Seven months of clicks. Not one recorded sale.

The programme had launched some seven months before the audit as a weight-management supplement programme on a premium plan tier, with access to enhanced reporting across commission group, product, device, publisher comparison, journey path, funnel and sector benchmarking. Payment status was green — online, within credit limit, accelerated payments. Link status was online. On paper the foundations were in place.

In practice the programme was entirely non-functional. Zero transactions had been recorded across its entire lifetime — not a single pending, validated or declined sale. The root cause was a conversion tracking failure. The platform’s tracking tag was present on the site and recording publisher clicks successfully, several thousand per month across the two months before the audit. But the conversion pixel was not executing when a customer completed an order, because the site’s checkout ran as a nine-step quiz-style funnel that remained on the same page throughout. A standard sale-tracking pixel has no page load to fire on in that architecture. The result was that every affiliate-driven sale went unrecorded, every publisher earned nothing, and the programme’s network index sat at zero.

The risks were immediate and structural. First, the tracking failure was haemorrhaging both revenue and publisher trust — publishers who have driven clicks for months with no commission eventually stop, and they do not come back. Second, the terms contained a direct contradiction: the general terms prohibited paid search advertising without prior written authorisation, while the PPC policy tab had every policy toggle set to permit PPC, direct linking and brand-name bidding. That is a compliance exposure that had to be resolved before the programme became functional, because PPC publishers would generate significant brand-term traffic the moment transactions started flowing. Third, more than fifty publisher applications sat in a pending queue with no evidence of review, and the backlog was visibly growing across the audit window. Fourth, the creative library was critically thin: fourteen text links and a single non-standard image banner, with no standard display sizes at all. Fifth, no communications had been sent to publishers in over six months, no triggered automations existed, no templates were saved, no offers or voucher codes were available, the product feed was empty, and branding guidelines were entirely absent despite the product being a health supplement subject to advertising regulation.

The quickest wins were the PPC terms alignment, under an hour of work; the validation period discrepancy, where the written terms and the system setting disagreed by ten days in a way guaranteed to cause disputes once transactions flowed; and the pending approval backlog, a few hours of review. The strategic priority, however, was singular: fix conversion tracking. Every other optimisation was blocked behind it.

The good news, and the reason the recovery plan is credible: the programme had genuine strengths justifying investment. Despite zero transactions and zero commission ever paid, more than 378 publishers had joined and been approved, with new applications still arriving daily. The premium plan tier provided reporting and tooling most competitors do not have. The website itself was well designed with clear pricing and prominent calls to action — and the quiz funnel that broke the tracking turned out, once traffic was routed into it properly, to be the programme’s strongest conversion asset rather than its weakest link.

There was one more finding that shaped everything that followed. This category has a specific failure mode: weight-loss programmes tend to die in month six, when claim complaints start arriving and publishers quietly delete the content rather than defend it. This programme had no branding guidelines, no approved claims list and no disclosure standard — which meant it was built to fail exactly that way. The compliance pack had to come before the recruitment, not after it.

2A. Metric scorecard

Several metrics read as zero rather than as a rating, because they genuinely were. Benchmarks are AME Reference Library values for the health and supplements sub-sector and are not client data.

2B. Area scorecard

A publisher base that assembled itself and stayed. Despite zero transactions and not one commission payment ever made, more than 378 publishers had joined and been approved, with new applications still arriving daily at the time of audit. That is genuinely remarkable and it is the programme’s single most valuable asset: a ready-made base waiting to be activated the moment tracking works. Most recovery projects have to recruit before they can grow. This one had to fix a pixel and send an email.

Premium platform access already paid for. The programme sat on a plan tier providing enhanced reporting across commission group, product, device, publisher comparison, journey path, funnel and sector benchmarking — tooling most competitors on entry-tier plans cannot match. None of it had been used, but none of it needed procuring either. The capability was bought and idle.

A well-built site with a strong conversion funnel. The website was well designed with clear pricing, a transparent multi-buy structure and prominent calls to action. The nine-step quiz-style checkout is the interesting part: it was the direct cause of the tracking failure, and it was also the best conversion mechanism in the programme. Quiz funnels in this category qualify intent and personalise the recommendation before asking for payment, and they convert substantially better than a conventional product page for exactly that reason. The finding here is not “replace the funnel”. It is “make the tracking work with the funnel you have”, and then route partner traffic into it deliberately.

Healthy payment and link status. Payment status was green, within credit limit, with accelerated payments enabled, and link status was online. In a category where programmes routinely carry billing warnings, having none is worth protecting — and it meant that once transactions began, publishers would be paid promptly and the index would recover quickly from zero.

An underserved category position. The niche was underserved in its market, with a clear product proposition and transparent pricing. The demand existed and the programme was not competing against a crowded field of well-run rivals. Once functional, it had room.

5.1 Overview

The programme had 383 publisher records — 375 active, seven departed, one rejected — and more than fifty further applications sitting unactioned in the pending queue. With zero transactions recorded, no revenue, commission, cost per acquisition or active rate could be calculated for any of them. The analysis is therefore about the shape of the base rather than its performance.

