A founder I know runs an EU-licensed crypto exchange. €290k a year on senior marketing — agency, law firm retainer, in-house lead. When I asked him what he thought he was actually paying for, he gave the answer everyone gives: "people doing work." When I asked what he was paying for if no work needed to be done that month, he had no answer. The retainer model is most easily described by what its buyers cannot describe.
A founder I know runs an EU-licensed crypto exchange. Fifty people, two product lines, a regulatory perimeter that gets stricter each quarter. Last year, his marketing budget had three line items: a senior in-house marketing lead at €120k, a paid-media agency at €8k/mo, and a crypto-specialist law firm on retainer at €6k/mo to clear marketing copy. The three together cost about €290k for the year. The marketing function shipped, regulator letters were avoided, the agency over-delivered on Q3 paid acquisition, and the firm and the team got along.
When I asked him what he thought he was actually paying for, he gave the answer that everyone gives, which is that he was paying for "people doing work." When I asked him what he was paying for if no work needed to be done that month — if no campaigns shipped, no press inquiries came in, no regulator wrote, no funnel needed redesigning — he had no answer. The retainer model is most easily described by what its buyers cannot describe.
This is the gap that a new pricing model is going to fill. I think — and most of what follows is downstream of this prediction — that within the next two to three years, the senior-expertise market will reorganize around a coverage-and-standby model rather than a retainer-and-deliverable model. Not because retainers are bad. Because AI made standby capacity scarce in a way it wasn't before, and the market always reprices the scarce thing.
Three observations sit underneath that prediction.
A senior marketer running a function five years ago was responsible for two genuinely separate kinds of output. The first was judgment — pricing decisions, positioning calls, risk reads on campaigns, hiring choices, narrative selection. The second was production — writing the brief, drafting the copy, building the deck, structuring the report, reviewing the campaign assets, monitoring the dashboards, summarising the Slack threads.
A senior marketer in 2026 is responsible for the same first thing and a much-shrunken second thing. Production has not gone away, but the marginal cost of production has fallen by something close to an order of magnitude for everything that can be expressed as text, structure, or pattern-match. The senior marketer's value used to be 30% judgment and 70% production-with-judgment-applied. Now the same value is 70% judgment and 30% production-with-judgment-applied, because the AI absorbed most of the bulk of the second category.
This is not a story about the senior marketer being replaced. The judgment is what was scarce all along; the production was what made the scarcity expensive to access. When you remove the production tax, the same scarce judgment becomes available at a price that would have been impossible to charge for it before. The senior marketer hasn't been disrupted. The way their time is sold has been.
The same EU-licensed exchange founder, when I pressed him a different way, did know one thing he was paying €290k/yr for: he was paying for not having to think about it. The agency-and-counsel-and-in-house combination didn't have to ship work every week to be valuable. They had to be available. They had to know the field. They had to pick up the phone when a regulator wrote, when a campaign caught fire on X, when a competitor announced something that needed a same-day response, when a customer comms event broke at 11pm.
Most of the senior expertise he was paying for, in other words, was insurance in the everyday-language sense — coverage against a category of bad outcomes whose timing he couldn't predict. Marketing-execution work was a useful side-effect of having that coverage, not the main thing he was buying.
The buyer side of the senior-expertise market has always known this implicitly. What's changed is that the supply side is starting to be able to say it. Once production becomes cheap, the standby capacity is the load-bearing piece — and standby capacity is, structurally, insurance.
A retainer prices a thing the buyer doesn't actually want. The buyer wants protection; the retainer prices hours-or-deliverables. When the seller delivers a lot of hours-or-deliverables, the buyer is happy in a busy-feels-productive sense and uneasy in an am-I-actually-getting-value-here sense. When the seller delivers few hours-or-deliverables, the buyer is uneasy in a what-am-I-paying-for sense even if the seller is doing the more important work, which is being available and watchful.
Retainer pricing therefore has an incentive structure that pushes both sides away from the actual value. The seller produces busywork to justify the retainer. The buyer cuts the retainer when the busywork looks expensive relative to the visible activity. The relationship cycles between "we should add an SOW" and "we should reduce the retainer," neither of which addresses what the buyer actually wants.
Insurance pricing solves the same problem differently. Insurance prices the thing the buyer wants — coverage. The buyer pays a known monthly premium. The seller defines, in published SLAs, what is and is not covered, what response times are guaranteed, what the cap on any single event is, what the escalation tree looks like. Both sides know what's true. Loss-ratio is calculated across the customer base, not within any individual relationship. The buyer doesn't feel guilty about quiet months. The seller doesn't have to manufacture deliverables. Premiums can rise as coverage scope grows, the way Apple Care premiums rise with device value or D&O premiums rise with company size — without the buyer feeling cheated.
This is a better-priced product in the literal economic sense. The buyer is paying for something whose value they can articulate and verify against the SLAs. The seller is selling a margin profile that scales with risk-pooled customers, not with linear hours.
The first thing people object to, when this argument is made, is that "insurance" implies a regulated insurance product. That objection is correct in the legal sense and wrong in the productisation sense. Insurance is a regulated word and a regulated category in the EU under IDD and Solvency II. A senior marketer cannot sell "insurance" as a literal product label without becoming an insurance broker. That is real and I take it seriously.
