Email’s Next Act: The Complete Innovation Reference

1

Overview

A reference document rather than an argument. It gathers every innovation in Email’s Next Act into one place and arranges them by what each is paid for, using the four rungs of the EARN ladder as the organising floors. There is a glossary at the end and an operating sequence in Part 8, which is the part to read if the question is where to begin rather than what exists. The consumer attention motion and the decisioning layer are both deliberately out of scope; Part 8 says why.

Email innovation has usually arrived as a succession of isolated features: a new editor, an interactive form, a deliverability upgrade, an agency service, an advertising unit. The result is a long list that is technically correct and strategically hard to hold in the head.

Email’s Next Act is different because the innovations depend on one another in a specific order. Primary inboxing makes attention possible. Composed-at-open makes the message current. Pay-in-Email and Tell-in-Email turn attention into revenue and declared data. An accountable operating layer takes responsibility for outcomes. ActionAds and a cooperative network turn trusted attention into governed media. The Factory makes all of it repeatable, and five instruments make it believable.

The organising principle is EARN. A rung is not a content category. It is a commercial position defined by capability, accountability and economics. The same Digest can be a brand-operated email, an interactive capability, an outcome programme run against a holdout, or a governed media surface. The customer job stays the same; the operating position changes. That is why the inventory below is arranged by rung — an item’s place in the list already tells you how it is sold.

Figure 1. The complete inventory on one page.

2

The Foundation

Four things sit below the ladder rather than on it. They are not commercial positions and nothing is sold on the strength of them alone. They come first because every innovation in Parts 2 to 5 fails without them, and because the most common way an email programme disappoints is that somebody bought a rung without the foundation underneath it.

Primary inboxing

NIVO addresses the probability that a message reaches the part of the inbox where a human might see it. This is the least glamorous item in the inventory and the one with the largest multiplier attached, because every downstream number — completed actions, verified lift, media yield — is a fraction of the messages that arrive somewhere visible. A programme with weak placement is not a weak programme; it is a programme measured on a denominator that was reduced before anything else happened.

Honest open measurement

Separating human activity from machine-generated opens and security scans. Privacy proxies and enterprise scanners have made the raw open rate close to useless as a signal, and a good deal of reporting has quietly become fiction as a result.

One discipline attaches, because the alternative is tempting. A human open is a health signal, never a billing unit. Pricing on verified opens would replace cost-per-send with cost-per-impression, recreate the volume incentive in a new form, and reward subject lines that earn an open without earning anything else. The measurement exists to tell the truth about the surface, not to create a new meter.

Living Emails, and composed at open

This is the enabling architecture for most of what follows, and it is a ladder of its own.

Figure 2. Five levels, and the only boundary that changes the economics.

L0 is static HTML with no interaction. L1 adds interaction that is self-contained — the recipient can do something, but nothing is written back. L2 adds writeback, so the server hears what happened. L3 composes the whole message per person, but still at the moment of sending. L4 composes it at the moment of opening.

Every level up to L3 makes a prediction. Only L4 makes a decision. That single boundary is what most of this document rests on: a payment needs a live price and an authorisation window, a booking needs current availability, an applied credit needs a real balance, and a declared preference needs somewhere it will be written back and later read. All of that is state at the moment of opening.

A static email is a prediction made at send. A Living Email is a decision made at open. The L3-to-L4 inflection is the highest de-risking priority in the whole programme.

One note for the record: L0, L3 and L4 are the fixed points. The descriptions of L1 and L2 above are the working reading and would benefit from being ratified before this becomes the version everyone cites.

Fallback-first delivery

Interactive rendering is not universally supported. It is Gmail, Yahoo Mail and Mail.ru, behind sender registration and a DMARC policy most brands have not yet set.

So the fallback build is not a courtesy path or a degradation mode. For a large share of any real list it is the primary experience, and it has to work on its own terms: a static render carrying state as of the moment of send, with a clear route into the live view. An email programme that only works in its interactive form is a demonstration rather than a product, and this is the single most common reason interactive email pilots fail to reach production.

