For most of its commercial history, email has been sold as transport. A provider accepts a message, renders it, routes it, delivers it and reports what happened. The unit of economics is the send: cost per thousand, cost per million, cost per campaign. The industry has spent decades making that transport cheaper, faster and more reliable, and it succeeded. The price of a send fell and kept falling, and the product became a commodity.
That was the right competition for its time. But the jobs email performs for a brand have moved well beyond transport. A customer can now act inside a message. A service notification can collect information the customer chooses to give. A publication can retain attention between transactions. A dormant customer can be recovered through the brand’s own relationship before it pays to buy them back. The value of what happens after the send has grown. The pricing has stayed attached to the send.
This essay argues that the gap is closing, and that it should. Its thesis is not that email should cost more. It is more demanding than that:
Price should move as close to business value as measurement allows.
That produces a new architecture for email: four jobs, two capabilities that change what a single message can do, and a ladder that runs from delivery economics to outcome economics. The previous essay in this series, Before the Slide, argued that marketing intervenes too late in the customer’s life. This one argues that the email business has been measuring and pricing the wrong thing. The old email business was paid-to-send. The new one gets paid to make customers act, stay and come back.
1
The Send as the Unit of Value
When email was transport, charging per send made sense. Its success created a trap.
The per-send model fitted the era that produced it. The provider ran the infrastructure — queues, IP addresses, authentication, rendering, throughput, retries, deliverability — and the brand bought reliable capacity. Those things were measurable, comparable between providers and roughly proportional to the number of messages sent. Volume was the natural unit.
But the unit shapes the incentives. As sending became cheaper, the easiest way for a provider to grow was to carry more of it. As email became cheaper than almost any other way of reaching a customer, the easiest way for a marketer to use the budget was to send more of it. Neither side was paid for what happened after delivery. When revenue rises with message count, restraint is economically unnatural: the provider can win even when the customer is receiving too much, and the brand can hit activity targets even as each additional message earns a little less attention than the one before. The result is familiar — large lists, high volumes, declining engagement, and a widening gap between delivered and noticed.
There is a further complication. Even the open, the industry’s favourite proxy for attention, is no longer clean. Apple’s Mail Privacy Protection downloads remote content in the background whether or not the recipient reads the email, precisely so that senders cannot reliably tell whether it was opened. That does not make email measurement impossible. It makes the old proxy a poor basis for value.
The question that matters now is what the email caused the customer to do, remember, continue or recover. Once that is the question, a single unit of price cannot fit every kind of email.
2
Four Jobs: SNDR
Every brand email should have one primary job: Sell, Notify, Digest or Reactivate.
The simplest way to see the change is to stop treating email as one thing. A brand’s emails do four different jobs, and the jobs line up against the state of the customer receiving them.
Figure 1. The four jobs of email, across a customer who is active, then fading, then dormant.
Sell is for the customer who is ready, or close to ready, to act. The job is to compress the distance between intent and transaction. Notify is for the customer who needs information or service: an alert, a confirmation, a statement, a renewal, a delivery update. The job is to serve them, and to learn from them while doing it. Digest is for the customer who is still in the relationship but may have nothing to buy today. The job is to retain their attention between transactions — the missing middle described in Before the Slide. Reactivate is for the customer who has gone quiet or dormant. The job is to recover them through the brand’s own relationship before paid media is needed.
Readers of earlier essays will notice a change. I previously called the fourth job Relate: communication that keeps the relationship alive without a transaction. That made sense when relationship content had no other home. It now has one. Digest carries that work, on a regular rhythm, so the fourth job can be sharper: Reactivate. NeoMails, the family of attention-earning email formats, sit under Digest.
The phrasing of the key sentence is deliberate. SNDR is a design discipline, not a claim that every email fits exactly one box. A Notify email may carry a small Sell action; a Digest may occasionally lead to a purchase. What matters is the primary job. If a brand cannot say what an email’s primary job is, it will struggle to decide its content, measure its result or choose its price.