But the metaphor is what gets the buyer to understand what they're buying. Apple Care isn't insurance in the regulated sense. Stripe Atlas isn't a corporate-services package in the legally-regulated sense. They are productised stand-ins for things buyers were previously paying for piecemeal — and the metaphor is what made the productisation possible.
When marketers and consultants get the framing wrong, they default to the "agency replacement" or "fractional executive" metaphors. Both of those frame the offering as a service — and services price by hours or by deliverables, which puts the offering back inside the broken retainer model. The insurance metaphor reframes the offering as a package of coverages — and packages of coverages price by premiums, with SLAs, with cap structures, with risk-pooling. That is a structurally different transaction from "you have someone working for you."
I think the model spreads from law to medicine to senior marketing to accounting to compliance to executive coaching, in roughly that order, over the next 36 months. Each industry has its own reasons for resisting the reframe and its own particular form of buyer that will demand it. The crypto industry — where I happen to operate — is the canary, because crypto's regulatory perimeter shifted three times during MiCA's rollout, and most companies operating in it cannot afford a full-time senior marketer or a €15k/mo fractional CMO and have therefore been buying neither and shipping copy that gets flagged in real time. They are the buyers most ready for an insurance-style coverage product, because the alternative is genuinely "ship marketing without senior judgment." The cost of the alternative is no longer hypothetical for them.
The same buyer profile exists in regulated medical devices, in regulated fintech, in regulated alcohol/cannabis adjacencies, and increasingly in any AI product touching consumer-facing claims. Same dynamics: regulatory perimeter expanding faster than the company can hire for, fractional senior counsel too expensive for the stage, in-house junior marketer producing copy that the company couldn't defend if asked. The match between this set of buyers and an insurance-style coverage product is the kind of match that is very obvious in retrospect.
If you're a buyer of senior expertise in 2026, I think there are three things to expect.
First: the price of senior expertise will rise on the highest tier and fall on the entry tier. Tier 3 fractional CMO retainers — the €15-30k/mo product — will hold price or rise, because the people qualified to do that work are individually more leveraged by AI and therefore each one of them does more. Tier 1 / Tier 2 productised access to the same expertise — the €1-5k/mo product, which mostly didn't exist as a category five years ago — will become a major segment. The middle gets crushed, in roughly the way the middle has been getting crushed across most expertise markets.
Second: the SLA is the product. Previously, "what does this consultant actually do?" was the question that determined whether the engagement was working. In an insurance-priced model, that question is replaced by "what does the SLA say, and is the seller honoring it?" That's a much easier question to evaluate. Buyers should expect to read SLAs and write SLAs into their contracts; sellers who cannot publish SLAs are signaling that they are still selling under the old model.
Third: standby capacity is the unit of value. If you are paying for a senior marketing operator to be available, demand to know how their availability scales with their customer base. A senior who has 6 customers on standby has different bandwidth than one who has 30. A coverage product with no cap on subscribers is a coverage product whose SLA you cannot trust under load. Look for caps. Look for tiers. Look for someone who is honest about what their capacity is, because the honesty of that disclosure is the honesty of the entire offering.
If you're a senior expert thinking about productising your own offering — which is most of the audience for this kind of essay — three things to consider.
First: write the SLAs first. Before pricing, before naming, before building. Sit down and write what you will actually guarantee, in time-bounded terms, to a paying customer. Most senior experts cannot do this exercise the first time and discover, in the writing, that what they thought they offered isn't what they actually offer. The discomfort of that exercise is the productisation work.
Second: define the cap structure before the sales motion. If you offer "unlimited" anything, you are pricing as a retainer with a different name. Insurance has caps because reality has caps. The customer should know, before they pay, what the boundary is and what happens past it. Building this before the first sale is the difference between a sustainable productised practice and a slow burn into the same retainer death-spiral.
Third: the metaphor is in the body copy, not the product label. "Marketing insurance" is what people understand. "[Your discipline] insurance" gets them to understand what they're buying without making promises you can't legally keep. Use the metaphor liberally. Avoid using the word as a literal product label or in pricing pages.
Most of what I've described above is what we built at NorthPoint for our crypto-marketing subscription product, AI Crypto CMO. The product wallet has four pillars — Coverage, Emergency, Signal, Oversight — priced at €2,500/mo against an assembled-stack equivalent of €18-50k/mo. We are early enough in design-partner stage that I cannot prove this works yet — but the math is the math, and the price math is the part of this argument I'm confident enough about to put in writing in advance.
The wider argument — that senior-expertise pricing is reorganising around coverage models — does not depend on whether NorthPoint specifically is right. I think it would be true if NorthPoint did not exist. NorthPoint is the version of the bet I am personally placing.
If you want to see what an insurance-priced senior-marketing product looks like in practice, the deep page is at northpoint.fi/ai-crypto-cmo. If you want to run a free MiCA check on your own marketing copy as a way of evaluating whether the underlying IP is real, that's at northpoint.fi/check/mica.
The product this essay is the philosophy of. Four pillars: Coverage, Emergency, Signal, Oversight. From €2,500/mo. Three design-partner spots open at €1,500/mo for the first three months.
The tactical companion. Five MiCA marketing checks an ex-exchange CMO runs before any banner ships, walked through on a composite token-yield example.
A buyer-side decision tree for the older retainer-style choices. Useful as the contrast frame for the insurance-priced argument above.