NeoMarketing: The Zero AdWaste Machine

Published May 23, 2026

How Atrium and Meridian end the Reacquisition Tax

1

Marketing is one of the largest line items on a modern brand’s P&L, and one of the least examined. Ten percent of revenue, sometimes more, flows out of the business through a combination of paid media, adtech fees, martech subscriptions, agency retainers, and the quiet losses that nobody lines up and measures — customers drifting, reactivation campaigns buying back what was already owned, attention decaying between quarters. The scale of the number makes it difficult to see. A brand cannot easily ask whether the tenth of its revenue it spends on marketing is producing growth or merely preventing collapse. The question has become unfashionable to pose, which is precisely why it needs to be asked again.

NeoMarketing is the operating system that poses it, and offers a structural answer. It begins from a single diagnosis: the $500 billion global AdWaste crisis is not a bug in martech. It is the natural behaviour of a system where customer engagement platforms lose customers and adtech platforms monetise the loss, with brands absorbing the cost and platforms capturing the margin. The leak is not an accident. It is the architecture.

This essay makes the case for a different architecture — two engines running on a shared substrate, with a commercial model that rewards vendors only when brands actually grow. Atrium is the attention engine, serving the drifting 80% through inbox-native attention with Zero CPM economics. Meridian is the outcomes engine, serving the revenue-generating 20% through Context Graphs, BrandTwins, and Alpha pricing. Together, they operationalise three commitments — Never Lose Customers, Never Pay Twice, Never Buy Fixed — and produce a closed system where customers are acquired once, retained continuously, and monetised without leakage.

NeoMarketing is not better martech. It is the machine designed to make AdWaste structurally impossible.

**

The Problem and the Doctrine

Why AdWaste is structural, not accidental — and the three commitments that name the refusal.

1. The $500 billion Reacquisition Tax is a tax, not an inefficiency. AdWaste is not a small leak. Globally, brands spend roughly $500 billion per year reacquiring customers they already owned — paying Google, Meta, Amazon, and the programmatic ecosystem to reach people whose email addresses, phone numbers, and transaction histories already sit inside the brand’s systems. A large share of what is called “acquisition” is in fact recovery. The category has normalised the absurdity. CFOs have learned to budget for it. CMOs have learned to report around it. Agencies have learned to optimise it. Platforms have learned to monetise it. What should have been recognised as a recurring tax has become invisible through repetition. But routine does not make something rational. It only makes it harder to notice. Once the tax is seen clearly, it becomes impossible to unsee — and impossible to keep funding without asking whether the brand is buying growth or merely paying rent on customers it once owned. The first step in NeoMarketing is not invention. It is recognition: AdWaste is the Reacquisition Tax hiding inside what marketing now calls growth.
2. Martech and adtech do not solve the leak. They collaborate in creating it. Customer engagement platforms are paid to send messages, manage journeys, orchestrate channels, and increase marketing activity. Adtech platforms are paid to recover the customers who stopped responding to those messages. Both sides are rewarded by the same underlying failure: customer attention decay. The CE vendor earns more when the brand sends more, whether or not sending works. The adtech platform earns more when the brand pays more to reacquire, whether or not reacquisition should have been necessary. Neither side is structurally punished when a relationship weakens. On the contrary, each side has a revenue model that tolerates or even benefits from drift. This does not require conspiracy. It is the predictable outcome of input-priced economics. One side monetises the attempt to engage; the other monetises the cost of having failed to engage. The brand absorbs both bills. The leak is not a failure of execution. It is a commercial architecture in which customer loss is profitable for everyone except the brand.
