Switched on, but not yet built.

The programme had generated a few dozen clicks, zero transactions and zero revenue across its first fortnight. That absence of sales is not itself a failing for a programme this new. The concern was that the fortnight had been spent accumulating publishers rather than preparing the programme those publishers would need in order to sell anything. Seventy-five partners had joined, but the joins were dominated by demand-harvesting types while demand-generating partners — content, editorial, influencer, community — made up roughly a quarter of the base. On inspection that quarter was thinner still: several partners labelled as editorial were mislabelled coupon aggregators, and one self-described content creator was actively promoting a competing supplement.

Underneath, the operational shell was largely hollow. There was no welcome email, so every newly approved partner received only the platform’s generic default and no instruction on how to promote the brand. There were no offers or discount codes, so partners had nothing to feature and no reason to prioritise the programme in any given month. There was no product feed, so comparison, CSS and shopping-directory partners — several of which had already joined — had no data to work with. The creative library contained two assets: a plain text link and a single small logo tile. The communication centre had never been used, and no triggered automations existed to catch a partner who joined and went quiet. The programme looked live from the outside and offered a joining publisher almost nothing to act on.

Three issues rose above the rest. First, commercial risk: an unpaid platform invoice sat against an account flagged as having exceeded its credit limit with no direct debit in place — precisely the condition under which a network can suspend tracking links, which would have silently broken the programme just as it was trying to establish itself. Second, margin risk: the single flat 20% commission was paid identically to every partner type, including cashback services and sub-networks that conventionally sit below the standard rate, and it was calculated on an amount that included sales tax — giving away commission on money the brand never collected. For a supplement brand that sector benchmarks place in a 5–15% band, 20% flat with no differentiation was both above the ceiling and structurally exposed the moment volume arrived. Third, compliance risk: the brand positioned itself as a natural alternative to GLP-1 medication, yet the programme terms contained no rules governing how publishers could make health claims, and the brand’s own careful regulatory disclaimers were nowhere reflected in what partners were contractually required to carry.

The quickest wins were almost all housekeeping and could be actioned within days: clear the invoice and set up direct debit to protect link status; write a real welcome email; create a launch offer and code; complete the profile, whose public website link was blank and whose description carried a spelling error and no commercial detail; and decline or reclassify the one pending application, a coupon site presenting itself as editorial.

The larger, slower work was strategic: deliberately recruit content, editorial and influencer partners suited to a US supplement; differentiate commission by partner type; and stand up the seasonal, communication and activation rhythm that turns a list of publishers into a performing programme.

The good news, and the reason this audit reads more like a build plan than a rescue: nothing here was broken by history. There was no leakage to unwind, no aged validation queue, no damaged partner relationship, because nothing had happened yet. The tracking was clean, the cookie and validation windows sensible, the brand website professional and compliant, and the 20% headline rate attractive enough to draw quality partners once the programme gave them something to promote. This programme needed a fortnight of disciplined building, not a rescue.

2A. Metric scorecard

Several metrics are correctly unrateable at two weeks live; the client document marks them as baseline rather than inventing a judgement. Benchmarks are AME Reference Library values for the health and supplements sub-sector and are not client data.

2B. Area scorecard

A clean, correctly configured tracking foundation. Tracking settings showed the click identifier automatically appended to all links, a default query-string append carrying source and campaign identifiers, and the brand website correctly registered. This matters because tracking is the one thing that is expensive to fix retrospectively: a programme that launches with broken or partial tracking spends its first months disputing missed transactions and rebuilding partner trust it has not yet earned. This programme avoided that trap entirely. To protect it, run a test transaction through a tracked link as soon as a first genuine or manual order is available, so end-to-end tracking is confirmed rather than assumed before volume arrives.

Sensible, sector-appropriate cookie and validation windows. The attribution window was set to thirty days and auto-validation to thirty days, with an email notification enabled two days before transactions auto-validate. For a considered-purchase supplement, a thirty-day cookie gives content and editorial partners a fair chance of being credited for the research-and-return behaviour typical of the category, and the pre-validation alert is exactly the control that stops cancelled or returned orders slipping through to payment. This is quietly mature validation hygiene for a two-week-old account and should be preserved rather than shortened.

A professional, compliant brand website that will convert affiliate traffic. The destination partners send traffic to was a well-built store with a clear hero proposition, a visible money-back guarantee, an accredited business badge, recognised payment marks and a full set of regulatory disclaimers correctly clarifying that the product is not a GLP-1 medication and advising vulnerable groups to consult a doctor. Conversion readiness is the single biggest determinant of whether affiliate clicks become commissionable sales, and this programme started with a page designed to convert and legally careful about how it does it. The protection task is to keep the affiliate journey pointed at pages this strong.

