Questions and straight answers
Written in plain English, including the questions we would rather not be asked. If something here is unclear or dodges the point, tell us and we will fix it.
The basics
Do you collect samples or run the laboratory work?
No. We begin when sequence files arrive. Sampling, filtration, preservation, extraction and sequencing all happen before we are involved, and the record of how they were done is declared by whoever submits the data. We check that the declaration is complete and well formed. We cannot check that it is true.
So what do you actually sell?
Processing. A laboratory or an organization sends us sequence data, we run it through one consistent method, and we return results that can be compared with anyone else's. We also sell a signed record showing exactly how each result was produced, which is what people need when a result has to hold up.
Are you a laboratory, a software company, or a standards body?
None of the three. We publish an open specification and run the infrastructure that applies it.
Is what you publish a standard?
No, and the difference matters. A standard comes from a recognized body after a formal balloted process: ISO, CSA, ASTM. We have written a specification: a careful public document that anyone may adopt, judged on whether it is useful rather than on who approved it. We explain why we are careful about this in We have not written a standard.
Is this a real company yet?
It is early. We launch in 2027. Right now we are talking to laboratories, agencies and buyers, and designing the study that will prove whether the idea holds up. We would rather be visible and unfinished than quiet and polished.
What don't you do?
We do not run samples, design assays, collect field data, sequence anything, interpret ecology, recommend laboratories, route work, or certify anyone's competence. We take sequence data that already exists and make the results comparable with other people's. That is a narrow job and we would rather do it properly than expand it.
Why won't you tell me what my results mean?
Because interpretation is where the value is, and an organization that both decides how results are expressed and sells the interpretation of them has a conflict that open licensing does not resolve. It also competes directly with the consultancies and agency scientists whose adoption we depend on. And interpretation is local — what a detection means depends on the system, the regulation and the history, none of which we hold. We explain the reasoning in We stop at the species list.
What if I don't know the answer to a required field?
We wrote about the reasoning in Say you don't know.
Money
What does it cost?
Roughly $40 per sample for processing, plus a monthly fee for the workspace where projects and clients are managed. Pricing will be published in full before launch. There is no negotiation and no volume favoritism. The price list is the price list, for everyone.
Do I need an account to use the free side?
At launch in 2027, every free function will require an account verified by email. The account will be free and always will be. Until launch, you can join the waitlist.
What is free?
The method itself, the identifier specification, the lookup that lets you check any identifier, all our validation data, and the record of every committee decision. Free, permanently. At launch, it will require an email-verified account.
Why give the specification away?
Because a specification you have to pay to read is one someone else will simply copy for free, and then theirs wins. Openness is the only version of this that works.
Do you charge laboratories differently from agencies?
No. Same method, same price, published openly. Charging one participant less than another would end the neutrality argument immediately, and the neutrality argument is the company.
Are you venture funded?
Not currently. We will say so plainly if that changes, because who funds us is a fair question to ask.
Neutrality
Are you trying to replace laboratories?
No, and we could not if we wanted to. Everything we do depends on laboratories producing good data first: the sampling, the extraction, the assay design, the controls, the judgment about when something looks wrong. That is the expensive, skilled part of this field and it is not ours. We handle what happens to the data afterwards, which is a much smaller job that nobody currently owns.
Why can't a laboratory just build this themselves?
Several have, and some of the software is very good — in a couple of cases better than anything we will ship in our first two years. The obstacle is not engineering. To make results comparable you need laboratories to send you their raw data, and no laboratory can sensibly send that to a competitor bidding against it for the same work. It is not about trust between individuals, most of whom know and respect each other. It is about position in a market.
What stops you becoming a laboratory later, once you have everyone's data?
It is written into how the company is governed, not just into our marketing. Changing it requires approval from a committee on which we hold no vote. That is the strongest commitment available to us short of asking you to take our word for it.
Who decides what goes into the specification?
An external science committee: agency scientists, academic taxonomists, an Indigenous data governance practitioner, people who use the data to make decisions, and two laboratory representatives drawn from competing laboratories. We staff it, fund it and publish for it. We do not vote.
Isn't ISO already doing this?