3. Never Lose Customers. Never Pay Twice. Never Buy Fixed. These are not slogans appended after the fact. They are the doctrine. Never Lose Customers names the mission: attention must be treated as inventory, drift must be detected early, and customer transitions must be managed before they become expensive. Silent drift — customers going quiet rather than unsubscribing — is the most expensive failure in marketing, because it is invisible until the reacquisition bill arrives. Never Pay Twice names the enemy: any conversion from a customer already reachable through owned channels should cost close to nothing. Paid spend should be for genuine discovery, not for recovering relationships that were allowed to decay. Never Buy Fixed names the mechanism: reject input-priced martech that gets paid whether or not outcomes materialise, and replace it with Alpha-linked economics where the vendor participates only when uplift is real. Each NEVER addresses a structural failure. Each has a direct economic consequence. Each imposes a line in the sand between the old category and the new one. A doctrine is not a list of features. It is a commitment to refuse certain behaviours in the present, and to build systems that make those refusals operational.
4. BRN and SNR are the two lenses that make the problem visible. NeoMarketing organises customers and messages through two orthogonal frameworks. BRN — Best, Rest, Next — describes where the customer sits in the relationship lifecycle. The Best (the top 20%) are the profitable core. The Rest (around 60%) are the drifting majority — not formally lost, but no longer paying real attention. The Next are unacquired. SNR — Sell, Notify, Relate — describes what the brand is saying. Most brands do Sell and Notify tolerably well: promotions, offers, product pushes, receipts, confirmations, reminders. Almost no brand does Relate well — communication that exists for the relationship independent of the next transaction. That missing Relate layer is precisely where attention decays. When the Rest segment receives only Sell and Notify, it becomes the staging ground for future reacquisition spend. BRN tells you where the customer is. SNR tells you why they are drifting there. The Reacquisition Tax lives in the cell most brands have left empty for years: Rest × Relate. That is where NeoMarketing begins its work.
5. One engine cannot solve a structural leak. NeoMarketing requires two. Atrium and Meridian exist because the problem is not singular. Atrium is the attention engine — it serves Rest and Next customers through inbox-native participation, attention economics, and Zero CPM communication. Its promise is simple and provocative: Zero the CAC. Meridian is the outcomes engine — it serves Best customers through Context Graphs, BrandTwins, decisioning, and Alpha pricing. Its promise is equally simple: Max the LTV. The two engines are complementary, not competitive. Meridian monetises outcomes. Atrium monetises attention. Without attention, there are no outcomes to monetise. Without outcomes, attention has no commercial purpose. Neither engine alone is sufficient. Meridian without Atrium leaves the drifting majority unserved and the Reacquisition Tax intact. Atrium without Meridian restores attention but leaves the highest-value customers managed by a weaker operating model. NeoMarketing is not a product suite made to look elegant. It is a dual-engine machine because no single engine can eliminate AdWaste across the full customer lifecycle.

