Rajesh Jain

Every NeoMarketing play is a move on one grid — the Transaction–Attention Table (TAT). Before you can run a play you have to build the map. A series on how, from five fields to a working radar.

The overview.  The Profit You Already Own introduced the Transaction–Attention Table as the map. This series is the build guide a CMO needs to commission that map and trust it — what data it takes, how to set each axis, and where to stop.

1  

The map comes before the machine.

Every transformation needs a map, and for NeoMarketing the map is the Transaction–Attention Table. The idea is simple. Rows show how far a customer has progressed in transactions. Columns show whether the customer is still paying attention. Put the two together and a brand can see what its dashboard normally hides: which customers can be grown, which must be protected, and which have already fallen into recovery.

Most CRM teams start the other way around — with actions. Send this journey, launch this offer, run this campaign, suppress this audience, target this cohort. That is activity-first marketing: it produces motion, not necessarily profit. The TAT changes the sequence. First map. Then diagnose. Then choose one play. Then measure the outcome.

Four questions before any play

Before approving any CRM or recovery programme, a CMO should be able to answer four questions:

1. Where are my customers on the TAT?
2. Which cells hold the most revenue at risk?
3. Which cells are moving in the wrong direction?
4. Which single play should run first?

If those four cannot be answered, the brand is not ready to spend — it is ready to guess.

Why RFM is not enough

The reflex answer to where are my customers is RFM — recency, frequency, monetary value. RFM is useful and incomplete, because all three of its variables are transaction variables: recency means transaction recency, frequency means transaction frequency, monetary means transaction value. RFM cannot see a customer who has stopped opening, stopped clicking, stopped visiting — until the transactions stop too. A customer who bought three times and still opens weekly is not the customer who bought three times and has ignored you for 90 days. Same RFM score, opposite trajectory. Transaction history tells you what happened; attention tells you what is about to.

The shared view

The deeper value of the TAT is organisational. Without it, every team argues from its own dashboard — CRM from opens and clicks, performance from ROAS, product from conversion, finance from cost — and none sees the whole base by transaction depth and attention health. The TAT creates one shared view. It tells the CRM team where to grow, the retention team where to protect, Progency where to recover, and the CFO whether paid reacquisition is a genuine necessity or a failure of earlier attention. The map is not the answer. It is the place where the right answer can finally be seen — and agreed.

2

The minimal data to start.

The power of the TAT is that it does not require a perfect data lake. A brand can build the Basic version from a handful of fields, one row per identified customer: a customer ID, preferably anonymised for analysis; lifetime transaction count; last meaningful attention date; and revenue or margin per customer.

The first field identifies the customer. The second places them by transaction depth — the row. The third places them by attention recency — the column. The fourth does something the first three cannot: it turns the grid from a headcount census into a money map.

Why revenue is the field that matters

That fourth field is the one teams are tempted to skip, and it is the one that makes the table a decision tool. A TAT built only on counts can tell you that a million customers are lost. Interesting, but not actionable. A TAT enriched with revenue or margin tells you which lost customers are worth recovering, which one-time buyers are worth pushing to a second purchase, and which repeat customers must be protected before they become expensive to re-buy. Counts describe the base; money ranks the work. Without revenue in the cells, every prioritisation argument collapses back into opinion.

What you do not need yet

A CMO does not need the full engagement event stream on day one — that is for the Advanced version. Basic can start from the latest trusted attention event already sitting in the engagement systems. The instruction to the analyst is therefore short: for every identified customer, find the lifetime transaction count, the last meaningful attention event, and the customer’s value; then place them on the grid. That is enough to act. The discipline here is to resist the urge to wait for completeness. The brand that builds a rough map this month will out-decide the brand still scoping a perfect one next quarter. Completeness is the enemy of the first decision; the map only has to be good enough to choose one play and measure it.

3

What counts as attention.

The most important construction decision in the whole table is not the row logic. It is the column logic — specifically, what counts as attention. Get this wrong and the map lies; and a map finance does not trust is a map no one acts on.

The column must be set by the last meaningful attention event — not the last order, not the last send, not the last impression, and emphatically not the last open if opens are polluted. The column should reflect one thing: whether the customer has recently shown they are still listening.

The hierarchy of trust

Order the signals by how much intent they reveal. A transaction tops everything. Then a website visit or product browse, then an app open or session, then a WhatsApp reply or reliable read, then a push tap, then an email click. The exact order can vary by category and channel maturity; the principle does not. Self-initiated signals — a visit, an app open — outrank solicited ones such as a click on something you sent. Passive or machine-fired signals are treated with suspicion.

The email open problem

This matters most for email. For years, opens were treated as proof of attention. That is no longer safe. Privacy protections, inbox prefetching and machine activity now fire a large share of opens with no human behind them. A customer who opens nothing but clicks once is more valuable than one who appears to open everything and never acts. The click is a customer signal; the open may be a machine shadow. So the rule is blunt: score the click, not the open, and if you keep the open at all, floor it near zero and tag it the lowest grade. This is not an argument that email is weak — it is an argument that measurement must be honest. Email clicks, AMP interactions, form fills, quiz answers, preference updates, pay-in-email actions and product taps are powerful, identified intent signals. Opens are low-grade supporting evidence, not the foundation of the table. The reason to be strict is downstream: every machine-fired open you let into the Strong column is a customer you will fail to recover because you believed they were fine.