The message grammar

One cross-cutting item, which prevents a recurring confusion. SNDR describes the job a message performs for the customer. It is orthogonal to the ladder.

Job	What it is for	Notes
Sell	Offers, promotions, conversion prompts	The one every brand already has, usually in excess.
Notify	Transactional confirmations and status	High open rates, almost never monetised, and the most under-used surface in the inventory.
Digest	Curated, recurring, low commercial intent	The format that earns a habit rather than a transaction, and where media inventory eventually becomes defensible.
Relate	Relationship content independent of any transaction	NeoMails is the Relate channel. Most brands have no Relate programme at all, which is why their only cadence is promotional.

Table 1. The four message jobs.

Two de-conflations worth stating plainly, because both appear in earlier working lists. NeoMails is not a Recover mandate. It is a Relate channel, and it can carry Protect, Recover or Grow work depending on how it is operated. A Digest is not a rung. It is a format, and it sits wherever the accountability sits.

Figure 3. Sixteen combinations. Any job can sit on any rung.

Key points 

• Four foundation items sit below the ladder: NIVO primary inboxing, honest open measurement, Living Emails composed at open, and fallback-first delivery.
• A human open is a health signal, never a billing unit.
• L0 to L3 all decide before the message is seen. Only L4 decides at open, and everything at Act, Run and Network depends on it.
• Fallback-first is not a degradation path. For most of any real list it is the primary experience.
• SNDR is the message grammar and is orthogonal to the ladder. NeoMails is a Relate channel, not a mandate; a Digest is a format, not a rung.

3

Line 1 — Email

The rung of delivery and earned attention, priced per send. Beyond the foundation, two things sit here.

The APU

The attention-processing engine, in four parts arranged as a loop.

Figure 4. Mu earns the open. Magnet earns the signal. Status earns the return. Ledger earns the invoice.

Mu recognises participation and attention. Magnet gives the recipient something to do inside the message, which converts attention into a signal. Status tells the customer what has changed and what comes next, which is what makes a return worth making. Ledger holds the memory and continuity that turn a sequence of sends into something with a history.

The important reframing is the operating question. Not how do we get this email opened, but how does this email raise the probability that the next one is opened? That is a different objective and it produces different emails.

Two placement notes. The APU’s attention-earning components belong here at Line 1; its monetisation component belongs at Line 4, because turning earned attention into inventory is a Network function — so the APU spans the ladder rather than sitting on one rung. And the return on it arrives through Carry rather than a unit price, which keeps Never Pay Fixed intact for our own infrastructure as well as the customer’s. An older definition places ActionAds inside the APU; that version and this one are both in circulation and the collision should be resolved formally.

One-Tap Subscribe

Acquiring a subscriber has always meant a landing page, a form, a confirmation and a drop-off at every stage. When the network operator is already processing the message, the prompt arrives pre-filled: one tap, explicit, logged, in-context consent from somebody demonstrably engaged. No lead form on any platform matches that consent quality, and the consent event itself is a durable asset rather than a compliance artefact.

Key points

• The APU is Mu, Magnet, Status and Ledger, arranged as a loop that ends in an invoice.
• The operating question is not this open. It is the probability of the next one.
• The APU spans the ladder: attention-earning at Line 1, monetisation at Line 4. It earns through Carry, not a unit price.
• One-Tap Subscribe produces the highest-quality consent event available in any channel.

4

Line 2 — Act

The rung where the customer completes something inside the message rather than being sent away to do it. Sold as a capability or usage fee when the brand operates it. This is the layer that constitutes Email for Revenue & Data.