2

Atrium: The Attention Engine

From primitives to products to network — how the inbox becomes a media asset.

6. Magnets, Mu, and Markets are the three primitives of attention. Atrium’s inner engine rests on three mutually dependent primitives. Magnets are 30-to-60-second interactive units embedded inside the inbox — quizzes, predictions, forks, polls, mini-games. The customer does not receive the message; they enter it, choose, tap, and see a result. Participation replaces consumption. Magnets produce revealed preference data that declared preferences from forms and surveys cannot match. Mu is the attention currency earned through participation — not given as a welcome gift, but accumulated through repeated engagement. A Mu balance is visible proof that attention has been paid, appearing in the subject line so the currency works before the content does. Markets is the burn destination. Without somewhere meaningful to spend Mu, the balance becomes decorative. WePredict — closed-group prediction markets run inside WhatsApp groups, office teams, family chats — gives Mu weight. Correct predictors split the pool at resolution. Losing Mu hurts because the balance took weeks to earn, and being wrong in front of people who know you is real consequence even without cash. Remove any one of the three and the loop collapses. Three primitives. One system. Magnets produce participation. Mu accumulates it. Markets gives it weight. Every product surface above them is built on this foundation.
7. NeoMails are a new class of email built on the primitives. Primitives alone are not a product. NeoMails are the product — a new class of email that is neither promotional nor transactional, but the Relate channel that SNR has been missing all along. Each NeoMail contains four elements in a specific order. The Magnet appears first — the reason to open, resolved inside 60 seconds. Mu appears in the subject line and the footer — the reason to return tomorrow. The BrandBlock appears after attention is earned, not instead of it — this is where the brand’s actual message lives, delivered to a reader who has already agreed to engage. And the ActionAd funds the send. NeoMails create cadence where campaigns create noise. They create habit where broadcasts create fatigue. They create signal where traditional email creates only opens and clicks. Crucially, they operate as a daily surface — not a weekly newsletter, not a monthly campaign, but a daily touch that compounds into a relationship. This cadence is the precondition for everything else Atrium can do. NeoMails are not a better email. They are a third category. Promotional emails interrupt. Transactional emails inform. NeoMails earn attention — and then the brand gets to speak.
8. ActionAds turn attention into revenue inside the inbox. Every NeoMail carries the potential for an ActionAd — an in-email action unit that completes inside the inbox, on authenticated identity, at the peak of attention. Advertisers pay for the completed action, not the impression. Two formats power the system. The One-Tap Subscribe ActionAd lets a customer subscribe to another brand’s NeoMails with a single tap, pre-filled from the platform identity — no landing page, no form, no confirmation loop, consent logged at the moment of interaction. The subscribing brand pays a transfer fee per confirmed NeoMail-active subscriber, a fraction of paid media cost. The form-fill ActionAd generates verified leads inside the email itself — contact details, a preference, a qualification — without the customer leaving. The advertiser pays a cost-per-lead fee, split between the platform and the publishing brand. That split is what closes the Zero CPM loop: instead of the brand funding a Relate message as a cost centre, the message funds itself. The relationship layer, which was previously invisible to the marketing P&L, now has an economic address. ActionAds resolve the oldest objection to email. The channel is no longer cheap but ineffective. It is measurably effective and structurally profitable.
9. NeoNet turns a single brand’s surface into a cooperative media network. One brand’s NeoMail surface is valuable. A network of cooperating brands’ NeoMail surfaces is transformational. NeoNet is the cooperative layer — an ad network where brands exchange access to their own active NeoMail audiences, first-party for first-party, with no auction and no platform intermediary taking rent. This matters because attention is not evenly distributed. A customer may be cold for Brand A and still highly active in Brand B’s NeoMails. In the old model, Brand A had to go to a paid platform to reacquire that customer at auction price. In the NeoNet model, Brand A reaches the customer through Brand B’s active first-party inbox surface at cooperative cost. Recovery points backward: a drifting customer is reached through another brand’s engaged audience. Acquisition points forward: a genuinely new customer discovers Brand A inside Brand B’s inbox and subscribes in a single tap. An identity enters NeoNet only after demonstrating live engagement — opening at least one NeoMail — so the network runs on present attention, not historical list volume. Every other ad network is adversarial by design: brands bid against each other, platforms extract margin, brands do not own the relationship. NeoNet is cooperative by design. It is the first ad network built for the brands, not on top of them.
10. Atrium is the Inbox Media Network — a new category of commerce media. Step back from the components and what emerges is a new category, parallel to but distinct from what already exists. Retail Media Networks monetise shopper attention at the moment of purchase intent, on the platform’s own surface — Amazon, Walmart, Instacart. The asset is purchase behaviour; access requires being a marketplace. Commerce Media Networks extend the same first-party shopper identity off-site — broader access, same intent-adjacent logic. The Inbox Media Network monetises relationship attention — between transactions, across cooperating brands, through action-based units that complete inside the inbox. The asset is the brand-customer relationship and the authenticated identity that underpins it. McKinsey projects commerce media will exceed $100 billion in US ad spend by 2029, growing faster than both traditional advertising and digital as a whole. Most brands assumed this model was unavailable to them because they lacked a marketplace or on-site inventory. But every brand with a NeoMail-active base has the asset. NeoMails create the surface. ActionAds monetise it. NeoNet extends it across brands. The database becomes a media asset — not because it contains contacts, but because it contains live, repeated, monetisable attention. The next major ad category may not be built on a new platform at all. It has been hiding in plain sight, inside the inbox, waiting for the product architecture that makes it visible.

3

Meridian: The Outcomes Engine

Best customers as portfolio positions — a five-layer architecture priced on Alpha.