4

Drawing the two axes.

With what counts as attention settled, the axes themselves are quick to draw.

The columns: 30 and 90 days

Set three columns by the date of the last meaningful attention event.

• Strong: within 30 days
• Weakening: between 30 and 90 days
• Lost: more than 90 days of silence

The 30-day cut identifies current listening; the 90-day cut identifies relationship breakage. A common objection arrives here: in long-cycle categories — insurance, furniture, jewellery, auto, real estate — customers simply do not buy often, so surely long silences are normal? This confuses the two clocks. A customer may not buy a sofa every month, but they can still read, click, browse, save and reply in between. Long purchase cycles do not justify long attention silences. This is the single most important distinction in the table: transaction recency and attention recency are not the same thing. A customer may not be due to buy, but they should never be invisible. If 90 days pass with no meaningful attention, the relationship is damaged even if the next purchase is months away — and recovery then needs a different machine than retention.

The rows: transaction depth

Rows are simpler: lifetime transaction depth, in three bands.

• None — identified customers who have not yet made a first purchase: subscribers, registered users, abandoned browsers, captured marketplace buyers; known, but not yet customers in the economic sense.
• One — customers who have bought exactly once; for many brands this is the largest and most important row, because a majority of first buyers never return.
• Repeat — customers who have bought twice or more, the Best relationships a brand must never let drift into paid reacquisition.

These three rows are the depth half of the BRTN segments — Best, Rest, Test, Next. Repeat maps to Best, One to Test, None to Next. Rest is the one that is not a depth at all but a state: the slipping-away pool that cuts across every row, splitting into R1, R2 and R3 — lapsed Best, Test and Next respectively. So the R is for Rest, not Recover; the columns we set next are how a customer falls into it.

Why the boundary is two, not three

The boundary that matters is between one purchase and two — and it must be visible, not buried inside generic RFM tiers. The move from first to second purchase is not just another conversion; it changes the economics of the customer. It often roughly triples lifetime value and makes every future purchase more likely. Setting the Repeat band at two-or-more puts that first-to-second inflection directly on the grid, where the whole game is played, instead of hiding it inside a broad repeat bucket. The table then shows, at a glance, how many customers are stuck after one purchase, how many remain attentive enough to move to a second, and how many have already gone quiet. That single view tends to reorder a CMO’s priorities on the spot.

5

Reading the nine cells.

Once rows and columns are set, every identified customer lands in one of nine cells — and each cell is a job, not a label.

Three verbs

The columns create three management verbs.

• Grow the Strong
• Protect the Weakening
• Recover the Lost

This is the first language a CMO should insist on: simpler than a hundred micro-segments and more useful than a generic lifecycle chart. The rows tell you how valuable or fragile the relationship is; the columns tell you whether it is still alive.

The nine jobs

Read across the grid and the plays name themselves.

• Strong / None is the First play: convert known prospects into first-time buyers.
• Strong / One is the Second play: move trial buyers to the crucial second purchase.
• Strong / Repeat is the Repeat play: keep the Best buying.
• Weakening / None needs nurture before the first-purchase window closes.
• Weakening / One needs protection before the second-purchase window closes.
• Weakening / Repeat is the quiet danger zone — valuable customers drifting before anyone in finance can see it.
• Lost / None usually warrants a low-cost revival or suppression.
• Lost / One is classic win-back: people who tried once and disappeared.

The most painful cell

And Lost / Repeat — coded R1 — is the most painful cell on the map: customers with proven value who have stopped listening and who, left alone, will be re-bought through adtech months later at full tax. This is the cell that most directly converts a brand’s own database into someone else’s margin pool, and it is the cell the dashboard is least equipped to surface, because the revenue has not yet visibly stopped. R1 is where recovery — Progency — earns its place, and where the holdout will later prove whether the lift was real.

Reading the nine cells is, in the end, reading where profit is leaking and which verb stops it.

6

Basic is enough to act.

There is a temptation to wait for the perfect table: complete event history, cross-channel identity, margin-level product data, predictive scores, attribution models, incrementality engines. All useful; none necessary to begin.

One question

Basic TAT treats attention as a date and asks one question of every customer: when was the last trusted signal that this person was still listening? That single question, answered across the base, is enough to run the first set of plays.

Reading Basic into action

The reads are direct. If One / Strong is large, run the Second play. If Repeat / Weakening holds high value, run Protect. If Lost / Repeat contains meaningful revenue, hand it to Progency. If None / Strong is growing, optimise the First play. And if Weakening is expanding across every row, your attention system itself is decaying — a warning no revenue report would have given you in time.

Simplicity is the point

Basic is the map every brand can build today, and it is deliberately simple, because simplicity creates agreement. The CFO can understand it. The CMO can commission it. The CRM team can act on it. The performance team can be challenged with it. Its purpose is not analytical perfection; it is managerial alignment. If the analyst cannot explain it, it is too complex; if the CRM team cannot act on it, the map is decorative. Basic is not a lesser version to be upgraded away from — it is the floor, and for most brands the floor is enough to start moving customers up and leftward while stopping the drift to the right. A simple map that ships this quarter beats a sophisticated one still being scoped next year — the cost of waiting for perfect data is every customer who drifts to Lost while the model is built.

7

Advanced: when attention becomes a score.