Innovation	What it does
Pay-in-Email	Financial authorisation inside the inbox, which is broader than a checkout: direct payment, confirming with a stored method, a UPI authorisation, approving wallet or reward-point value, a renewal, accepting a quote, or completing an application step. The principle is one line — the transaction should end where the attention begins.
Tell-in-Email	Declared data captured at the moment of engagement: a preference, an intent, a consent, a timing choice, a renewal date, a category interest, a piece of feedback. One tap, no form, no landing page. Exact and consented, where inferred data is probabilistic and decaying.
Magnets and SmartBlocks	Interactive units inside the message — calculate, choose, declare, confirm, participate. The mechanism by which attention becomes a signal rather than an impression. SmartBlocks are the B2B expression.
Multi-step flows	Sequences completed entirely in-message: a booking against live availability, a renewal, a KYC or verification step, an application resumed from where it stopped, a preference centre that needs no login.
BrandBlocks	Reusable, brand-approved components. Less an innovation than the thing that makes the others repeatable, because a capability requiring bespoke build per campaign is a service disguised as a product.

Table 2. The Act rung. Pay-in-Email earns revenue now; Tell-in-Email earns the data that makes the next decision smarter.

Two disciplines attach. A customer who completes inside the inbox has proved they are reachable for free and should leave the paid retargeting pool the same minute — Never Pay Twice at the level of one person on one afternoon. And on the same audience and the same intervention: charge for the capability or participate in the outcome, never both.

Figure 5. The same Living Email, two commercial models, never both at once.

That second rule also disciplines the sale, because a commercial team forced to choose has to be clear about whether it is selling tooling or accountability — and the customer can tell which one they bought. Ambiguity here is where outcome pricing quietly dies: an operator paid a fee regardless has no exposure, and an operator with no exposure is an agency with a dashboard.

Key points

• Pay-in-Email is financial authorisation inside the inbox, not merely a checkout button.
• Tell-in-Email captures declared data — exact and consented, where inferred data decays and grows more expensive every year.
• Magnets, SmartBlocks, multi-step flows and BrandBlocks make the rung repeatable rather than bespoke.
• A customer who completes in the inbox should leave the paid pool the same minute.
• On the same audience and intervention: charge for the capability or share in the outcome, never both.

5

Line 3 — Run

The rung where the provider’s role changes. Somebody takes accountability for a defined customer state, operates the interventions, holds a control group, and earns only on the improvement they can prove. Act changes what an email can do; Run changes who is responsible for what it accomplishes.

Innovation	What it does
Progency	The accountable operating layer that sits after the CRM and before the auction. Not a fifth product and not a services wrapper.
Recover	Mandate for customers gone dark. The counterfactual is adtech — the money the brand would otherwise spend buying that person back through a platform that already holds their address. Connection first, then recovered attention, then conversion.
Protect	Mandate for valuable customers whose attention is cooling. The counterfactual is drift: left alone they become lost, and a lost customer is reacquired later at several times the cost of having kept them. The hardest of the three to sell, because its return is a bill that never arrives.
Grow	Mandate for the attentive. The counterfactual is a slower next purchase and margin left on the table.
Alpha pricing	Beta is what would have happened anyway. Alpha is the verified lift above it. Carry is a share of the Alpha, and only the Alpha. The economic unit is the incremental completed outcome above an agreed baseline — not an email, an open, a click or an attributed conversion.
The holdout gate	A hard gate in the automation layer: if the concurrent control is not held, the campaign does not run. An innovation rather than a policy, because a discipline that depends on somebody remembering it under quarter-end pressure is not a discipline.
Decision Traces	A record per intervention: context, eligible pool, treatment, channel, holdout status, expected outcome, cost, actual outcome, resulting state. Individually a row; accumulated, the asset that does not commoditise when the models do.
MGEs	Martech Growth Engineers — the delivery model that makes an outcome commitment staffable, and the organisational transition from Act-rung to Run-rung work.
The Alpha Audit	The front-door diagnostic. A brand discovers the size of its own leakage before anybody pitches a remedy, which is a considerably better opening than a capability demonstration.

Table 3. The Run rung. Three mandates, one per attention state, each a bet against a different counterfactual.

Two floor separations

These are the most frequent source of vocabulary drift in the whole inventory and both are worth being pedantic about.