11. Best customers do not need more automation. They need a different operating system. Traditional customer engagement reached its limits not because it lacked features, but because its logic was wrong for the highest-value relationships. Best customers are the top 20% of a brand’s base — the relationships where the economics are deepest through cross-sell, repeat purchase, preference formation, advocacy, and long-duration value. They are also the most sensitive to irrelevance, mistiming, fatigue, and generic treatment. More workflows do not solve the Best customer problem. Workflows can automate known branches. They cannot continuously reason about an individual customer’s evolving state, value trajectory, tolerance for outreach, or next-best investment. A marketer can design fifty segments; a marketer cannot continuously manage ten thousand high-value relationships as markets of one. This is why Meridian exists. It starts where customer engagement platforms stop: at the point where the brand no longer wants more sophisticated machinery for sending messages, but a system capable of making decisions at N=1 scale with measurable commercial accountability. The ceiling of CE appears precisely where relationships become too valuable to be managed as segments, but too numerous to be managed by humans alone.
12. Meridian’s five-layer stack turns data into commercial decisioning. The architecture is not decorative. Layer 1 is the familiar commodity layer of data sources — transactional, behavioural, world, nano — organised through the TWIN framework. Necessary, but not defensible. Layer 2 is where Meridian becomes structurally different: Context Graphs. The Customer Context Graph models evolving customer state — preferences, momentum, fatigue, trajectory. The Product Context Graph holds offers, substitutes, complements, margin realities, and constraints. The Decision Trace Graph records what was done, why, what alternatives were considered, and what actually happened. That third graph is the compounding layer. Layer 3 holds the Functional Agents — Insights, Segmentation, Content, BrandTwins — reading from the graphs and producing structured outputs. Layer 4 is the Decisioning Agent, the brain that orchestrates those outputs and decides what happens next for each Best customer. This is where Alpha is generated. Layer 5 is the Co-Marketer, the human-facing surface through which review, explanation, approval, and oversight occur. Context Graphs run under Atrium too, but at cohort resolution; Meridian uses them at individual resolution. Same infrastructure, different views of the same customer. Data is table stakes. The architectural breakthrough is not that Meridian has agents — it is that the agents sit on top of working memory and decision traces, inside a system where the core intelligence is measured by Alpha, not by activity.
13. Three phases make autonomy commercially believable. One reason AI rhetoric in marketing feels ungrounded is that it leaps directly to the endpoint. Meridian does not. It progresses through three phases, each a distinct product with distinct pricing. In Phase 1, the CSM is assisted by M-Agents — Meridian’s collective of task-specific marketing agents. AI augments. The human proves the method on a defined Best customer cohort against a baseline. Pricing is hybrid — Alpha1 — a modest base fee plus Alpha upside. The method is not yet proven at scale, the Decisioning Agent does not yet have enough traces to operate autonomously, and caution is appropriate on both sides. In Phase 2, the CSM becomes a Martech Growth Engineer and the Alpha Agent begins to absorb more of the execution burden under MGE supervision. Capacity rises dramatically — one MGE can now manage roughly 10X more brand accounts than before, because the agent handles execution and the human focuses on calibration and oversight. Pricing shifts to Alpha1.5 — reduced base, larger Alpha share. In Phase 3, the Autonomous Co-Marketer runs the work and the human governs outcomes, exceptions, and trust boundaries. Pricing moves to Alpha2 — 100% variable, paid on Alpha alone. Each phase earns the right to the next. Trust compounds operationally; commercial terms compound commercially. A radical end state becomes credible only when the path to it is staged, operationally earned, and commercially sequenced.
14. The Alpha Agent is trainable because the system produces its own corpus. The Alpha Agent is not magic. It is the Decisioning Agent operating inside the Alpha pricing structure under MGE supervision, trained on a loop that Meridian itself creates through operation. Context Graphs hold the input state — who the customer is, how the relationship is evolving. Decision Trace Graphs hold the judgement layer — chosen action, rationale, rejected alternatives, expected outcome, actual outcome. Together they form a proprietary training set of something far more valuable than generic marketing data: expert decisions tied to measured business consequences. Supervised fine-tuning teaches the Alpha Agent to approximate strong MGE judgement. Reinforcement learning tunes it against actual Alpha — actions that generated uplift get positive reward; actions that damaged CRR (Click Retention Rate) or triggered fatigue get negative signal. Weekly MGE calibration loops keep the agent current as brand contexts shift. Competitors can license stronger base models tomorrow. What they cannot do overnight — or in a year — is replicate years of decision traces accumulated from real MGE work on real Best customer cohorts at real brands, with real Alpha outcomes measured against real Beta baselines. The architecture is open; the corpus is proprietary. The moat is the traces, not the model. Any competent team can build an agent. Not every team can build a corpus of proven decisions that makes the agent measurably better every quarter.
15. Alpha pricing changes the meaning of a martech vendor. The most radical element of Meridian is commercial, not technical. Traditional software sells access to capability. Meridian sells participation in realised uplift above a pre-agreed baseline. The brand’s existing revenue trajectory is Beta — agreed and locked before execution begins. Meridian’s job is to generate Alpha above it. Carry is calculated on Alpha alone. If Alpha is zero, Carry is zero. No retainer. No base fee. No safety net on the vendor’s side. The analogy is hedge fund economics with one important improvement. Hedge funds charge “2 and 20” — a 2% management fee regardless of performance, plus 20% carried interest. The manager earns the 2% whether they beat the market or not. Meridian removes the 2%. There is only the equivalent of carry, calculated against a fixed benchmark. This is not a cosmetic change to pricing. It changes incentives, posture, trust, procurement, and internal accountability on both sides. It also changes category identity. A vendor that charges fixed fees regardless of whether retention improves is a supplier of tools. A vendor that earns only when measurable uplift appears has stepped into a different relationship entirely. Meridian is not interesting because it has AI. It is interesting because it is willing to put that AI inside a contract where the vendor’s own economics depend on the brand’s outcomes. That is what makes it an outcomes engine rather than a CE upgrade.