Once the Basic table is trusted, the Advanced version can be built — and the order matters: trust first, sophistication second.

Attention as a score

Advanced TAT does not replace the grid; it enriches it. The rows are still transaction depth, the columns are still set by attention recency, but inside each cell every customer now carries an attention score. The score combines three things you can defend out loud: signal type (self-initiated beats solicited), recency (points decay, so the score falls on its own if nothing fresh arrives), and depth (a transaction beats a browse, a browse beats a click, a click beats an open). A practical method is a capped blend — take the top meaningful signal, add a small contribution from the next two, cap at one hundred — rather than letting forty hidden weights decide the one screen a CFO is looking at. An auditable five-input score beats a black-box forty-input one every time.

Recency sets the column; the score enriches the cell

One rule protects the whole thing from false precision: recency sets the column; the score only enriches the cell. Never let the score redraw the grid. If the score decides the column, the table becomes a black box and teams debate weights instead of acting. A customer with a strong signal 95 days ago still belongs in Lost by the column rule — the score may tell you to prioritise them within that lost cell, but it does not move them out of it. A customer with a weak signal yesterday belongs in Strong, even if the score flags them as low priority there. The column answers one question — are they still listening? The score answers a different one — how strong is the attention inside that state? Keep the two separate and the map stays legible; merge them and it turns opaque.

What the score unlocks

Held to that discipline, Advanced unlocks three things Basic cannot.

• First, within-cell ranking: not every Lost / Repeat customer deserves equal recovery spend, and the score decides who gets email, who gets WhatsApp, who gets a concierge call, who gets NeoNet routing and who gets suppressed.
• Second, trajectory: a score falling week after week matters more than one that is merely low but stable, and a falling score in the Strong column is the earliest possible drift warning — the chance to protect a customer before they cross the 90-day line.
• Third, a real Click Retention Rate: attention-score decay across the base is attention churn, measured rather than asserted, and because attention decays before revenue does, it is a far earlier warning than revenue churn.

Basic tells you where customers are; Advanced tells you which way they are moving. Basic is the map; Advanced is the radar.

8

Commissioning the map, and the first move.

A CMO does not need to specify every formula, but the commissioning brief should be clear enough to prevent the common mistakes. It comes in two stages.

The brief

Build Basic first. Use customer ID, lifetime transaction count, last meaningful attention date and customer value. Set the attention date from click-grade or self-initiated signals; floor email opens. Use 30 and 90 days as the default column cuts and lifetime count for transaction depth. Show both customer count and revenue or margin in every cell. Name the two cells with the highest money at risk. Recommend one play, not six. Only once Basic is accepted, build Advanced: add signal scoring, recency decay, a capped blend and trajectory — and use the score to rank customers inside a cell, never to move them between columns. Keep the weights visible and every output auditable.

The one screen

The output should fit on a single screen: the nine-cell table, customer count by cell, revenue or margin by cell, the attention trend by cell, the highest-risk cell, the recommended first play, and the holdout design that will measure it. The tests are practical. If the analyst cannot explain the table in five minutes, it is too complex. If the CFO cannot see why a cell is being prioritised, the model is too opaque. If the CRM team cannot act on it, the map is decorative. The TAT exists to decide, not to decorate.

The first move

The first move after building the map should not be a grand transformation. It should be one measurable play against one leaking cell. For many brands that will be Second — move attentive One-band customers to a second purchase, often the fastest route to LTV expansion. For others it will be Protect — stop valuable Repeat customers drifting from Weakening into Lost. For brands with large lapsed pools and heavy paid reacquisition it will be Recover — hand Lost / Repeat to Progency and measure the lift against a three-arm holdout of worked, untouched control and adtech arms. The choice should be driven by the table, not by organisational habit. A campaign calendar asks what we are sending next week; the TAT asks where profit is leaking and which play fixes it first. That is the difference between activity and accountability.

The discipline

The table fails two ways: over-engineered too early, or reduced to a pretty segmentation chart with no economic consequence. The discipline that avoids both is short. Transaction depth sets the rows. Attention recency sets the columns. Customer value sizes the opportunity. The attention score ranks priority inside a cell. Holdouts prove whether the play worked. Everything else is optional. This is why the TAT belongs at the very front of the Alpha Audit: before a brand buys new software, switches agency, raises spend or celebrates another ROAS report, it should know how many of its customers are still listening, how many are drifting, how many have gone dark, how much revenue sits in each state, and how many will be re-bought through adtech if nothing changes. Those are not analytics questions. They are profit questions — and they are the map your dashboard was built not to draw.

Published June 25, 2026

Marketing’s most expensive problem is the one your dashboard cannot show you: how much you pay to re-buy the customers you already have. This is the whole argument — the map, the maths, and the one move to make on Monday.

1

The most expensive line in marketing is the one nobody reports.

Your marketing dashboard reports opens, clicks, revenue and return on ad spend. Every one of those is a measure of activity — of how busy the engine was. None of them reports the one thing quietly thinning your margin: how much of your growth you are paying for twice. A performance dashboard is built to show motion, not waste, so the most expensive problem a brand has is the one its own reporting is structurally unable to surface.

It helps to restate the job. A marketing team has one real task — to produce the next profitable transaction — and there are only two ways to add profit to it. Win that transaction in less time, by moving customers up the ladder on a channel you own. Or pay less tax to make it happen, by not renting back customers you already have. Profit is margin minus marketing cost; push the margin up and push the cost down, and everything else is detail.