This also moves email strategy away from the vocabulary of formats. “Newsletter”, “blast”, “trigger”, “transactional” and “promotional” describe how a message is built or sent. SNDR describes what the customer needs the message to do. And it changes the planning question. Instead of asking how many emails to send, a team asks four separate questions: how often do customers need to be sold to, notified, kept attentive and reactivated? There is no reason to assume those cadences are the same.
The four jobs divide into two pairs. Sell and Notify make existing email work harder. Brands already send these messages; the opportunity is to make each one more effective. Digest and Reactivate solve customer loss. They create value email has rarely been asked to deliver. The pairs need different products and — as the rest of this essay argues — different prices.
3
The Two Foundations: Get Seen, Get Action
Two capabilities change what a single email can do.
Get seen. Delivery is not enough. An email that lands in spam, is buried, or consistently reaches a part of the inbox the customer rarely checks is worth less than one that lands where they look. Better placement changes the probability of being seen at all. At Netcore we call this layer NiVO. The important word is measurable: no sender can promise a particular inbox tab to every recipient, because placement depends on the mailbox provider’s algorithms, the recipient’s behaviour, authentication, reputation and content. The case for a placement premium therefore has to rest on evidence — panel or seed measurement against the brand’s current setup, with a clear baseline.
Get action. A conventional email asks the customer to click out, open a site or app, log in, navigate and then act. Every step loses people. AMP for Email lets supported email clients render interactive components inside the message, so the customer can answer, choose, confirm, respond and, in time, begin a payment without leaving. The point is not animation. It is friction removed.
Figure 2. Two foundations under Sell and Notify. Digest and Reactivate use the same rails, but are sold on results.
Both come with limits that belong in the product design, not hidden in the pitch. AMP renders only in supported clients — for many consumer markets, most importantly Gmail — so every AMP email travels with an HTML or plain-text version in the same message, and that fallback has to be designed first rather than treated as a courtesy. Gmail also routes AMP requests through its own servers and strips cookies, so any interaction that needs the customer to be recognised has to use secure access tokens rather than an ordinary logged-in session.
The two capabilities should also be kept separate in thinking. Getting seen is a placement problem, and its evidence is inbox reach. Getting action is an interaction-design problem, and its evidence is completed tasks and reduced friction. Bundling them can create a stronger product. Stacking them as two surcharges creates the wrong conversation with a procurement team. The commercial case should come from the combined improvement for the customer, not from the provider’s cost structure.
4
Sell and Notify: Making Existing Email Work Harder
If the email measurably does more, it can be worth more.
For Sell, the product is the compressed journey. Yes in Email lets a customer accept an offer, confirm a booking or approve a renewal without leaving the message. Pay in Email extends that towards payment. The claim is not that every payment must settle inside the inbox; it is that as much of the decision and authorisation as possible should happen there, with a secure hand-off where the payment system requires it. The value can be tested directly against a plain-HTML control: more completions, fewer abandoned steps, less time to act.
For Notify, the product is the AMPlet: a small interactive unit inside a message the brand already has a reason to send. An order confirmation can ask for a delivery preference. A statement can let the customer update a detail. A renewal notice can capture a yes or no. A service message can ask one question that improves future personalisation. This is data the customer gives intentionally because the request is useful at that moment — better than anything inferred from a click.
Advertising inside Notify messages needs care. Service messages are the most trusted mail a brand sends, and commercial content can damage that trust or change how the mailbox provider treats them. In regulated alerts it may be inappropriate altogether. The first use of Notify is service action and data; monetisation is a later and conditional layer.
A single design principle covers both jobs: remove one step. If a Sell email removes a login, a page load or a form field, it has created value. If a Notify email removes a call to customer service, an app visit or a separate trip to a preference centre, it has created value. The measures follow the promise: completion and abandonment against a control for Sell; service resolution, data captured and avoided contact for Notify. That is why better email should not be sold as richer creative. Richness is not the outcome. Fewer steps are.