Partly, and we are glad they are. ISO 17805:2026 covers water sampling and preservation, and CSA W214:21, already published, covers reporting requirements and terminology. Those cover how a sample should be taken, handled and reported. What no published standard yet covers is what happens to the data afterwards: how a sequence keeps a stable identity across years and laboratories, how reference database versions are tracked, and how two laboratories' results are actually compared. That is the part we work on.
So are you a standards body, or not?
Not in the sense ISO or CSA are, and we should not pretend otherwise. They convene committees, run formal ballots and publish through national member bodies. We publish an open specification for a narrow gap, and we operate the infrastructure that applies it. Over time the sensible outcome is for what we specify to be taken up by a formal standards body. We would regard that as success rather than as being displaced. We have not written a standard explains the distinction in full.
Why not just wait for ISO to get to it?
Because standards processes take years, and in the meantime data is being produced that will not be comparable to anything. Somebody has to run something in the interim, publish what they learn, and hand the evidence to the bodies doing the formal work. That is a more useful role than waiting.
Are you accredited or endorsed by ISO?
No. We are not accredited by, endorsed by, or affiliated with ISO, CSA Group, or any standards body, and we will never imply otherwise. We adopt published standards by reference, which anyone is entitled to do.
Two competing laboratories on the same committee sounds unworkable.
That is intended. Two rivals on the committee is what makes it credible. If it is occasionally awkward, that is the mechanism working.
Do you take money from the laboratories you serve?
Yes, for processing, at the same published rate as anyone else. What we do not do is hold equity in a laboratory, take exclusivity, or offer preferential terms. Those are the arrangements that would actually compromise us. A laboratory paying the list price for a service is a customer, which is a relationship standards bodies have always had.
Isn't this what a certification body does?
It is close, and the comparison is useful. The testing, inspection and certification industry has operated on exactly this principle for over a century: an independent party whose value comes from having no stake in the outcome. Those organizations manage the obvious conflict through structural separation enforced by accreditation, under standards including ISO/IEC 17025 and ISO/IEC 17065. We are not accredited and do not claim to be. But the model is not novel, and we would rather learn from an industry that has run it at scale than invent our own version of impartiality.
Where is my data stored and processed?
Stored in the United States, under United States jurisdiction. Processing infrastructure moves to a United States region on our first federal or tribal contract, and no later than the end of 2027. Until then, some processing occurs outside the United States. We would rather tell you that than have a procurement review discover it. If your program requires in-country processing from day one, say so and we will tell you where we are before you commit to anything.
Will you recommend a laboratory to me?
No. We do not route samples, operate a marketplace, rank providers, or maintain a preferred list. If you ask us who to use, we will decline — not to be unhelpful, but because a reference layer that steers business toward some laboratories and away from others is not a reference layer. We will happily tell you what to ask them.
Isn't the whole thing a conflict of interest?
Judge it by structure rather than intent. We earn per sample processed regardless of which laboratory produced it and regardless of what the result says. We hold no vote over our specification. The specification is openly licensed, so anyone can fork it if we behave badly. Those are the checks. If you think they are insufficient, we would genuinely like to hear why.
Technical
What is environmental DNA?
Every living thing sheds genetic material: skin cells, mucus, scales, pollen, spores. Filter a water or soil sample and you catch it. Sequence what you caught and you get a list of which species have recently been there, without ever seeing an organism.
What is environmental RNA, and how is it different?
DNA is durable and hangs around for days to weeks, so finding it means something was there recently but not necessarily that it is still alive or even local. RNA falls apart within minutes to hours, so finding it means something living was active right there, very recently.
Which methods do you support today?
DNA metabarcoding, across water, soil, sediment and air. Targeted qPCR and dPCR, hybridization capture and environmental RNA are on the roadmap and are not supported at launch. We say this plainly because most United States invasive species work is qPCR, and anyone arriving here for that should know before they spend time on us.
What is a persistent identifier and why does it matter?
Species names change as taxonomy is revised. If your records are stored by name, a revision quietly rewrites your history and you cannot tell later whether your data changed or the naming did. Store them against a permanent identifier instead, with the name attached as a dated label, and a revision updates the record rather than breaking it.
Do I need to keep my raw sequence files?
Yes, and most people do not. The report from a laboratory is its interpretation on the day, using the reference library available at the time. The raw files can be reanalyzed years later against better references. Ask for them, including the control samples, and ask before you place the order rather than after.
Will this work with data I collected years ago?