The largest avoidable cost on most brands’ marketing line is the same: paying adtech to re-buy customers already sitting in the database. That is AdWaste — and it is invisible precisely because the dashboard counts the re-purchase as a win, not as a customer you owned and lost and bought back.

Push the margin up, push the cost down. Everything else is detail.

2  

Attention is the lead indicator your dashboard cannot see.

Every transaction is downstream of attention. Nobody buys from a brand they have stopped noticing. That makes attention the lead indicator and revenue the lagging one — and the gap between them is where brands lose customers without realising it. By the time a sales report dips, the attention left months earlier; the dashboard simply had no column for it.

Track attention and you can act before the loss rather than after it — re-engage a fading customer while there is still a relationship to save, instead of paying to reacquire them once they are gone. This is possible for any identified customer: anyone reachable by email or mobile, which for most brands is the majority of their value.

Revenue tells you what already happened. Attention tells you what is about to.

3

A better map than RFM: the Transaction–Attention Table.

RFM ranks customers by what they have already done. The Transaction–Attention Table adds the axis that predicts what they will do next: attention. The rows are transactions since the customer’s first order; the columns are attention right now, split at thirty and ninety days. Ninety days of silence — even in a long-cadence category like insurance or furniture — means the relationship is broken, whatever the purchase history says.

The columns hand you three jobs. Strong attention is to be grown. Weakening attention is to be protected before it slips. Lost attention has to be recovered. Most of any brand’s base sits in the weaker rows, and the whole game is moving them up while stopping the drift rightward — the drift RFM is blind to.

The rows hide the single most important number in consumer marketing: the second purchase. For most brands, 60-65% of customers buy once and never return. The second transaction roughly triples lifetime value, and each purchase after it makes the next more likely. One purchase is a trial; the second is a customer. Almost everything worth doing is in service of getting from one to two — and then never letting two go dark.

One purchase is a trial. The second is a customer — and it roughly triples their value.

4

Five plays your CRM team already runs — and a sixth it was never built for.

Read as moves on the table, the work resolves into a small set of plays. Capture turns anonymous and intermediated buyers into known customers — the on-ramp onto the grid. First moves a customer from nought to one. Second moves them from one to two, across that tripling inflection. Plus-One keeps the best customers buying. Attention pulls drifters back before they are lost. Those five are the Grow and Protect work your CRM team already does — and agents plus in-channel interactivity now make each of them sharper and cheaper to run.

The sixth play is different in kind. Recover takes the lost column — the customers whose attention has gone dark. Nobody owns it, and that is not an accident. A CRM team is built and measured to keep the engaged engaged; the lost are always next quarter’s problem, so they are never worked. They fall through the floor — straight to adtech.

Five plays climb the rows and hold the columns. The sixth wins back the customers everyone else has given up on.

5

Why recovery is a different machine.

Here is where most teams go wrong: they treat recovery as a harder win-back campaign. But by the time a customer is in the lost column, the channel is still open and the customer has simply stopped listening. Sending a sharper offer down a channel no one reads only trains them to ignore you faster. Recovery is not a better campaign; it is a different machine.

That machine runs a different sequence: attention, then repetition, then conversion. It re-earns the open, builds a habit of opening, and only then asks for the sale. Two kinds of message do the first two jobs — and they are exactly the messages a conventional CRM programme never sends. Relate, perhaps twice a week, carries no offer at all; it is a reason to open, not a discount, and its only job is to rebuild reachability. Digest, weekly, is a recurring read that turns a re-opened inbox into a habit. Sell comes last, when the customer is paying attention again, and it closes in the channel — pay-in-email or pay-in-WhatsApp — with no detour to a website where the intent leaks away.

The contrast is the whole point. The CRM team starts at conversion, because that is what it was built to do. Recovery starts at attention, and earns the right to sell. That is also the honest answer to the question every operator asks next — if recovery were this valuable, why is my team not already doing it? Because it was never built for this job, and was never measured on it.

The CRM stack was built to send. Recovery has to earn the open before it asks for the order.

6

The AdWaste you are funding — and the tax that comes with it.

Today the lost column is handed to adtech by default. For many brands, around 70% of repeat transactions come back through that rented channel — at a 30-35% tax, for a return on ad spend of roughly 3. You are paying a platform to re-buy a customer already in your own database. That is the AdWaste: paying twice for what you already own, and calling the second payment ‘performance’.

Every route to a transaction carries a tax, and the further the route from your own channels, the bigger the bite:

Route	Tax (cost as % of revenue)	Note
Organic · direct	~0–5%	the cheapest revenue you have
Owned CRM	~5–10%	the everyday Grow and Protect work
Recovery (pre-adtech)	~15%	the missing rung — half the adtech tax
Adtech	~30–35%	≈70% of recovery today, at ROAS ~3
Intermediated · marketplace	~35–40%	you may not even own the identity

The job is to move transactions up toward the cheapest proven route, and to spend adtech last — for what nothing cheaper could reach. The sixth play deserves an owner that does exactly this. Call it a Team 6: a pod, internal or outsourced, with one number to hit — a return on ad spend of 6, hence the name — working the lost column on owned channels, before the auction, at half the tax. If you would rather not build it, the done-for-you version is Progency, which the next two essays go inside.