The pricing principle follows. Package the capabilities into clear products rather than stacking a feature tax. Charge the premium where the capability renders and where the improvement can be demonstrated. Over time, some Notify uses can move closer still to value and be priced per completed action rather than per message.
5
Digest and Reactivate: Solving Customer Loss
When email retains attention or recovers a customer, it should be paid for the result.
Digest and Reactivate are different in kind. They are not upgrades to a message the brand was going to send anyway. They address value email has rarely owned: keeping a customer’s attention between transactions, and recovering it once it has gone. Before the Slide set out the underlying economics — customers go dormant in attention before they go dormant in purchases, and marketing spends heavily at both ends of that slide and almost nothing in the middle. Brand Digest slows the flow into dormancy. Progency Recover reverses it. Neither is a better send, so neither should be priced as one.
For Digest, the first useful unit is not a send and not a raw open. It is an Engaged Edition: one person, one edition, at least one verified human action. Ten actions in one edition are still one Engaged Edition. The rule fits in four words — no attention, no charge — and it protects the product from bad incentives. Billing on every click would reward the provider for adding clickable clutter; billing on opens would reward subject lines that overpromise. One unit per person per edition asks only whether the edition earned human participation. Because the reader chooses how often the publication arrives, a daily reader can be worth more than a twice-weekly one — but only because they chose it.
Where a Digest carries advertising, the brand’s share of that revenue can be credited against the bill, so that a well-read publication can end up costing the brand little or nothing. That is a state the economics can reach, not a promise made on day one, and its accounting belongs in a contract rather than an essay.
At scale, both jobs point to the top of the ladder: pricing on lift verified against a concurrent holdout. For Digest, that means slower migration from engaged to quiet and, over time, lower reacquisition. For Reactivate, delivered as a done-for-you service, it means customers and revenue recovered.
Figure 3. Two commercial logics: premium messaging for better sends; engagement and outcomes for customer loss.
6
The Pricing Ladder
Price moves as close to value as measurement allows.
Put the four jobs together and the commercial model forms a ladder with four rungs. This ladder, rather than any single price, is the essay’s central idea.
Figure 4. The pricing ladder. Each rung sits closer to value and demands stronger measurement.
- Per send. The provider is paid for transport. The measurement question: was the message sent and delivered?
- Premium send. The provider is paid more because the message has added capability or measurably better placement. The question: did it render, did placement improve, did completion improve?
- The provider is paid when a person does something verifiable. The question: did the reader take part in this edition?
- The provider is paid from value added relative to a concurrent holdout. The question: did the intervention change the result?
Each rung needs better measurement than the one below it, which is why pricing innovation cannot be separated from measurement innovation. Outcome pricing without an agreed baseline becomes a negotiation. Engagement pricing without a clean event definition becomes a billing dispute. A placement premium without evidence becomes a claim. The closer price moves to value, the more precise the measurement has to be.
Nor is the ladder a migration schedule. A highly standardised notification may stay on a premium send because that is simple and auditable. A Digest may begin on engagement pricing because its long-term economic effect cannot yet be seen. A mature recovery programme may go straight to outcome pricing because its success event is already clear. The principle is to choose the closest measurable unit of value without pretending to measure what cannot yet be measured reliably.
This is Never Pay Fixed in practical form. It does not mean everything becomes variable at once: transport still has costs, infrastructure still matters, and brands still need predictable budgets. The point is directional. As value becomes observable, more of the economics can depend on value delivered. Pricing, in other words, is evidence for the larger change: email is moving from carrying messages to carrying actions, attention and outcomes.
7
What Changes for the Email Provider
A provider paid for outcomes has to stop optimising for sends.
The most important consequence of outcome economics is not the invoice. It is the provider’s behaviour. A provider paid on volume has an obvious growth path: more campaigns, more recipients, more frequency. A provider paid for engaged attention or recovered customers has a different one: make the communication worth receiving.
Figure 5. The incentive shift, from maximising sends to maximising wanted, repeated attention and results.
That flips several operating habits.