The three mandates are not the four zones. Zones — Retain, Finish, Recover, Acquire — answer where the work sits. Mandates answer what is being changed in the customer’s state. An unfinished renewal is a Finish-zone job, and it can occur at any attention state, so it may run under Recover or under Grow depending on whether the customer is dark or engaged. Finish is an excellent offer name and the easiest thing in the catalogue to sell. It is not a mandate. Recover appears on both floors and means something different on each.

Simulated judgement and measured Alpha never share a currency. Models, priors and backtests decide what to try; they get no vote on what is paid. The holdout decides what is paid; it has no opinion on what to send. A system must never promote itself using its own predictions as evidence.

How many arms a test needs

Arm	What it establishes	When it is needed
Current best effort	Beta — the baseline, what happens anyway	Always. Without it there is no Alpha and nothing to invoice.
The intervention	The treated rate	Always.
Existing paid reacquisition	The alternative the brand is already buying, at the tax it already pays	Recover only. Protect and Grow have no adtech comparator, because the customer is still reachable.

Table 4. Protect and Grow need two arms. Recover needs three — the question is not whether owned recovery beats nothing, but whether it beats what the brand already pays a platform to do.

One constraint on this rung is not a vocabulary matter but the largest practical limit on the model: the throttle on outcome pricing is working capital, not demand. An operator paid only on verified lift funds the interventions before being paid for them, and the measurement window runs to weeks or months. Appetite is not the scarce input; a balance sheet that can carry the gap between doing the work and proving it is.

Key points

• Progency is an accountable operating layer sitting after the CRM and before the auction.
• Three mandates: Recover against adtech, Protect against drift, Grow against a slower next purchase.
• Beta, Alpha, Carry. The unit is the incremental completed outcome above an agreed baseline.
• The holdout is a hard gate in the automation layer, not a clause in a contract.
• Three mandates are not four zones. Finish is a zone and an offer name, never a mandate.
• Recover needs three test arms, because its real competitor is the paid reacquisition already being bought.
• The throttle on outcome pricing is working capital, not demand.

6

Line 4 — Network

The rung where earned attention carries adjacent demand and the direction of the money reverses. It is last for a plain reason: a dead inbox has no inventory.

Figure 6. Same components. The brand changes seats and the money arrow turns round.

ActionAds are the unit. Not a banner — a completable action inside the message, so the advertiser buys a finished action or a declared interest from an identified person rather than a probabilistic exposure. Because the identity is first-party and authenticated, the advertiser pays for certainty rather than probability while paying less, since no auction is extracting margin from every impression.

NeoNet does two distinct jobs, and collapsing them loses the more defensible one. As a media network it routes adjacent demand into earned attention — a Line 4 function. As a cooperative recovery mechanism it lets one brand reach a customer through another brand’s earned attention, paid on outcome — a Line 3 function. Same infrastructure, two commercial models, two rungs.

ZeroCPM is not a product on this rung or any other. It is the scoreboard: the point at which Lines 2, 3 and 4 have offset Line 1. It cannot be bought from a vendor by negotiating a send rate down, because the send rate is the smallest and most contested term in the equation.

What makes the inventory different

Four things separate this from sponsored newsletters. The attention was rebuilt deliberately, so it can be rebuilt again. The surface is the primary customer relationship rather than a side channel, which changes how much abuse it may absorb. The unit is an action, not an impression. And only live attention enters the pool — a dormant address sits outside the network until it proves itself with an open, so audience quality is structurally higher than any purchased list or lookalike, and improves as the network grows.

The governor

Ad load must be set by what happens next, not by what fills today. The metric is attention-adjusted yield — value taken now, net of the effect on the attention that will or will not be there in six weeks. A programme optimising raw yield will always over-monetise, because the revenue lands now and the cost lands in a future quarter. A marketplace that consumes its own supply is not a network; it is a liquidation.

The structural protection is the reason Run precedes Network. An operator paid on carry cannot over-monetise the surface, because the carry depends on the attention surviving. If opens fall, Recover misses, Protect misses, Grow misses, and the operator’s own income falls with them. The commercial model is the governor — not a promise not to spoil the surface, but an arrangement in which spoiling it is immediately expensive to whoever holds the dial.