Genuine approval discipline at the application stage. Against seventy-seven approved partners, thirty applications had been declined — and the declined set was concentrated exactly where it should be: low-quality coupon and editorial aggregators in markets with little relevance to a US supplement. Someone was reading applications rather than rubber-stamping them, which is the foundation of a clean programme. The way to extend the strength is to apply the same scrutiny to the low-value sub-network and coupon partners that were admitted, and to pair the discipline with proactive recruitment so the base is shaped rather than merely filtered.

A competitive headline commission that will attract quality partners. A 20% default sits at or above the top of the supplement range and, as a recruiting signal, it is genuinely attractive — content and influencer partners scanning for programmes to promote will notice it. The strength is real. The task is to keep that attractiveness for the partners who earn it while removing the margin exposure of paying the same rate to cashback services and sub-networks. Used deliberately, the 20% becomes a recruitment magnet rather than a blanket cost.

5.1 Overview

The programme had seventy-five joined partners, thirty declined applications and one pending. No partner had yet driven a transaction, so revenue, commission, cost per acquisition and active rate could not be calculated. The analysis in this section is therefore about the shape and quality of the base being assembled rather than its performance — which, at two weeks live, is the more useful question anyway.

The base had been assembled largely through a cluster of sign-ups over a few days shortly after launch, consistent with mass-market partners auto-attaching to a newly launched programme rather than being individually recruited.

5.2 Type distribution

Grouping these, demand-harvesting partners — discount, cashback, sub-network, comparison, shopping directory — made up well over half the base, while demand-generating partners — content, influencer, editorial, media, community — sat at roughly a quarter. That quarter was weaker still on inspection: the editorial count was inflated by mislabelled coupon aggregators, and one self-described content creator was actively promoting a competing supplement, so the genuinely usable demand-generating base was smaller than the raw count implied.

Set against the health sector’s optimal mix, the distortion is clear. Discount-code partners are over-weighted, sub-networks run at more than three times the recommended ceiling, and the content and influencer partners that should form the majority are the minority. The programme had, in effect, been built upside down for its category.

5.3 Concentration

With no transactions, revenue concentration cannot be measured; there is no top-one or top-ten dependency to report. This is worth stating plainly because it is an opportunity rather than a gap: the programme can still shape which partners come to dominate its revenue before any concentration sets in. Every other audit in this library is trying to unwind a concentration that already exists. This one could simply avoid creating it. The goal is to ensure the eventual top tier is populated by content, editorial, community and influencer partners that create incremental demand, rather than by the sub-networks and coupon sites that currently dominate the count.

5.4 Device performance

Device-level performance was not yet measurable with no transactions recorded, though the profile confirmed the site is mobile-optimised and the brand website renders responsively. Device reporting is available on the plan and should be reviewed as soon as transactions accrue — particularly in this category, where a large share of discovery happens in mobile community and social contexts and the purchase may complete elsewhere.

5.5 Pending approvals

One application was pending: a partner presenting as editorial content whose own description identified it as a coupon and deals site outside the programme’s target market. This is a mislabelled type on a market-restricted programme and fits the profile of the low-value applicants already, correctly, declined. The recommendation was to decline it or request clarification rather than approve.

5.6 Extended analysis

The base was broad but shallow. Its breadth — seventy-five partners across sixteen promotional types — flattered the count while masking that the partners most able to sell a trust-led supplement were barely present.

Geography compounded it. Although the programme was market-restricted, roughly two thirds of joined partners were based outside the target market, many of them coupon and content partners whose audiences are not the buyers this brand needed. Seven joined partners had blank or unsubscribed contact emails, which would limit their reachability through any communication plan before one even existed.

Two further data-quality signals warranted a check before any spend or attribution flowed. Several operators appeared under more than one publisher ID, which carries a double-payment and transparency risk. And the creator account noted above was promoting a rival supplement, which places it outside the content tier rather than inside it.

The strategic read. Deprioritise further sub-network and low-relevance coupon growth. Verify the cluster of near-identical sub-network entities admitted in a single window before any of them transact. Pour recruitment energy into US health and wellness content creators, editorial titles, communities and influencers. Protect the approval discipline already being shown. Renegotiate the flat commission that currently rewards harvesters and creators identically. Build the content and influencer layer that is almost entirely missing — because in this category, it is the only layer that can grow.

No partner had transacted, so dispositions are based on type, relevance and risk rather than performance. At this programme size the matrix covers essentially the entire active base rather than a top twenty.