The single most important fact in this section is that seven publishers had left. Given that none of the 378 had ever earned a single commission across seven months, the more surprising number is that 375 stayed. That is the asset the recovery was built on, and it had a limited shelf life.

5.2 Type distribution

The variety of publisher types present was reasonable — editorial, content creators, communities, comparison, shopping directory and mobile traffic were all represented. The problem was not the mix. It was that no partner of any type could earn anything, and most types additionally lacked the basic materials to promote: no feed for the comparison and shopping partners, no display creative for the display partners, no offer for anyone, and no claims guidance for the content and creator partners who carry the most regulatory exposure.

Four or more sub-networks had been approved without vetting, which represents both a concentration risk and a compliance concern — sub-networks bring opaque downstream traffic, and the programme had no terms requiring disclosure.

5.3 Concentration

Revenue concentration cannot be measured with zero transactions. As with any pre-revenue programme this is an opportunity rather than a gap: the programme could still determine which partner types come to dominate its revenue before any concentration set in. The specific goal here was to ensure the eventual top tier was populated by content, community and creator partners — the types that can carry a regulated category responsibly — rather than by unvetted sub-networks whose downstream traffic nobody can see or stand behind.

5.4 Device performance

Device performance was not measurable with no transactions recorded. The relevant forward-looking point concerns the quiz funnel: a multi-step qualification flow behaves very differently on mobile than on desktop, and completion rate by device is the single metric that should be instrumented first once tracking works. In this category the majority of partner traffic arrives on mobile, and a funnel that loses people at step four on a small screen will cap the entire programme invisibly.

5.5 Pending approvals

More than fifty applications sat pending with no evidence of review, and the queue grew measurably across the twelve days of the audit alone. This is the clearest single indicator of the programme’s dormancy: inbound interest was arriving daily and nobody was there to receive it.

The recommendation was to clear the backlog with proper type and market scrutiny — not to bulk-approve it. Given that four unvetted sub-networks were already in the base, adding fifty more partners without review would have compounded the problem the audit had just identified.

With no transactions recorded, dispositions rest on type, relevance and risk rather than performance. The matrix below is grouped by partner type rather than named partner, because at this stage the type-level decision is the one that matters.

There was no relationship management of any kind. No top-publisher identification, no segmentation, no publisher tags, no outreach, no review cadence. With no transactions and no communication in over six months, there was in a literal sense no relationship to manage — only a list.

That is a harsher position than it first appears, because relationship damage had already occurred silently. Every one of the 378 approved publishers had at some point decided this programme was worth their traffic, sent some, and received nothing. Seven had formally left. The rest had almost certainly deprioritised the programme without telling anyone, which is what publishers do.

The relaunch sequence matters more than the segmentation. The conventional advice — tier the base, establish a cadence, brief the top twenty — is correct but premature. The first relationship action had to be an honest relaunch communication sent after tracking was verified, not before: confirming that tracking had been broken, that it was now fixed, that the fix had been tested, and that commission would flow from a stated date. A programme that has taken publishers’ traffic for seven months without paying for it does not get to open with a promotional newsletter.

Then the structure. Once transactions begin recording, tag the base by type, tier and compliance status; identify the emerging top tier by verified revenue; establish monthly contact with the leading content and community partners; and put a quarterly review in place. The compliance pack should be reissued to every partner at relaunch, not just to new joiners, because the existing base has been promoting without guidance for months.

Competitor exposure. In weight management, content and creator partners promote multiple brands and are unusually sensitive to compliance risk — a single claim complaint can cost them a platform account, not just a commission. A programme offering no claims guidance is not merely unhelpful; it is a liability those partners will choose to avoid once they understand it. That is why the compliance pack sits ahead of recruitment in every plan in this document.

Partner discovery tooling was functional and, at the time of the audit, was actively displaying quality recommendations on screen — relevant partners in the right market and the right categories. Zero invitations had ever been sent. The tool was working perfectly and had never been used.

The publisher base had accumulated entirely through inbound application, which explains both its size and its composition. Inbound growth attracts whoever finds you; it does not produce a chosen mix. It also explains the four unvetted sub-networks: they applied, nobody reviewed, they were approved.

Recruitment must wait for two gates. This is the single most important sequencing point in the audit, and it runs counter to the instinct to fix a quiet programme by adding partners. Recruiting into a programme where the conversion tag does not fire produces new publishers who earn nothing and leave — converting an inbound asset into a burned reputation. And recruiting into a regulated category with no approved claims list produces exactly the month-six collapse this category is known for: partners publish, complaints arrive, partners have no defensible language to point at, content comes down.

So: tracking first, compliance pack second, recruitment third. Not because the recruitment is less important, but because doing it earlier destroys its own results.

Where to recruit once the gates are cleared. The priority types are content and editorial publishers writing credibly in the category, community operators with engaged memberships, and creators with genuine personal experience. In this category the partners worth having are the ones whose audiences trust them precisely because they are cautious — and cautious partners are exactly the ones who will read the compliance pack before they say yes. A published claims standard is not overhead in this category. It is the recruitment pitch.