Key points

• Network is last because a dead inbox has no inventory.
• ActionAds sell a completed action from an authenticated identity, so the advertiser pays for certainty and pays less.
• NeoNet is two functions on two rungs: media routing at Network, cooperative recovery at Run.
• ZeroCPM is the scoreboard, never the offer.
• Only live attention enters the pool, so quality is structurally higher than any purchased list and improves as the network grows.
• Ad load is governed by attention-adjusted yield, and an operator paid on carry cannot over-monetise the surface.

7

The Living Email Factory

An innovation stack does not become a business until it can be produced repeatedly, safely and cheaply. Everything in Parts 2 to 5 is a capability that has to be manufactured, or it is bespoke work with a product name attached. The Factory is the production layer beneath all four rungs.

Component	What it does
AMPGenie	A native authoring environment for interactive and Living Emails — visual and code-assisted, with the fallback build produced alongside the interactive one rather than after it.
Trove	A public, searchable repository of real emails: patterns, formats, mechanics, seasonal treatments. Worth noting separately as the strongest acquisition asset in this inventory — a free inspiration surface that lets marketers discover what is possible before entering a sales process, and milled.com has already demonstrated that the traffic exists.
Studio	Creative and content production, including the copy and asset variants that personalisation at scale requires and that no brand team can staff manually.
Component libraries	BrandBlocks, SmartBlocks, templates and workflow recipes, so Sell, Notify, Digest and Relate experiences are assembled rather than recreated. This is the difference between a capability and a campaign.
Connectors	Payment, consent, identity, catalogue, availability and writeback integrations. Unglamorous, and the actual reason projects slip.
Testing and fallback tooling	Render verification across clients, fallback validation, and the checks that stop an interactive email shipping as a broken static one.

Table 5. The Living Email Factory. Not a rung and not a venture — the production layer beneath all four.

A naming note, since the collision is easy to make: this is the Living Email Factory, not an Email Foundry. Foundry is a separate venture with an unrelated thesis, and blurring the names blurs two architectures. The Factory answers a narrower operational question: how do the experiences on every rung become repeatable products rather than bespoke projects?

Key points

• The Factory is the production layer: AMPGenie, Trove, Studio, component libraries, connectors, testing and fallback tooling.
• Trove is the strongest acquisition asset in the inventory, because it brings the audience in before any sales process starts.
• Component libraries are what separate a capability from a campaign.
• It is the Living Email Factory, never an Email Foundry. Foundry is a separate venture.

8

The Instruments

Most email dashboards report activity. This layer reports whether attention, customer economics and incrementality are improving — and it is the half most often left out of discussions of email innovation, which is odd, because it is the most persuasive material in the inventory. Every one of these five numbers is measurable today, before a brand adopts anything at all.

Instrument	What it measures	Why it is uncomfortable
CRR — Click Retention Rate	Whether engaged attention persists from one period to the next.	It falls while campaign metrics look stable, which is why decay is usually noticed two years late.
Real Reach	The ninety-day engaged base against total list size.	It reveals that list size — the number most often reported upward — is close to meaningless.
REACQ%	The share of existing customers being bought again through paid channels.	It puts a number on money spent reaching people the brand could already reach for nothing.
Adtech-to-martech ratio	Spend on renting attention against spend on owning it.	The ratio is usually somewhere between twenty and fifty to one, and nobody has been asked to defend it.
Alpha Generated	Verified uplift above a pre-agreed baseline.	It is the only number here that cannot be produced without a control group.

Table 6. The instruments.

These have two jobs. Before an intervention they diagnose, without requiring the brand to accept any pitch. After an intervention, holdouts and Decision Traces prove. No holdout, no claim.

Their quieter power is that they change who can hold the conversation. A CMO sees disappearing attention and unnecessary reacquisition. A CFO sees a measurable baseline, an avoided cost and an incremental contribution. An operator sees which customer states and which interventions are working. Three audiences, one instrument set, no translation layer between them.