4

The Machine

Why the two engines compound into something greater than their parts — and why this is not better martech, but what replaces martech.

16. The two engines are connected by shared substrate, and that is what makes NeoMarketing a system. Atrium and Meridian are not merely adjacent products. They sit on common foundations. Context Graphs operate underneath both engines, at different resolutions — cohort-oriented and attention-sensitive under Atrium, individual and decision-rich under Meridian. The same infrastructure, different views of the same person. Mu travels with the customer across experiences, providing visible continuity of attention and participation. The broader intelligence substrate — ArtificialPeople, world-model-based consumer archetypes — powers BrandTwins in Meridian at the individual level and cold-start personalisation in Atrium at the cohort level. The same archetype intelligence that informs a BrandTwin for a Best customer informs the initial Magnet selection for a newly reactivated Rest customer. This matters because it prevents fragmentation. A Rest customer reactivated through Atrium does not arrive in Meridian as if from nowhere. The system already has memory. A Best customer starting to drift does not disappear into a separate recovery universe. The same substrate carries the relationship context through the transition. Without shared substrate, NeoMarketing would be two clever products. With it, it becomes a closed system in which memory, state, and attention survive movement between engines.
17. Customers move between engines, and the system should not forget them when they move. Real customers do not sit permanently inside neat segments. A Best customer can begin fading quietly. A Rest customer can be reactivated through consistent NeoMail engagement. A Next customer can enter through one surface and deepen through another. If the infrastructure treats these as separate worlds, the brand ends up rebuilding understanding at each transition — which is precisely how value leaks. NeoMarketing is designed to make those transitions native. Atrium captures and restores attention in the low-cost, high-scale part of the relationship. When a Rest customer’s engagement deepens — Mu balance growing, streaks forming, Magnet participation becoming habitual, spending trajectory turning upward — the BrandTwin activates and Meridian takes over at N=1. The customer has graduated, and the architecture has absorbed the transition without rebuilding anything. When a Best customer’s engagement begins to decay — signals in the Context Graph flagging fatigue, silence, reduced category affinity — the BrandTwin raises the alert before dormancy is visible, and the customer is moved back into Atrium’s attention loop. This happens before full reacquisition becomes necessary. The customer moves. The memory should not. That is what makes the dual-engine model operationally powerful rather than merely conceptually elegant.
18. The engines compound because attention and outcomes reinforce each other. Atrium and Meridian are not simply complementary. They form a flywheel. Atrium restores or builds low-cost attention at scale, especially among the Rest. Reactivated customers graduate into Best as engagement deepens. Meridian then converts these Best customers into higher LTV through N=1 treatment and autonomous decisioning. That higher LTV makes the economics of attention investment more attractive, expanding the Relate channel and producing more NeoMails, more Magnet participation, more Mu activity, more Market behaviour, more signal. Richer signal sharpens ArtificialPeople. Sharper archetypes improve BrandTwins and cold-start personalisation. Better personalisation deepens engagement further. Each rotation produces a slightly better system than the rotation before. Meridian without Atrium lacks its upstream attention machinery and pays the full Reacquisition Tax. Atrium without Meridian restores attention but leaves downstream value creation underdeveloped. Together, they produce a structure in which customers are acquired once, retained continuously, and grown without the leakage that has defined the old martech-adtech bargain. The compounding lives in the interaction between the engines, not in any single component. Each engine is strong alone. Together, they produce economics neither could achieve on its own.