Every route to a sale carries a tax. Take the cheapest proven one first — and adtech last.

7

The maths — and the honest version.

The appeal is simple arithmetic. Recover a dollar of revenue through adtech at a return of 3 and the ad cost is about a third of it; recover the same dollar on owned channels at a return of 6 and the cost is about a sixth. Move the recovery from one to the other and you save roughly a sixth of that revenue — about 17% — and because it is a cost you simply stop paying, it falls straight to profit. For a business on a 10% operating margin that recovers a meaningful share through paid media today, shifting that share to owned recovery and tightening the five owned plays can roughly double operating profit. On $100m of revenue, around $10m of profit becomes around $20m.

Now the part a keynote has no time for, and which a written argument owes you. That 17% saving is the same at any gross margin. It is cost arbitrage — you are re-routing the same revenue more cheaply — and the gross margin cancels out of the comparison entirely. But gross margin governs a different question: whether recovering at that return is worth doing at all. A recovery only profits when its return on ad spend clears one divided by the gross margin — about 1.4 at a 70% margin, but 6.7 at a 15% margin.

So a return of 6 is comfortably profitable for higher-margin, replenishment-led categories — beauty, supplements, grocery, pet — where the timing is predictable and the margin is generous. For thin-margin categories it barely clears break-even, and there recovery still wins, but on the near-zero marginal cost of an owned email rather than on the headline return. The arithmetic of the saving is universal; the wisdom of the activity depends on your margin. Both belong in the same sentence.

Half the tax is double the return — by arithmetic. Gross margin doesn’t shrink the saving; it decides whether the recovery was worth running.

8

Why you can believe the number.

A claim this large needs a proof a brand can run on its own data, not a benchmark from someone else’s. That proof is the holdout. Split the lapsed base at random into three arms: one is worked, one is left completely untouched, and one is given to adtech. The untouched arm is a baseline the brand audits for itself — proof that some of those customers would have returned with no help at all. Recovery is then billed not on the customers it brought back, but on the lift above that control, reported as a range rather than a falsely precise figure. No lift, no bill.

That single mechanism is what separates this from a vendor’s spreadsheet. The number is measured against the brand’s own control, which the vendor cannot move, rather than asserted from a deck. The test of whether a partner believes its own numbers is whether it will hold back a control group and take its fee only on the lift.

And the honesty has to travel with the limits. Recovery needs identifiable customers, repeat or renewal economics, real owned channels, and enough lost-customer volume to run a control. It is not the right first move for a pure land-grab brand still proving product-market fit, for a genuinely no-repeat category, or for a marketplace seller with no path to direct identity. A brand that cannot hold back a control is not ready — and that, itself, is the first useful finding.

The test of whether a partner believes its own numbers is whether it will hold back a control and take its fee only on the lift.

9

What to do Monday — and where it leads.

None of this starts with a migration. It starts with a diagnosis. The front door is the Alpha Audit: bring your own data and get back four numbers — how much repeat revenue you re-buy through paid media, how many proven buyers are quietly fading, how much value sits in customers who have gone dark, and how many of your buyers you ever convert into known customers — plus a cohort map and one recommended first move. Not six moves. One.

That first move is almost always the same, and its power is that it is mechanical. Suppress your active customers from retargeting, redirect the saved spend to owned channels, and measure the lift against a holdout. You do not have to believe a doctrine to run it — only remove an audience, redirect a budget, and compare outcomes. The performance team may argue the retargeting was incremental; the holdout settles the argument with your own data, and it tends to pay for the audit by the end of the quarter.

From there the path is a ladder, each rung earned rather than assumed: the audit first; then CRM 2.0 for the five plays that Grow and Protect; then a Team 6 — built or bought as Progency — for the lost column, once the recovery economics justify it; then scale only what the holdout has proven. That is the whole of NeoMarketing in one line: stop doing marketing for its own sake, and start making profit from the customers you already own.

You are almost certainly paying twice for customers you already own. The map shows where, the maths shows how much, the holdout proves it on your own data, and the first move is small enough to make on Monday. Stop paying twice. — Never Lose Customers. Never Pay Twice. Never Pay Fixed.

   The argument in brief

Question	Answer
The problem	You pay adtech to re-buy customers already in your database — AdWaste your dashboard counts as a win.
The map	The Transaction–Attention Table: rows are transactions, columns are attention. Strong → Grow, Weakening → Protect, Lost → Recover.
The inflection	60–65% buy once and never return; the second purchase roughly triples lifetime value. The game is getting to two.
The sixth play	Recover the lost column — a different machine that earns attention before it asks. Build a Team 6, or buy it as Progency.
The maths	Half the tax is double the return (~17% of moved revenue, saved as cost). Gross margin sets the break-even, not the saving.
The proof	A three-arm holdout the brand audits itself. Paid on lift above its own control. No lift, no bill.
Monday	Run the Alpha Audit; suppress active buyers from retargeting; redirect to owned; measure against a holdout.

 

Published June 24, 2026

Find the disease your dashboard cannot see — four numbers, on your own data, in about three weeks.

Core idea.  The Alpha Audit is not the proof that NeoMarketing works. It is the front door that lets a brand test the doctrine on its own data — before buying anything else.

