Los Angeles, August 23rd, 2018 — Yesterday, Exact Sciences gave the diagnostics industry a useful reminder: the market does not reward biomarkers simply because they are scientifically interesting. It rewards them when they become easier to use, easier to order, easier to explain, and easier to scale.
Exact announced a U.S. co-promotion agreement with Pfizer for Cologuard, its non-invasive stool DNA test for colorectal cancer screening. Pfizer brings serious scale, physician access, and marketing muscle. Exact will continue to run the test. In response, Exact's stock jumped sharply.
The signal from investors is clear: the market is not celebrating the discovery of a new biomarker or a new test — Cologuard is already approved and on the market. What changed is the perceived ability to turn that test into a more deeply embedded part of routine care.
That is the real lesson! In precision diagnostics, we still spend too much time talking about biomarkers as if discovery is the hard part and adoption is the afterparty. Find the right signal, validate the biology, publish the paper, and surely the market will follow.
Unfortunately, healthcare does not work like a TED Talk. A biomarker creates value when it is used, not when it is discovered. That sounds obvious, but the industry still behaves as if scientific novelty is the finish line rather than the starting line. In the real world, a biomarker has to survive a long and often brutal chain of practical questions: Is it on the agenda when meeting with the doc? Who orders it? When? What clinical question does it answer, and how does it change patient care or management? Is that care change reimbursed? Can the workflow absorb it without creating yet another portal, PDF, login, or fax-machine-adjacent experience?
Break any link in that chain, and even a technically excellent test can fail to matter. That is why the Exact/Pfizer announcement is so interesting. It is a reminder that diagnostics scale when biology gets wrapped in infrastructure, including provider education, payer access, logistics, ordering behavior, follow-up, and trust.
In other words, Cologuard is a workflow story. And that is true across precision diagnostics. In early stage discovery and translational settings, the industry often rewards "more": more analytes, more omics layers, more resolution, more signal. I'm guilty of that! But the market does not primarily reward data. It rewards uncertainty reduction. And those are not the same thing!
Data is only valuable if it helps a clinician (or biopharma partners) make a better decision under real-world constraints. The best diagnostics platforms collapse ambiguity and reduce the number of plausible next steps.
That is why some diagnostics companies succeed even when they are not the most comprehensive in a purely scientific sense. They win because they are clearer, more actionable, and better integrated into a decision someone already cares about. They may not be the biggest data buffet in the room. But they answer the question that matters.
Clinical integration is where this becomes visible. A biomarker may be analytically valid and clinically valid, yet still fail commercially if it arrives too late, requires too much interpretation, or sits outside treatment pathways. In that case, the test exists in the abstract, but not in the operating system of care. And in healthcare, "abstract" is where good ideas go to become posters at conferences! This is why adoption rarely tracks scientific merit one-to-one. The market is not a meritocracy of biological insight.
I have to admit that (despite my beef with them) payers understand this instinctively. They reimburse decisions that lead to better patient outcomes, not data. Providers understand it too, even if they say it differently. Their daily challenge is knowing whether information affects patient care. If a biomarker does not alter management, it becomes one more input competing for attention in an already crowded clinical environment. The last thing clinicians need!
That is why integration is not a downstream commercialization issue. It is the design problem. Diagnostics companies should start with the end in mind! The "first" question should be "Can this become actionable in a real clinical pathway?" The companies that understand this build differently. They focus less on breadth for its own sake and more on where uncertainty is most expensive. They ask which decision is hardest, most consequential, or most constrained — and then design around that decision.
That often leads to a narrower claim, but a stronger one. It is better to be indispensable in one high-value decision than vaguely interesting across many.
This is also why commercial partnerships matter so much in diagnostics. In therapeutics, a product can sometimes create its own gravitational field. In diagnostics, gravity is harder. The test has to fit into existing clinical behavior and patient flow. You cannot simply sprinkle a biomarker over the healthcare system and hope it becomes standard of care.
The Exact/Pfizer deal is a useful reminder that the "product" in diagnostics is bigger than the assay. It is the assay plus the entire clinical workflow, from ordering pathway to the next clinical action. The science gets you into the conversation; the workflow determines whether you stay there! That may be the central lesson for precision medicine right now. The field is not short on biomarkers. It is short on biomarkers that can survive the journey from signal to system. Integration is what turns it into a market.