A supplement customer cannot feel a vitamin D capsule working. That fact sits underneath most subscription supplement businesses in Europe. The formulation can be good and the customer can take it every morning for four months, and all they observe is the money leaving.
Churn is rarely a rejection of the product. It is boredom. Brands answer with new flavours, streaks and coaching, none of which touches the actual problem, which is that the customer cannot observe what they are paying for.
Blood testing is the obvious response and also a trap for the careless. The honest answer first: a baseline plus a retest works for the narrow set of nutrients whose markers respond inside a subscription cycle, which means vitamin D, ferritin, B12, folate and homocysteine. For everything else a test shows the customer where they stand rather than whether the product worked. And no laboratory result lets a brand say more about its product than EU law authorises.
Why is supplement churn a measurement problem?
The value of a supplement subscription is a belief rather than an observation, and belief decays. A number does not, because it belongs to the customer.
A baseline also gives a brand its first honest segmentation. The British Society of Gastroenterology guideline puts the lower limit of normal for serum ferritin at 15 to 30 micrograms per litre, with a level below 15 indicating absent iron stores. A customer at 12 has a reason to buy an iron product and a customer at 90 does not, so measurement shrinks the honestly addressable market before it grows the retained one.
Which markers move on a timescale a subscription can use?
Marker kinetics rather than the billing cycle should set the retest date.
Vitamin D. The circulating form, 25-hydroxyvitamin D, has a half-life of roughly 15 days. Accumulation towards a new steady state reaches about 94 per cent after four half-lives, so the arithmetic gives roughly 60 days before the number stops climbing and 90 before it settles. One controlled winter study in healthy men found equilibrium 25-hydroxyvitamin D rising about 0.70 nmol/L per extra microgram of daily cholecalciferol, or roughly 17.5 nmol/L at 25 micrograms a day.
Ferritin. Slow and easy to misread. The same guideline recommends continuing iron for around three months after haemoglobin normalises, making twelve weeks the earliest sensible retest. Ferritin is an acute phase protein, so infection or hard training can lift it while iron status is unchanged.
Vitamin B12. Serum B12 rises readily on oral supplementation, which is the difficulty. The British Committee for Standards in Haematology guideline states there is no gold standard test for cobalamin status, that serum cobalamin is first line with plasma methylmalonic acid second line, and that definitive cut-off points are not possible. A move from 250 to 600 pmol/L is a real change in the analyte and a weak statement about the customer.
Folate. The same guideline found serum folate equivalent in diagnostic capability to red cell folate and makes it the first line test. It is responsive, but folate also comes from food, so a rise cannot be attributed to a capsule unless intake is accounted for.
Homocysteine. The cleanest case. A meta-analysis of twelve randomised trials covering 1,114 people found folic acid at 0.5 to 5 mg daily reduced blood homocysteine by 25 per cent, with a 95 per cent confidence interval of 23 to 28 per cent, standardised to a pretreatment homocysteine of 12 micromol/L and folate of 12 nmol/L. About 0.5 mg of vitamin B12 added a further 7 per cent, taking the worked example from 12 to 8 or 9. Reductions were larger where pretreatment homocysteine was higher, so the baseline identifies responders in advance.
One caveat covers all five. That meta-analysis closed by saying large trials were still needed to show whether lowering homocysteine reduces vascular risk. A marker moving is not a health outcome.
Which ingredients have no marker worth retesting?
Most of them. Collagen peptides, most botanicals, adaptogen blends, mushroom extracts and greens powders have no routine blood marker that shifts in a way a customer would read as evidence, so a test bolted onto them produces a report nobody can connect to the purchase. The EU Register of nutrition and health claims records many non-authorised entries where the claimed effect was judged not substantiated, including coenzyme Q10 and resveratrol. Those are verdicts on dossiers rather than substances, but a retest is unlikely to record what a funded application could not.
Worse are the markers that move for other reasons. Lipids, inflammatory and glycaemic markers respond to diet, sleep, illness and training at once, so attributing a change to one capsule is indefensible.
Where does EU health claims law draw the line?