The governance principle underneath all five: uplift is measured against a pre-agreed baseline with incrementality checks, and there is a shared ledger running from actions to outcomes. Without that, these are five more dashboard numbers.

Key points

• Five instruments: CRR, Real Reach, REACQ%, the adtech-to-martech ratio, Alpha Generated.
• Before an intervention they diagnose; after one, holdouts and Decision Traces prove.
• All five are measurable today, before anything in this inventory is adopted.
• One instrument set serves the CMO, the CFO and the operator without translation.

9

Putting It To Work

The inventory is not a menu. Read as a list it is thirty-odd items and the list is not the point; read as a sequence it is a loop that gets better each time round.

Figure 7. Six stages, and the sixth feeds the first.

Diagnose. The Alpha Audit and the five instruments locate the leakage, and produce a number the brand owns rather than a claim it has been offered.

Prioritise. Name one pool. Agree the counterfactual — what would otherwise happen, and what the brand is currently paying for instead. Fix the measurement design before the treatment begins, never after.

Build. The Factory assembles the treatment, fallback first.

Operate. At the appropriate rung: as a capability the brand runs, as an outcome somebody is accountable for, or as network inventory.

Prove. A concurrent holdout and incrementality checks.

Learn. Write the Decision Trace. The next diagnosis begins better informed than the last, which is the only compounding in the system that a competitor cannot buy.

The order matters more than any single stage. A programme that begins at Build has skipped the two stages that tell it what to build. A programme that skips Prove has produced activity and called it value. And a programme that skips Learn does the same work again next quarter at the same cost.

The horizon

One item is real, probably significant, and deliberately not in Parts 2 to 5. Agent-ready Email is the idea that an Agent Action Manifest travels inside an ordinary message, so a customer’s assistant can read the available actions and act on them without the human opening anything: the renewal renewed, the appointment moved, the payment authorised, by software acting on the person’s behalf.

It is sequenced behind composed-at-open for a straightforward reason. An inbox people do not open is not made valuable by making it machine-readable. Human attention first, agent attention afterwards. The manifest is a small piece of engineering on top of a Living Email; it is not a substitute for having built one.

What is not here

Two exclusions, stated so their absence is not read as oversight.

The consumer attention motion — the persistent world, Circles, standing, the always-today thread — is the supply side that makes Line 4 large, because no individual brand has enough to say to earn a daily open. It is a substantial body of work and belongs in its own document.

The decisioning layer. Lines 3 and 4 are not humanly operable at scale: thousands of customers, individual attention states, concurrent holdouts, a next-best action for each person. Every innovation on those two rungs assumes something is making the decisions, and that something is a subject in its own right rather than a footnote to this one.

Key points

• Diagnose, Prioritise, Build, Operate, Prove, Learn — and the sixth stage feeds the first.
• Fix the measurement design before the treatment begins, never after.
• A programme starting at Build has skipped the stages that tell it what to build.
• Agent-ready Email is the horizon, sequenced behind composed-at-open. A machine-readable inbox nobody opens is worth nothing.
• Two exclusions by design: the consumer attention motion, and the decisioning layer.

10

Glossary

Terms in alphabetical order, with the floor each one belongs to. Where a term is contested or awaiting ratification, that is noted.