If you kept the raw sequence files, usually yes. If you only kept the reports, largely no. This is the single most common regret we encounter.
Does this work for soil, or only water?
Both, and also sediment and air. Soil is in some ways harder — genetic material binds to particles and is patchy at centimeter scales, so replication matters more. But the comparability problem is identical, and soil work suffers from it at least as much because reference coverage for soil fauna and fungi is thinner than for fish.
Does this work in marine environments?
Yes. Marine work brings its own complications. Depth becomes a primary variable, sample volumes are often larger, and a detection can originate a long way from where it was collected. Offshore monitoring also tends to run for decades across changing vendors, which is precisely the case where a shared reference frame earns its keep.
Which domain are you focused on?
Every substrate, through submitted sequence data. Most sites carry obligations in more than one — a mine has a receiving stream, a port has a shoreline and a seabed, a wind lease has a cable route running inland. Today those are surveyed separately and reported in incompatible formats, and nobody can give the operator one comparable picture of the site.
Can AI solve this instead?
Not this part of it. Machine learning is genuinely useful in environmental DNA — for taxonomic assignment where references are sparse, for combining eDNA with acoustic and satellite data, for forecasting. What it cannot do is create provenance. A model trained on results that were never comparable produces confident answers that are still not comparable, faster. Compression of the analysis layer does not substitute for a record of how each input was made.
Why don't you just build a better reference database?
Because the constraint is not curation, it is collection. A reference database can only contain organisms somebody has physically found, vouchered and sequenced. The largest private reference libraries in this industry were built over a decade of fieldwork and are held as competitive assets, which is exactly why a shared reference layer cannot depend on owning one.
Limits
Is environmental DNA approved by the EPA?
Metabarcoding is not an EPA-approved method. It is a survey and screening tool. Targeted assays are used in specific federal and state programs under their own protocols. If a detection matters, your agency has a confirmation process and that is the one that counts.
Will Common Measure make my results legally defensible?
It will make them auditable, reproducible and comparable, which is a large part of what defensibility means in practice. It will not by itself make an unapproved method approved. Anyone telling you otherwise is overselling.
Does a detection mean the species is there?
It means its genetic material was in your sample. DNA travels and persists, so it could have come from upstream, from a bird, or from something that died a week ago. That gap between detection and presence is real, and pretending otherwise is how the field loses credibility.
Can eDNA tell me how many organisms are present?
No. It tells you what was probably there, not how much. Read counts are affected by so many things other than abundance — body size, shedding rate, water chemistry, soil binding, primer efficiency — that treating them as a population estimate is not defensible in any substrate.
What if the specification turns out to be wrong?
Then we publish the correction, keep the old version available, and record what changed and why. Every result records the version it was produced under, so nothing gets silently reinterpreted. A specification that cannot admit error is not a credible specification.
What could make this whole idea fail?
Three things, honestly. Laboratories may decide their data is too commercially sensitive to share with anyone, whatever the governance. The comparison study may show a shared reference frame does not improve agreement enough to be worth the effort. Or nobody may ever be required to compare results, in which case this stays a nice idea nobody funds. We are testing all three before we build much further.
Getting involved
How can I take part?
Comment on a draft, join the cross-laboratory study, propose a change to a definition, tell us where you think the idea breaks, or bring us the requirements of a database your results have to reach. All of it is free and none of it requires an account.
What are you most looking for right now?
Laboratories willing to take part in the comparison study, particularly ones that compete with each other. And objections. Especially objections.
I think you are wrong about something. Do you want to hear it?
Yes, more than almost anything else. We are early enough to change course cheaply, and that window closes.
Can I use your work commercially?
Yes. The specification, the vocabulary and the validation data are published under CC BY 4.0. You may build on them, adapt them and sell products based on them, including in competition with us, as long as attribution is kept.
Who do I contact?
info@commonmeasure.bio, or use the form. We reply to everything, usually within two business days.
Longer answers
Some questions need more than a paragraph. These pieces work through them.
Practice & Adoption
The metered side of Common Measure
The same three people, and the moment each one decides Common Measure is worth paying for.
Read it →Practice & Adoption
The free side of Common Measure
Three examples of people who rely on eDNA results, use Common Measure, and never pay us.
Read it →Something missing?
If a question you have is not here, or an answer above reads like it is avoiding the point, tell us. We will add it or rewrite it.
Ask us