Regulation (EC) No 1924/2006 applies to claims in commercial communications, whether in labelling, presentation or advertising, so it governs the marketing email as firmly as the tub. Under Article 10(1), health claims are prohibited unless authorised and included in the lists provided for in Articles 13 and 14, and the everyday list is the Annex to Commission Regulation (EU) No 432/2012.
Article 2(2)(1) defines a claim as any message, in any form, implying that a food has particular characteristics, and Article 2(2)(5) defines a health claim as one implying a relationship between a food or its constituents and health. Those definitions show where the line sits. A laboratory report is a measurement about a person rather than a message about a food, so delivering a customer their own result is not in itself a health claim. Marketing that places the result beside the product implies a relationship between food and health, and the Regulation then applies in full. A private report carrying no product messaging does not.
Three further provisions catch brands out. Article 10(3) allows a general reference to wellbeing only alongside a specific authorised claim. Article 12(c) bars claims referring to recommendations of individual doctors or health professionals. Article 3(e) prohibits claims referring to bodily changes that could exploit fear, which is the risk in marketing a low result as alarming.
Separately, Article 7(3) of Regulation (EU) No 1169/2011 prohibits attributing to any food the property of preventing, treating or curing a human disease, and Article 7(4) extends that to advertising. In Germany, section 11 of the LFGB turns that into a domestic prohibition covering advertising as well as labelling. Disease risk reduction claims are possible, but Article 14 requires individual authorisation.
Article 13(5) does allow a claim based on newly developed scientific evidence, with an option to request protection of proprietary data, through the procedure in Article 18. Retest data from your customers is a reason to start that process rather than a shortcut around it, and Article 6 puts substantiation on the operator. Two German points close this out. Advertising a diagnostic procedure can fall under section 1(1) no. 2 of the Heilmittelwerbegesetz where the statement relates to detecting illness, and section 8(3) of the UWG lets competitors demand that you stop.
What a marker change can and cannot support in marketing
| Marker |
What an authorised claim lets you say |
What the change cannot be used to claim |
| 25-hydroxyvitamin D |
Wording from Regulation (EU) No 432/2012 such as "Vitamin D contributes to the normal function of the immune system", where the product is at least a source of vitamin D |
That the rise prevented an infection, or any wording implying the product prevents, treats or cures a disease |
| Ferritin |
"Iron contributes to the reduction of tiredness and fatigue" or "Iron contributes to normal oxygen transport in the body", on the same source condition |
That a named customer's tiredness was resolved by the product, which goes beyond the authorised wording and, unsupported, is misleading under Article 3(a) |
| Vitamin B12 |
"Vitamin B12 contributes to normal energy-yielding metabolism" or "to normal homocysteine metabolism" |
That a higher serum number proves a functional deficiency was corrected, since the haematology guideline records no gold standard test and no definitive cut-off |
| Folate |
"Folate contributes to normal homocysteine metabolism", with the Article 10(2) accompanying statements |
That the customer's diet was inadequate, which Article 3(d) restricts, or that the rise came from your capsule rather than from food |
| Homocysteine |
Only the nutrient claims above, because the authorised wording is about metabolism rather than about the number falling |
That a lower value reduces cardiovascular risk, which would be an Article 14 disease risk reduction claim and needs individual authorisation |
The table concerns wording, not product quality. Strong internal data still leaves a brand limited in public to the authorised sentence.
What does the operational lift actually look like?
Less than founders fear on logistics, more than they expect on copy. A diagnostics partner should own the laboratory network, collection kits, courier and cold chain, the ordering surface, structured machine-readable results rather than only a PDF, white-labelled delivery under your brand, and the data protection pack with an enumerated sub-processor list and EU hosting.
The brand owns everything the customer reads and everything touching a claim: offer design and pricing, panel selection, marketing copy and its compliance file, expectation setting before the first draw, and the response when a result worries someone. Interpretation of an individual result stays with the treating clinician, and the support script has to route to that.
Accreditation belongs to the analysing laboratory rather than the platform, so ask which laboratory runs your samples. Two further items need a named owner: the bad day, meaning a haemolysed sample or a failed marker, and the export format if the relationship ends.
What happens when the retest shows nothing?