Term	Definition
ActionAds	Governed, completable advertising units inside email. Line 4. The advertiser buys an action or a declared interest, not an impression.
Agent Action Manifest	A machine-readable description of the actions available inside a message, so a customer’s agent can act on them. Horizon item, sequenced behind composed-at-open.
Alpha	Verified lift above a pre-agreed baseline, measured against a concurrent holdout.
Alpha Audit	The front-door diagnostic. Supersedes all earlier names for it, including TripleA Audit.
AMPGenie	Native authoring environment for interactive and Living Emails, producing the fallback build alongside the interactive one.
APU	Attention Processing Unit: Mu, Magnet, Status and Ledger. Spans the ladder — earning attention at Line 1, monetising it at Line 4. An older definition places ActionAds inside it; the collision is unresolved.
Attention-adjusted yield	Value taken today, net of the effect on the attention that will or will not be there in six weeks. The governing metric for ad load.
Beta	What would have happened anyway. The baseline, agreed in advance and measured rather than asserted.
BrandBlocks	Reusable, brand-approved content components. Factory layer.
Carry	A share of the Alpha, and only the Alpha. No lift, no fee.
Composed at open	The L4 level: the message is assembled from current server-side state at the moment it is opened rather than the moment it is sent.
CRR	Click Retention Rate. Whether engaged attention persists from one period to the next.
Decision Trace	A record linking context, treatment, holdout status, expected outcome, actual outcome and resulting state. The compounding asset.
Digest	A curated, recurring message format. A SNDR job, not a rung — it can sit at any of the four.
EARN	Email, Act, Run, Network. Four rungs of rising accountability, and four lines on the brand’s P&L.
Fallback-first	Designing the non-interactive render as the primary experience rather than as a degradation path.
Finish	A zone, and an offer name. Never a mandate. An incomplete job can occur at any attention state.
Grow	Run mandate for attentive customers. Counterfactual: a slower next purchase and margin left on the table.
Holdout	A concurrent, randomly assigned control group receiving the current best effort. Enforced in the automation layer, not the contract.
Ledger	The APU component providing memory, continuity and proof. Earns the invoice.
Living Email	An email composed at open from current server-side state. L4 on the Living Emails ladder.
Living Email Factory	The production layer beneath all four rungs. Never called Email Foundry, which is a separate venture.
Magnet	An interactive unit inside a message that converts attention into a signal. Earns the signal.
MGE	Martech Growth Engineer. The delivery model that makes an outcome commitment staffable.
Mu	The attention currency. Recognises participation; earns the open. Not purchasable, poolable or transferable.
NeoMails	Daily attention-earning Relate emails. A channel, not a mandate — it can carry Recover, Protect or Grow work.
NeoNet	Cooperative infrastructure with two commercial expressions: media routing at Line 4, and outcome-based recovery at Line 3.
Net Email Cost	Delivery and content cost, less capability, outcome and media revenue. The brand’s scoreboard.
NIVO	Primary inboxing and deliverability. Foundation layer.
One-Tap Subscribe	Pre-filled, in-context consent captured inside a message. Line 1.
Pay-in-Email	Financial action or authorisation completed inside the message. Line 2.
Progency	The accountable operating layer, sitting after the CRM and before the auction. Line 3.
Protect	Run mandate for cooling customers. Counterfactual: drift into a lost state and later reacquisition at several times the cost.
Real Reach	The ninety-day engaged base against total list size.
REACQ%	The share of existing customers being re-bought through paid channels.
Recover	Run mandate for customers gone dark. Counterfactual: paid reacquisition through a platform that already holds the address. Also a zone name — the two meanings differ.
SmartBlocks	Interactive engagement units in a B2B context. Line 2.
SNDR	Sell, Notify, Digest, Relate. The message grammar, orthogonal to EARN. Supersedes the three-part SNR.
Status	The APU component that shows what changed and what comes next. Earns the return.
Tell-in-Email	Declared data, consent, preference or intent captured inside the message. Line 2.
Trove	A public repository of email patterns and designs. Factory layer, and the strongest acquisition asset in the inventory.
ZeroCPM	The point at which Lines 2, 3 and 4 offset Line 1. A scoreboard, never an offer.
Zones	Retain, Finish, Recover, Acquire. Where the work sits, as distinct from the three mandates, which describe what is being changed.

Table 7. Reference glossary.

**

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

Read as a list, this is thirty-odd innovations and the list is not the point. Read as a ladder, it is four commercial positions built on one foundation, produced by one factory, and checked by five instruments — and almost none of it is technology that will remain scarce. What will remain scarce is the willingness to measure the surface without flattery, to hold a control group you could have sold to, and to let the media line grow only when the attention has proved it can take it. The list of innovations will keep changing. The architecture should not.