1

The diagnosis your dashboard cannot run

Your marketing dashboard reports opens, clicks, revenue and ROAS. Every one of those is a measure of activity — of how busy the engine was — and none of them tells you the one thing quietly thinning your margin: how much of your growth you are paying for twice. A performance dashboard is built to show motion, not waste, so the most expensive problem a brand has is the one its own reporting is structurally unable to surface.

The Alpha Audit is the instrument that surfaces it. It is not a platform migration, not a Progency contract, and not a demand that the CMO reorganise the team before seeing value. It is a fixed diagnostic: you bring your data, it hands back four numbers, a cohort map and one recommended first move — in about three weeks in its fullest form, and from a single file in its lightest.

The right analogy is bloodwork, not surgery. A blood test does not cure the disease; it tells the doctor what is actually happening before anyone prescribes treatment. The Alpha Audit does the same for customer economics — it reads where a brand’s revenue is leaking time and tax, the two levers the wider doctrine is built on. Nothing in your stack changes to run it.

The audit is a front door, not a trapdoor — a small, safe, fast first step.

Where the dashboard answers ‘how much did we sell and how busy were we,’ the audit answers a sharper question, and it is the one a board actually wants answered: how much of last quarter’s spend was structurally avoidable? Not all paid spend is waste, and the point is not a moral argument against adtech. The point is to separate necessary spend from avoidable spend — customer by customer, route by route. The first ask is intentionally small: upload, see, then decide.

2

The four numbers it returns

The audit does not begin with a maze of dashboards. It begins with four numbers, and each number is a diagnosis that points to exactly one Alpha Move.

Paid Repeat Leakage is the share of repeat revenue bought back through paid media — customers you already owned, re-rented from a platform; it triggers Grow. The Weakening Pool counts proven buyers whose attention is fading before any sales report would catch it; it triggers Protect. Rest Recoverable Value is the revenue sitting in former Best and Test customers who have gone dark; it triggers Recover. Identity Capture Rate is the share of intermediated buyers and anonymous visitors you ever convert into known customers; it triggers Capture.

The Big Four diagnostics — and the Alpha Move each one triggers.

The four are not vanity metrics; they are budget-routing metrics. Apart, they are interesting. Together, surfaced on a single page, they turn a vague sense that ‘marketing could be better’ into a board-ready slide with a number on it — and a number is something a CFO can act on. The Playbook explains what each move does once you have the number; the audit’s only job is to put the number in front of you, measured on your own data rather than asserted from someone else’s benchmark.

3

Where the data lives — and why the audit is tiered

The reason most brands have never seen these four numbers is not that the maths is difficult. It is that the numbers live in different systems that were never joined. The order file is the spine of the whole audit — customer ID, order date, value, channel — and almost every brand already has it. From the order file plus a last-engagement signal, the audit can compute Rest Recoverable Value and the Weakening Pool straight away.

The other two are harder, and it is worth being honest about why. The attention signals — last open, click, read, app session — live in the engagement platform. Identity Capture needs the marketplace feed and the anonymous web traffic. And Paid Repeat Leakage, the most important number, needs the hardest join of all: orders against paid-media spend and audiences, connected by an attribution rule. That is the single biggest data obstacle in the audit, and the reason the fourth number is the last to arrive.

No single system holds the whole truth — the order file is the spine; everything else hangs off it.

A word on the obvious source. GA4 is a useful input for the attention layer — visits, sessions, engagement — but it is event-centric, not customer-centric, increasingly aggregated and consent-gated, and it holds none of your order economics or discount data. GA4 can be an input to the audit; it can never be its spine.

Because the data lives in those tiers of difficulty, the audit is delivered in three tiers of effort — and only the first is truly self-serve, by design. Tier 1 is instant: you upload one order export and, in minutes, see the reachability half of the picture with a readiness score. Tier 2 is connected self-serve, attempting the full join through read-only connectors. Tier 3 is the assisted Full Alpha Audit, where a Martech Growth Engineer reconciles the messy reality of siloed teams and finance data, and where Paid Repeat Leakage gets computed properly.

Tiering turns the audit from a consulting engagement into a product.

Tier	Minimum data	What it returns	What it cannot yet prove
Tier 1 · Instant self-serve	Order CSV: customer ID, order date, value, channel; optional last-engagement date	Reachability map, Real Reach, Weakening Pool, Rest Recoverable Value, readiness score	Paid Repeat Leakage; true offer-tax economics
Tier 2 · Connected self-serve	Read-only connectors to commerce, engagement, GA4/BigQuery and the ad platforms	Attempts all four numbers, with confidence labels and missing-data flags	Inconsistent IDs, agency files, finance data may still defeat the join
Tier 3 · Assisted Full Audit	MGE-assisted reconciliation of orders, paid audiences, finance, offers and marketplace feeds	Full Alpha Audit with reliable Paid Repeat Leakage and effective tax	Nothing material; remaining gaps reported as confidence bands

 

The tiering is not a compromise; it is the product. And if a brand cannot convene the data owners at all, that is not a reason the audit fails. The inability to assemble the data is the audit’s first finding — a brand that cannot see its own customer picture is exactly the one leaking most through the gaps between its teams.

4

What the brand gets back, and the one move it makes first

The audit hands back a readout, not a dashboard maze: the four numbers, a cohort map, the readiness score, a ranked list of avoidable-spend pools, and one recommended first move. Not six moves. One. A diagnostic that ends with too many recommendations becomes consulting theatre; one that ends with a single concrete test earns the right to the next conversation.