Sooner or later a customer who took the product exactly as directed opens a report identical to the last one. That is a churn event and a data point at once, so the offer has to be designed for it.
Four decisions reduce the damage. Pick the retest panel from the baseline, so only markers that were low get remeasured. Set the interval from the marker's kinetics, meaning at least eight weeks for vitamin D and twelve for ferritin. Write the flat result email before launch, explaining adherence, absorption and timing rather than defending the product. Offer a next step that is not more of the same.
An illustrative example follows, with every input an assumption to replace with your own. Assume a subscription at 40 EUR per month at a 60 per cent gross margin, giving 24 EUR of gross profit per retained month. Call the combined cost of one baseline and one retest T. Absorbing testing into the subscription breaks even only if it buys T divided by 24 extra retained months. At T of 120 EUR that is five extra months, and at 240 EUR it is ten. Ten extra months of average customer life is a very large claim for a retention feature, which is why the arithmetic usually pushes brands to price testing separately.
Where Aniva fits
Aniva provides the diagnostics layer a supplement brand would otherwise assemble from four vendors: orchestration of accredited partner laboratories, kits and logistics, an ordering surface with an API, white-labelled result delivery and the compliance framework. Aniva is not itself a laboratory, and accreditation including RiliBÄK and ISO 15189 sits with the analysing partner laboratory such as ZOTZ|KLIMAS.
The same infrastructure runs Aniva's consumer membership at 199 EUR per year for more than 100 biomarkers, with results in the app in about a week and draws in Germany and Finland performed by licensed physicians. Panels scale beyond 2,000 markers through add-ons and data is stored in the EU under GDPR. Details sit on the diagnostics page, and companion pieces cover the same model for clinics, digital health teams and wellness creators.
The bottom line
Testing does not make a weak supplement work. It makes a good one legible and a weak one visible, which is the real trade. The brands that benefit sell nutrients whose markers respond inside a quarter and write copy against the authorised claim list.
Frequently asked questions
Can a supplement brand say that its product raised a customer's vitamin D?
Not as a marketing message, unless the wording matches an authorised claim. Under Article 10(1) of Regulation (EC) No 1924/2006, health claims are prohibited unless authorised and on the lists provided for in Articles 13 and 14. Showing an individual their own result privately is a measurement about a person, while publishing a before and after tied to the product is a health claim.
How long after starting a supplement should the retest happen?
It depends on the marker rather than the billing cycle. 25-hydroxyvitamin D has a half-life of roughly 15 days, so a new steady state takes two to three months and a retest before eight weeks reads a number still rising. For iron, the British Society of Gastroenterology guideline recommends continuing treatment around three months after haemoglobin normalises, which puts the retest at twelve weeks.
Is showing a customer their own blood result a health claim?
Normally not on its own. Article 2(2)(1) defines a claim as a message implying that a food has particular characteristics, and Article 2(2)(5) covers messages implying a relationship between a food and health. A private laboratory report says nothing about the food. Placing it beside product messaging is what turns it into a claim.
Which markers belong in a supplement brand's baseline panel?
Start with the nutrients the product contains and whose markers respond inside a quarter: vitamin D, ferritin with an inflammatory marker beside it, B12, folate and homocysteine. Markers no product in the range can influence give information without giving a reason to stay.
What should a brand do when the retest shows no change?
Treat it as a planned outcome. Explain adherence, absorption, timing and the reference range, avoid defending the product, and offer a next step that is not a larger order. A nutrient that never moves in your customer base says something about the dose.
Who holds the laboratory accreditation, the brand or the diagnostics partner?
Neither, in most setups. Accreditation such as ISO 15189 and RiliBÄK sits with the laboratory that analyses the sample. A diagnostics platform orchestrates those laboratories and provides logistics, software and compliance, and it should name the analysing laboratory on request.
This article is general commercial and regulatory information for supplement brands. It is not legal, medical or nutritional advice, and it does not describe any diagnosis or treatment. Legislation is summarised for orientation, and the authoritative texts are the published instruments cited. Laboratory accreditation and reference ranges belong to the analysing laboratory, and interpretation of any result remains with the treating clinician. The unit economics example is illustrative.