There was not yet a performance-based top twenty to manage, because no partner had transacted and click volume was spread thin across the base. Relationship management at this stage is therefore forward-looking: identifying the partners most likely to become the programme’s future top tier, and beginning deliberate relationships with them before competitors do.

The partners worth investing in immediately were the content, influencer and community partners already in the base — a small handful — plus any US health and wellness creators recruited deliberately. For a supplement in this category, these are the partners who build the trust and education that converts, and they respond to personal contact, early access to offers, bespoke codes and content support far more than to a generic newsletter. Two of them had joined weeks earlier and never been contacted at all.

The immediate actions were to reach out individually with a warm welcome, provide a tested launch code — ideally single-use or prefixed so each partner’s contribution is separately attributable — and offer bespoke creative or a modest rate uplift for genuine content partners.

A risk that has not arrived yet. Because no exclusive or creator codes existed, there was no last-click overwrite risk on the day of the audit. That risk arrives the moment creator codes are issued into a base that also contains coupon and cashback partners. The protection task is to enable voucher attribution and consider an assist reward before content partners are asked to invest, so their contribution is not captured by a downstream coupon click. Doing it afterwards means explaining to a creator why their sale went to someone else, which is a conversation that ends relationships.

Content-freshness checks, business reviews, media-pack collection and partner-day activity were all premature at two weeks live; the operating calendar schedules them for the point at which a genuine top tier exists. Meeting and call history should be logged from the first outreach so a relationship record exists as the programme grows.

Recruitment was the programme’s biggest growth lever by a wide margin, and it was close to dormant. Partner discovery tooling was available on the plan and presented a full set of publisher, technology, influencer and brand-partner discovery tools, but there was no evidence of invitations sent or recommendations actioned, and the dashboard confirmed no marketplace offers had been made. The publisher base had grown by absorption — mass-market coupon and sub-network partners joining of their own accord — rather than by deliberate targeting.

This matters more here than in any other programme in this library, and it is worth being blunt about why. In a category adjacent to GLP-1 medication, paid social and search are restricted. Ad accounts get disapproved. Landing pages get flagged. There is no paid channel to fall back on while the affiliate programme matures. Recruitment was not the growth lever; it was the only lever.

Rebuilding the invitation pipeline. Discovery tooling allows filtering by promotional type and region, so the priority was to search for US-based health and wellness content creators, editorial publishers, community operators and influencers, and to invite them with a personalised message leading with the commission rate, the money-back guarantee and the launch offer once it existed. Invitations should be tagged by type so the growing base can be managed against a target mix, using tagging tooling that was at the time entirely unused.

Where the real partners are. The most valuable partners in this category are not in any media kit and do not appear in a marketplace search. They are people with genuine, personal experience of the category — the community operators, the creators documenting their own journey, the writers whose audiences trust them precisely because they are not an advertiser. Finding them is manual work. It is also the entire growth plan, and it is what a restricted category forces you to do properly rather than buy your way past.

What not to do yet. The marketplace should be reviewed for seasonal and niche placements suited to the supplement calendar, but judged on fit, cost, likely incrementality and margin — not taken speculatively. At zero revenue, paid tenancy is premature and should wait until the programme has proven it converts.

One free channel being ignored. The publisher invite link auto-approves anyone who joins through it, so given the volume of low-value partners already in the base, it should be shared only with specific trusted partners and never posted publicly. Meanwhile the brand website carried no affiliate programme call to action at all — adding one creates a low-cost inbound recruitment channel for exactly the content and creator partners the programme needs, and costs nothing but a page.

9.1 Current state

The programme ran a single commission group at a flat 20% applied identically to all partners. There were no bespoke rates, no scheduled changes, no bonus rules, no basket-value tiers and no assist or product-level commissioning. The calculation base included sales tax. In other words, commission was not yet a strategy; it was a single number applied to everyone.

As a headline recruiting signal, 20% is attractive and should be preserved for the partners who earn it. As an economic structure it is exposed. Sector benchmarks place this category at 5–15%, with the wider supplement range reaching 20% at the top. Premium conventions then adjust from the standard rate: content and influencer partners typically a few points above, cashback, sub-network and comparison partners a point or two below. A flat 20% pays the demand-harvesters — who dominate the current base — the same as or more than the demand-generators the programme should reward, and does it on a tax-inclusive base that gives away commission on money the brand never collected.

With zero transactions there was no loss on the day of the audit. That is precisely why it was the moment to restructure: before volume makes the leak real, and before partners become accustomed to a rate that would later have to be reduced under notice rules. Cutting a partner’s rate is a relationship cost. Never having offered it is free.