In most brands that first move is the same, and it is the move the audit almost always recommends. Suppress your active customers from retargeting, redirect the saved spend to owned channels, and measure the lift against a holdout. It works because it is mechanical — the brand does not have to believe a doctrine, only remove an audience, redirect a budget and compare outcomes. The performance team may argue retargeting is incremental; the holdout settles the argument with the brand’s own data, and it tends to pay for the audit itself by the end of the quarter.

The readout should make that first move feel safe: which customers are eligible for suppression, how large the holdout should be, what window to measure, what revenue is at risk — and what not to do. Do not cut broad prospecting first, do not remove all retargeting overnight, do not declare victory on a 30-day read if the category has a longer cadence, and do not count gross recoveries as Alpha.

5

Where the front door leads — and who should not walk through it

The audit is the entrance to a ladder the rest of the doctrine has already described: the audit first, then CRM 2.0 for Grow and Protect, then Progency for the lost column when recovery economics justify it, then scale after proof. But the entrance is built to be useful on its own — the door opens inward, and the brand is never pushed through it.

The front door leads to the NeoMarketing ladder — but each step must be earned.

And, as with everything else in the doctrine, it is honest about who it is not for. The Alpha Audit needs identifiable customers, repeat or renewal economics, usable owned channels, and enough volume to measure a lift. It is not the right first move for a pure land-grab brand whose only job this year is to acquire; not the right doctrine for a no-repeat category; not a magic wand for a marketplace seller with no path to direct identity. A brand that cannot hold back a control group will get a thin audit and an honest readiness score saying so — rather than a confident number it has not earned.

Most diagnostics are sold as the opening move of a long engagement you are expected to continue. This one is designed to be genuinely useful even if you stop after the readout. The promise is deliberately modest and therefore powerful: upload one file, see what can be known, learn what cannot yet be known, and run one test that tells the truth.

—

The essay in brief

Question	Answer
What it is	A fixed diagnostic that returns four numbers, a cohort map, a readiness score and one first move.
What it is not	Not a platform migration, not a Progency contract, not a rip-and-replace, not a claim without the brand’s own data.
The product ask	Tier 1: upload one file and see what can be known. Tier 2: connect systems. Tier 3: assisted full truth.
The first move	Suppress active buyers from retargeting, redirect to owned channels, and measure against a holdout.
The closing promise	Four numbers. One first move. No commitment until the brand’s own data has earned it.

Landing-page CTA.  Find the disease your dashboard cannot see. Upload one file, see three of your four numbers, learn what is missing, and run one test that tells the truth.

* * *

The Alpha Audit replaces the generic claim — brands waste money reacquiring customers — with the specific number: this is how much your brand is paying to do it. Four numbers. One first move. No commitment until the brand’s own data has earned it.

Published June 23, 2026

Recovery is a two-step job — earn the attention back, then convert it — and a general-purpose CRM stack is built for neither. Inside the two engines, and how the model that runs them is built.

1

The recovery problem a CRM stack cannot solve

Recovery is not the same job as Grow or Protect, and treating it as one is why most win-back fails. By the time a customer is in the lost column, the attention is gone. The channel is still technically open — the email still delivers — but the customer has stopped listening. Sending a sharper offer down a channel no one reads does not recover anyone; it just trains the customer to ignore you faster.

A general-purpose CRM stack is built to do one thing well: send. It pushes batch campaigns through owned channels and measures opens and clicks. That is delivery, not recovery. Asked to win back the lost, it does the only thing it knows — it sends, and when nothing happens it discounts harder, which accelerates exactly the loss it is trying to reverse.

Recovery is two steps, in a fixed order: attention first, transaction second. Earn the right to be heard again, then ask for the sale. A delivery tool skips the first step entirely, because it has no engine for it. Lined up against what recovery actually requires, the gaps in a send-first stack are structural, not incidental:

What recovery requires	What a CRM stack does instead
Re-earn attention before asking	Built to send the next campaign
Sequence attention, then transaction	Optimised for immediate response
Decide customer-by-customer	Runs on rules and broad segments
Learn across many brands	Each brand isolated in its own instance
Be paid on the recovery it proves	Charges a fixed fee regardless

You cannot retarget your way back into a relationship. You have to earn the attention first — and that is a job for two purpose-built engines, not one general-purpose stack.

2

Atrium — the engine that earns the attention back

Atrium is the attention engine, and in recovery it is pointed at the lost. It has three parts. NeoMails are the Relate vehicle — relationship content that stands on its own, independent of any transaction; email today, extensible to WhatsApp and beyond. ActionAds are the monetising unit that can sit inside a NeoMail. NeoNet is the cooperative network that places those ActionAds across brands.

Recovery is a two-step job: Atrium earns attention back; Meridian converts it.

On a lost customer, Atrium does the opposite of what a CRM stack does. It does not lead with an offer. It leads with utility, recognition and relevance — a reason to open that is not a discount — because the first job is to rebuild reachability, not to extract a sale. Only once a customer is opening and engaging again has Atrium produced anything worth converting.