- Never optimise for sends. Optimise for the frequency readers choose and the engagement that follows.
- Let volume grow only when readers want more. A publication can be produced every day, but the customer chooses how often it arrives. The best high-volume reader is not the one who was messaged seven times; it is the one who asked for seven editions and keeps engaging with them.
- Measure the next interaction, not only this one. A subject line that wins today’s click and costs tomorrow’s trust is a bad optimisation.
- Use concurrent holdouts. If the provider is to share in improvement, the baseline cannot be whatever happened last quarter. It has to be the brand’s current best effort, running at the same time.
- Keep the ledger auditable. When pricing depends on events and outcomes, brand and provider need a shared record of actions, cohorts, exclusions and results.
It changes product management too. A volume business asks, “How do we make campaigns easier to launch?” An outcome business has to ask, “Why will the customer choose to engage again?” That question pulls the provider into editorial quality, interaction design, experimentation, identity, measurement and incrementality. It demands more accountability. It also offers a way out of commodity pricing. The provider stops being only the pipe and starts sharing responsibility for what the pipe produces.
8
What Would Prove This Wrong
The architecture should be tested as hard as the products inside it.
- Brands will not pay for better sends. Once interactive email and better placement are measured against a control, the improvement proves too small to justify a premium.
- Engagement cannot be metered cleanly. Brands and providers repeatedly dispute whether an action was human, meaningful or billable.
- Advertising demand does not arrive. The ad credit stays theoretical because too few advertisers buy the inventory at sufficient quality.
- Outcomes cannot be bought on. Holdout-verified lift proves too slow, too noisy or too contested for brands to accept meaningful variable pricing.
There is a broader risk too. Customers may not want more interactivity in email; they may prefer the inbox as a simple notification layer and the app or website as the place where things get done. If so, some of these capabilities will stay niche. That is why the new economics has to be earned use case by use case. The point is not to decree that all email becomes outcome-priced. It is to stop assuming that every valuable email job must be priced like commodity transport forever.
9
Two Transformations
This essay and Before the Slide are two halves of one argument. One changes when marketing acts: from the transaction and the loss to the attention in between. The other changes what email gets paid for: from the send to the result.
Figure 6. The customer-side and provider-side transformations, side by side.
Underneath both is a single shift. Marketing has mostly treated attention as something to consume: bought, spent on an offer and bought again. The new email model treats attention as something to retain, and prices email by what it does with the attention it is given.
Email became one of marketing’s most important channels because it was cheap to send, easy to measure and owned by the brand. Those advantages remain. What changes is the unit of ambition. A Sell email should not be judged only on delivery if the customer can complete the action inside it. A Notify email should not be judged only on receipt if it can resolve a service task or collect useful data. A Digest should not be judged on an open if it can retain voluntary attention over months. A Reactivate programme should not be judged on message volume if it can recover customers before the brand pays to buy them again.
The email industry spent thirty years improving the cost of sending. The next decade will be about improving — and pricing — what happens after the send.
The old email business was paid-to-send. The new one gets paid to make customers act, stay and come back.
Key points
- Email’s old unit of value was the send. It fitted an era of transport, and it rewards volume over outcomes. Even the open is no longer a clean measure of attention.
- Every brand email should have one primary job: Sell, Notify, Digest or Reactivate. SNDR is a design discipline. Relate is retired; its work now sits under Digest.
- Two capabilities change what an email can do — get seen and get action — each with limits that belong in the design: supported clients, HTML fallback, secure tokens, measured placement.
- Sell and Notify make existing email work harder. Remove one step; price the measured improvement in bundles; move some Notify uses towards per-action pricing.
- Digest and Reactivate solve customer loss. The Engaged Edition — no attention, no charge — is the first unit; verified lift is the destination.
- The pricing ladder — per send, premium send, engagement, outcome — is the central idea. Each rung needs better measurement, and the ladder is a principle, not a schedule.
- A provider paid for outcomes must stop optimising for sends, measure the next interaction, and share responsibility for what the pipe produces.
