuine clinical significance, was reported present by three companies and absent by four, from the same stool. One company issued a consumer report based on a replicate in which more than half the sequencing reads could not be identified at all. ### Bio-Pipeline Ledger Research-grade microbiome sequencing: a legitimate and productive scientific tool when methods are standardized and results are interpreted against matched controls. Validated for research, not packaged as a personal health readout. Direct-to-consumer stool microbiome kits: widely sold, using varied methods, with methodological variability equal to or greater than biological variability for seventeen of eighteen commonly reported genera. Not analytically validated. Clinical stool testing for specific pathogens: ordered by a physician for a defined question, such as a Clostridioides difficile infection, and validated for that purpose. A different category of test entirely. Fecal microbiota transplantation and approved live biotherapeutics: validated for recurrent Clostridioides difficile infection, with approved products available. Narrow and genuinely effective. Microbiome-guided diet and supplement recommendations: generated by the consumer tests above from the data described. No validated basis, and the study found the same stool classified as both healthy and unhealthy, with divergent recommendations. ### The Clinical Reality Check The authors' own conclusion is restrained and worth quoting in substance: methodological variability can be of the same magnitude as biological variability, which means caution is needed in interpreting and acting on these results. A legal scholar who studies this market put it more bluntly in commentary, saying the tests may not be analytically valid, let alone clinically valid, and that they are very lightly regulated if at all. Analytical validity is the first of two hurdles and this study suggests these products do not clear it. The second hurdle is clinical validity: even a perfectly accurate list of your gut bacteria would need to connect to a health outcome and an action that improves it. That science is genuinely in progress, and there is no scientific consensus on what a healthy microbiome even looks like, which makes the comparison against a reference population somewhat circular. None of this means gut health is unimportant or that nothing can be done about it. The interventions with actual evidence behind them are unglamorous and do not require a test: dietary fibre from a range of plant sources, which is the single best-supported lever, and attention to whether specific symptoms warrant a proper medical workup rather than a consumer kit. The practical rule is the one that applies across this whole category. Persistent digestive symptoms, unexplained weight loss, bleeding or a change in bowel habit are reasons to see a physician, not reasons to mail a sample to a company. A test that returns a different answer depending on who processes it cannot support a decision, however confident the report looks. ![](https://storage.ghost.io/c/93/20/932004ad-b501-4cef-8a02-28e1473c42cb/content/images/2026/11/gut-microbiome-test-kits-disagree-2-cinematic.jpg) ###