9.2 Recommended architecture

9.3 Budget impact

Because there was no revenue yet, restructuring commission carried no immediate cost and every future benefit: it prevents the programme from ever paying its top rate on a tax-inclusive base to low-incrementality partners.

Modelled forward, differentiating the base to roughly 12–15% and dropping cashback and sub-network partners to 8–12% materially reduces blended commission cost per order against the current flat structure, while the premium content and influencer tier concentrates spend on the partners most likely to grow the programme. Excluding sales tax from the base removes a further leakage of commission on non-revenue.

The recommendation was to implement the differentiated structure before the first significant wave of transactions, communicate it to partners with the required notice, and keep the headline 20% only for a defined premium tier where it functions as a recruitment and reward tool rather than a blanket cost.

The client document carries each task with fifteen columns. The fields are laid out here as a card because fifteen columns is unreadable on any screen. No field has been dropped in the transform.

Day 0–30: stabilise and build

This programme did not need a rescue. It needed a fortnight of disciplined building. Almost nothing in the first thirty days costs money — it is configuration, writing and decisions that were simply never made.

Day 31–60: tune

Day 61–90: scale

On sequencing. Voucher attribution appears in the second month rather than the third, and deliberately so. It has to be live before the first creator code is issued, not after — because the first time a creator’s sale is captured by a coupon partner at the last click, the relationship is usually over, and in this category those relationships are the entire growth plan.

12.1 Standing cadence

12.2 Retail calendar moments

This category’s calendar is severely front-loaded: the single largest demand moment arrives in January, which means partner briefing has to happen the previous November. A programme that launches into its own peak has already missed it — the demand arrives before the partners do, and no amount of budget compresses an eight-week editorial lead time.

No — but it needed a fortnight of building, not a rescue. Nothing in this programme was broken by history. There was no leakage to unwind, no aged validation queue, no damaged partner relationship, because nothing had happened yet. That is an unusual and genuinely valuable position: every finding in this audit could be fixed before it had cost anything.

What held it back. Three things. First, the operational shell was hollow — no welcome email, no offer, no feed, no creative, no communication. A partner who joined found a programme that looked live and offered nothing to act on. Second, the partner mix had assembled itself rather than being chosen, filling with demand-harvesters in a category where only demand-generators can grow the programme. Third, commercial and regulatory exposure sat unaddressed: an unpaid invoice that could take the links offline, a flat rate paid on a tax-inclusive base to every partner type identically, and no health-claim rules for a product positioned adjacent to prescription medication.

The constraint that shapes everything else. This category cannot advertise. Paid social and search are restricted for anything adjacent to GLP-1 medication — accounts get disapproved, landing pages flagged. That single fact changes the weighting of every recommendation in this document. The affiliate programme was not one acquisition channel being optimised alongside others. It was the only channel available, which means partner-mix drift is not a quality issue to be tidied up later. It is the growth plan failing quietly.

The first five tasks, and why that order. One: settle the invoice and set up direct debit, because everything else is irrelevant if the tracking links go offline. Two: publish the health-claim and disclosure terms, because that is the only finding carrying regulatory rather than commercial risk. Three: restructure commission before volume arrives, because doing it now is free and doing it later costs relationships under notice rules. Four: write the welcome email, because it is the only guaranteed moment of partner attention and seventy-five partners had already been wasted. Five: complete the profile, because it is the cheapest possible improvement to how the programme presents itself.

What the brand should not do yet. No paid marketplace tenancy at zero revenue — placements should be judged on proven conversion, not bought on hope. No further sub-network or low-relevance coupon growth. No creator codes issued until voucher attribution is enabled, because the first creator whose sale is captured at the last click by a coupon partner will not come back. And no public sharing of the auto-approving invite link, given the base already contains more low-value partners than it should.

What to review in 30 days. Is the invoice settled and the direct debit active? Are the health-claim terms published? Is commission differentiated on a tax-exclusive base? Is the welcome email live and is the profile complete? Is there a live, tested offer? Has the demand-generating share of new joins started to move?

What requires external evidence before a final conclusion. An end-to-end test transaction is needed to confirm tracking rather than assume it. The brand’s approved claims language is required before terms can be published. Order value and bestseller data are needed to complete the profile and the welcome email. Legitimacy verification is required on the near-identical sub-network cluster before any of them transacts. And subscription attach data by acquiring partner will be needed before creator performance can be judged on anything more than first-order revenue — which in a subscription category is the least informative half of the picture.