9.1 Current state

The programme ran a 4% base commission against a sector benchmark of 10–20%. The profile advertised a performance bonus above that base, but commission manager access was denied during the audit, so no tiered structure could be verified — and a rate structure that cannot be inspected cannot be managed, which is itself a finding.

At the base rate, theoretical earnings per click fell far below the sector benchmark range even under optimistic conversion assumptions. That matters in a specific way here: once tracking is fixed and transactions begin recording, publishers will finally be able to calculate what this programme pays them. If the answer is a quarter of what the category norm offers, the tracking fix will simply reveal an uncompetitive programme rather than unlock a competitive one.

There was no differentiation of any kind. A community partner producing original, carefully worded content in a regulated category earned the same 4% as an unvetted sub-network passing through opaque downstream traffic.

9.2 Recommended architecture

9.3 Budget impact

With zero recorded revenue there is no current commission spend, so restructuring costs nothing today and prevents a great deal later. That is the same argument that applies to every pre-revenue programme, but it lands harder here because of the sequencing: the rate change and the tracking fix should land together, so that the relaunch communication to 378 dormant publishers can say two things at once — tracking now works, and the rate is now competitive. Sending that message twice, months apart, wastes the only genuine attention moment the recovery gets.

Modelled forward at the click volumes the programme was already generating and at sector-benchmark conversion rates, the difference between the current base and a sector-floor rate is the difference between a programme publishers tolerate and one they actively choose. The activation bonus is the cheapest line in the table and, against a base of 378 partners who have never earned a commission, likely the highest-returning.

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: make it functional and make it safe

The sequence in this plan is unusually rigid and deliberately so. Recruitment into a programme that cannot track produces publishers who earn nothing and leave. Recruitment into a regulated category with no claims standard produces the month-six collapse the category is known for. Both gates come first.

Day 31–60: tune

Day 61–90: scale

On launching after the peak. This category’s demand peaks in January. A programme that is not fully functional, compliant and stocked with briefed partners by the preceding November cannot serve that peak, and launching into it half-built wastes the year’s best traffic on a programme that is not ready to convert it. Launching deliberately after the peak, with the compliance pack and tracking verified first, is the better trade — it costs one season and buys a channel that survives.

12.1 Standing cadence

The first line of this calendar is not standard. A daily check that transactions are recording would be unnecessary in most programmes; in one that ran seven months without noticing its conversion tag had never fired, it is the most important line in the table.

12.2 Retail calendar moments

No — it was not ready to operate, let alone scale. A programme that has recorded zero transactions across seven months is not underperforming. It is switched off in a way that looks switched on, which is considerably more dangerous, because a dashboard showing clicks and no sales is indistinguishable from a programme with a conversion problem. Nobody looked closely enough to tell the difference for seven months.

What held it back. One thing, absolutely, and then everything else. The conversion tag never fired because the checkout ran as a single-page quiz funnel with no page load for a standard pixel to execute on. Until that was fixed, commission increases, creative investment, recruitment and seasonal campaigning would all have produced exactly the return they had produced so far, which was nothing. Behind it sat a second gate that mattered almost as much: no branding guidelines, no approved claims and no disclosure standard for a regulated weight-management product.

Why the compliance gate ranks with the tracking gate. Weight-management programmes have a characteristic failure mode. They grow, and then in month six the claim complaints start arriving, and publishers — who were never given approved language, never told what they could and could not say, and have no assurance the brand will stand behind them — quietly take the content down rather than fight. The programme does not collapse loudly; it unwinds. Publishers do not avoid this category because it converts badly. They avoid it because they have been burned, and a programme with no claims standard is telling them, accurately, that it will burn them too.

The first five tasks, and why that order. One: fix conversion tracking and verify it with a real test order, because nothing else has any effect until it works. Two: publish the compliance pack, because it must exist before a single new partner is recruited or the recruitment produces a month-six collapse. Three: resolve the PPC contradiction, because the exposure crystallises the moment transactions begin. Four: reconcile the validation discrepancy, for the same reason. Five: clear the pending backlog, because inbound interest was arriving daily and being thrown away.

What the brand should not do yet. No recruitment until tracking is verified and the compliance pack is published. No promotional communication to the existing base until a test transaction confirms commission will actually flow — sending a promotional newsletter to 378 publishers who have earned nothing for seven months would be worse than continued silence. No seasonal campaign investment into a peak the programme is not equipped to serve; launching after the peak, properly built, is the better trade.

What to review in 30 days. Are transactions recording with correct value and attribution? Is the compliance pack published and issued to the entire existing base, not just new joiners? Do the terms and the system state the same validation period? Is the PPC position consistent? Is the pending queue clear? Has the base commission moved into the sector band?

What requires external evidence before a final conclusion. An end-to-end test order is required to confirm tracking rather than assume it — and given the history here, assumption is exactly what caused the problem. The brand’s approved claims language is needed before the compliance pack can be published. Commission manager access must be restored before any tier structure can be verified. Funnel completion data by device is needed before the quiz architecture can be optimised. And traffic-source disclosure is required from the unvetted sub-networks before the first transaction records against any of them.