19. The market opportunity is not the martech market. It is the value locked inside AdWaste. The conventional software market for martech is approximately $50 billion globally — large, crowded, increasingly commoditised, squeezed by consolidation and AI pressure. Growth within this market is essentially zero-sum: vendor X wins a deal that vendor Y loses. NeoMarketing does not compete in this market; it addresses a different pool. The AdWaste crisis is ten times larger at $500 billion — the value trapped in reacquisition, poor retention, broken customer memory, weak outcome accountability, and channel models that decay by design. That is the pool Atrium’s Zero CPM economics attack directly. Beyond that sits participation in the transaction economics of e-commerce itself through Meridian’s Alpha pricing — a percentage of incremental revenue compounding as Carry across a portfolio of brand relationships, quarter after quarter. Traditional software asks, “how much will the buyer pay for the tool?” NeoMarketing asks, “how much value can be unlocked if customer drift stops and outcomes are underwritten?” The first question has a ceiling — the buyer’s martech budget. The second does not, because it is measured against actual economic value created. NeoMarketing does not merely enlarge the martech business. It changes which business you think you are in — from selling software to capturing value from preventing waste and underwriting growth.
20. NeoMarketing is what comes after martech, not what improves martech. Every successful technology category eventually completes itself. Once it does, the next category is not defined by doing the same thing a little better. It is defined by changing the architecture, the economics, and the buyer relationship. Customer engagement as software is nearing that point. The leading platforms have the features. The integrations work. They all now have agents, or will soon. The remaining gains inside the category are incremental. NeoMarketing is not chasing those increments. It is changing the frame. On the attention side: cooperative networks, self-funding owned channels, the inbox as a media asset rather than a communication cost. On the outcomes side: outcome underwriting, Alpha economics, staged autonomy, operator accountability. Underneath both: shared substrate, memory, and state that let customers move without being forgotten. The distinctive claim is not “better martech” or “martech with AI.” Those phrasings collapse the category transition back into the old vocabulary. The distinctive claim is the full stack — Magnets, Mu, Markets on the attention side; Context Graphs, BrandTwins, Decisioning, Alpha pricing on the outcomes side; NeoMails, ActionAds, NeoNet as the product and network layers; ArtificialPeople and shared substrate underneath. Plenty of vendors will build pieces of this. Few will build the whole architecture. Fewer still will agree to be paid only when it works. NeoMarketing is the machine designed to make AdWaste structurally impossible. That is not a feature improvement. It is the argument for what replaces martech when the old category has finally run out of room.

**

Closing

Atrium earns attention. Meridian generates outcomes. Shared substrate carries memory across both. The Three NEVERs define the refusal: Never Lose Customers, Never Pay Twice, Never Buy Fixed. Zero AdWaste defines the destination. The Reacquisition Tax is the enemy. The Inbox Media Network is the new ad category waiting to be named. Alpha pricing is the new commercial contract that makes outcome underwriting real.

This is not a better tool. It is the first serious attempt to build an operating system in which customer drift is detected before it becomes spend, attention is earned before it is rented, and vendors are paid only when brands measurably win. Once a CMO has seen that framing clearly, the old architecture becomes impossible to defend. The numbers on the new side are simply too large, and the incentives too much better aligned.

Never Lose Customers. Never Pay Twice. Never Buy Fixed.

Zero AdWaste.

This is NeoMarketing.