Atrium is also designed to pay for itself, and it is worth being honest about the status of that claim. The intent is that ActionAds fund the NeoMails on a ZeroCPM basis. That is a design goal the early pilots must prove, not a fact to assume — the first real test is simply whether attention yield covers a meaningful share of sending and content costs. The output of Atrium is not a sale. It is a customer who is reachable and paying attention again — the raw material the second engine needs.

3

Meridian — the engine that converts restored attention

Meridian is the conversion engine. It is a proprietary model paired with an Alpha Agent that turns earned attention into transactions. It is the part of Progency that is never sold and never shown — but it is not magic, and it is worth being precise about what it does even while keeping how it does it closed.

Meridian decides. It prioritises which lost customers are worth recovering at all; it decides, for each one, whether the next move is to ask for more attention or to ask for a transaction; it selects the content, the offer, the channel and the timing; it learns from what happened across cohorts; and it writes every decision and every outcome back to the Context Graph. That last habit is what makes the system improve rather than merely run.

What ships to the brand and what stays inside are two different things. The in-house team, on CRM 2.0, gets the M-Agents — Insights, Audience, Content and Decisioning — visible, useful, and theirs to run. Meridian itself stays in the black box. The analogy is a quant fund: the model is the secret the fund is paid for, not a product it ships. A brand buys Meridian’s results through Progency; it does not licence Meridian, any more than an investor licences the fund’s strategy.

4

Building Meridian — the data substrate

Recovery is a per-customer sequential decision problem: for each lost customer, a series of choices over time, each depending on what happened before. A campaign tool has none of the three things that problem needs — state, per-customer models, and a memory of outcomes. Meridian is built around supplying all three.

State and memory live in three Context Graphs, with the intelligence substrate above and the Alpha Agent on top.

State and memory live in three Context Graphs. The Customer Context Graph holds who each customer is. The Product Context Graph holds the catalogue, affinities and substitutes. The Decision Trace Graph holds every decision the system has made, with its context and its outcome. The Decision Trace is the important one: it is both the recovery memory and the auditable record of actions to outcomes that the commercial model depends on.

Beneath the graphs sits the intelligence substrate that lets recovery reason about one customer at a time. BrandTwins are per-customer models, maintained at scale by TwinFactory and built on ArtificialPeople — the world-model behavioural archetypes that give a twin its priors before much first-party data exists. This is the layer that makes N=1 recovery possible: a decision tuned to this customer, not to the segment they happen to fall in.

5

Training the Alpha Agent

The Alpha Agent is a decision policy, not a content generator. Its output is choices: who to work, when to ask for attention versus a transaction, what to send, when, and through which channel. Building it is a question of how a policy like that is trained — and the honest answer has four parts.

First, the cold start. With no history, the agent is bootstrapped by behavioural cloning — it learns to imitate the best human operators. This is the quiet purpose of the Martech Growth Engineer pilots: every decision an MGE makes, and the outcome it produced, is a training example. The humans running today’s recoveries are writing the training set for the machine.

Second, the reward. The agent is not trained to maximise opens or clicks, which are easy to game and weakly tied to money. It is trained against measured Alpha — the holdout-verified lift, the same three-arm control that prices Progency commercially. The number that pays Progency is the number that trains its agent. That makes this closer to offline reinforcement learning and contextual bandits than to ordinary supervised learning.

A compounding loop: act, measure the lift against a holdout, write the outcome, update the policy — a little better each time.

Third, simulation. Live experiments are slow and costly, so candidate policies are rehearsed offline first. ArtificialPeople provide a sandbox to score a policy before it is deployed, and BrandTwins let the agent ask ‘what would this specific customer do if…?’ at N=1. Simulation reduces the cost and risk of the cold start — but it only approximates reality, so the live holdout remains the ground truth that overrides any simulated result.

Fourth, cross-brand learning. Because every brand’s recoveries write to the same kind of Decision Trace, the agent learns patterns that hold across many brands — more brands make recovery sharper for all of them, done privacy-by-design with federated patterns rather than raw personal data crossing brand boundaries. That single design choice is both the moat and the privacy answer: the network effect lives in the learned patterns, not in shared customer lists.

The build, in one line.  The same holdout that prices Progency is the reward that trains the agent — and the MGEs running today’s pilots are writing tomorrow’s training set.

6

Why two engines beat one stack

A general-purpose CRM stack is asked to acquire, grow, protect, recover and report — and is therefore optimised for none of them. Progency does one job with two engines tuned for exactly that job, and the two-step sequence is the whole reason it works. Atrium does the step the stack skips — re-earning attention. Meridian converts what Atrium restores. Neither engine alone is enough; sequenced, they recover customers a delivery tool cannot.

Old win-back	Progency recovery
Send the offer	Earn the attention first
Discount harder	Rebuild relevance
Measure campaign response	Measure attention restored, then conversion
Retarget through paid if no response	Recover before falling through to adtech

And the honest limits travel with the design. Recovery needs enough lost-customer volume to run a control group. It works where customers are identifiable. Atrium’s self-funding economics are a goal the pilots must prove, not a settled fact. NeoNet is emerging, not finished. And recovery sits between CRM and adtech — it replaces neither.

* * *

The CRM stack was built to send. Progency was built to recover — and recovery, it turns out, is a different machine: one that earns attention before it asks for a sale, decides one customer at a time, and is paid only for what it can prove it brought back. Never Lose Customers. Never Pay Twice. Never Pay Fixed.