Whoever declares something dead usually sells the replacement.

There's a proposal on the table. Account management can come down by a third, because buyers prefer to work things out for themselves these days. There's a percentage next to it and the name of a research firm. The numbers are unambiguous.

Nobody at the table asks what year the research is from. Or which market it covered. Or what the same report said two paragraphs further on.

Three months later the headcount is gone and the work turns out not to have disappeared, but moved. Except now nobody is sitting in the chair it moved to.

Why a death notice lands so well at the board table

A death notice gives you the feeling of a decision while it’s actually avoiding one. X is dead, so we cut X. No diagnosis, no trade-off, no ratio to work out. You’ve chosen without choosing.

Richard Rumelt calls that the core of bad strategy: the form of a decision without the substance. Bad strategy rarely comes from stupidity. It comes from avoidance. The hard part gets skipped and what’s left looks like decisiveness. It is exactly the kind of work that belongs at the board table and, in most companies, never gets there.

I checked four death notices from the past year against the research. In three of the four, the number in the proposal didn’t match the number in the study. That’s a finding in itself.

Four death notices, checked

The account manager is redundant. Gartner reported in March 2026 that 67 percent of B2B buyers prefer to buy without a salesperson. That number travelled across LinkedIn all spring. Two months later Gartner published a second number from the same research wave: 69 percent go to a salesperson anyway, to check whether what the AI told them holds up. Same sample, same questionnaire, roughly 645 buyers.

That isn’t a contradiction. It’s one buyer with two needs. Gathering information he can do himself. Being willing to carry a decision is something else. And note what was actually measured: a preference, not behaviour. Only 45 percent actually used AI in their most recent purchase.

The middleman gets skipped. The standard case is Nike, which announced in 2017 that it would focus on roughly forty partners. That gets retold as “from 30,000 to 40”. It never went that way. Nike differentiated its network and concentrated investment and exclusive product with forty partners. The rest carried on selling.

What happened afterwards matters more. In 2024 Nike reversed the direct-to-consumer course and invested in the channel again. In the most recent quarter, sales through retail partners grew 5 percent while its own direct sales fell 4 percent. The best-known exhibit under this death notice is now the evidence for the opposite.

The reason is prosaic. Selling direct doesn’t remove the work, it moves it into your own warehouse. Stock, single orders, returns, a rush job on a Friday afternoon. The margin per order goes up, and so does the cost of handling that order.

The trade show is dead. The same funeral every year, and every year the halls are full. Both claims are too easy. In the United States, B2B trade show attendance at the end of 2025 was still around 6 percent below its 2019 level, according to the CEIR index. In the Netherlands, the Monitor Beurzen for 2024 counted 15 percent more visits and 19 percent more stand space than the year before. Two markets, two directions. Pick either number and you can tell whatever story you like.

The real question isn’t whether the trade show is alive, but what you hold it accountable for. On scanned badges, a trade show is almost always a bad investment. On twenty conversations you can’t otherwise arrange in a year, usually not. And the loss is rarely in the show itself. It’s in the weeks afterwards, when almost nobody follows up on those twenty conversations.

SEO is dead. In March 2025 Google gave a number of its own for the first time: more than five trillion searches a year, over thirteen billion a day. That’s considerably higher than the 8.5 billion that has been doing the rounds in presentations for years. So search isn’t shrinking, it’s growing.

At the same time, in the first four months of 2026 roughly 68 percent of US Google searches ended without a click to an external site, according to SparkToro and Similarweb. Zero click, it’s called. How often an AI Overview sits at the top, that block of machine-written answer above the blue links, honestly, nobody knows precisely: the measurements range from 16 to over 40 percent, depending on who measures and how.

More searches, fewer clicks. That isn’t a death. It’s a split into two ways of being found: the search result itself, and the answer built on top of it. Someone has coined a word for the second one, GEO, generative engine optimization. It isn’t a choice between the two. You can be present in both.

The pattern under all four

Look at who pronounces the death notice. In most cases he sells the replacement. Whoever sells GEO declares SEO dead. Whoever sells software declares the salesperson dead. That doesn’t make him a liar, but it does make him an interested party. Reason enough to look up the research yourself.

And look at what none of the four mention. In what ratio, then?

That’s precisely the hard part. A binary you can explain in an eighty-word post. A ratio you have to work out, for your market, your customers, your margin. Which is why you hear the binary everywhere and the ratio almost never.

Every death notice is a diagnosis somebody else made for you

This is where the strategic problem sits, and it’s bigger than a wrong number.

Commercial problems come in three kinds. They don’t know you at the moment that counts. They don’t choose you, even when you’re better. Or they don’t stay, and you don’t see it coming. Discover, choose, stay. Every company gets stuck on one of those three at some point, and which one it is determines where the first euro should go.

Every death notice silently points at one of them. “SEO is dead” and “the trade show is dead” are about discovering. “The account manager is redundant” is about choosing. “The middleman gets skipped” is about who carries the work after the purchase, and therefore about staying.

Adopt the death notice and you’ve let an outsider determine which track your problem sits on. Someone who doesn’t know your revenue per customer group, hasn’t seen your quote outcomes and has never spoken to the buyers who walked away. He does have one thing: a product on that track.

Roger Martin says strategy consists of choosing where to play and how to win there. A death notice answers that first question for you, for free, and wrongly. That’s an expensive free service.

The ratio is the work

The best-known ratio in our trade is Les Binet and Peter Field’s 60/40: sixty percent of the budget to brand building, forty to direct activation. What usually falls away is that it’s an average from a database, not a norm. New entrants sit higher, established brands lower. Byron Sharp has publicly attacked the rule. WARC found in 2024 that shifting from activation only to brand building plus activation raised returns by ninety percent on average, and the other way around lowered them by forty percent.

Who’s right about 60/40 isn’t the interesting part. The interesting part is that the entire discussion is about a ratio, and not about which half gets to stay.

That ratio is different for you than for your competitor. Different deal size, different margin, different customers, different sales process. Which is exactly why you can’t lift it from a post. And exactly why a competitor can’t lift it from you. A binary is free to copy. A ratio is a position.

What that means concretely for the work

Per track, the starting point shifts from “which half do we cut” to “which ratio do we hold”. That changes the work.

Discovering. The starting point is that you get chosen from what someone already knows, not from what happens to be findable at that moment. The Ehrenberg-Bass Institute estimates that at any given moment roughly five percent of business buyers are actually in the market. That’s a rule of thumb based on purchase cycles, not a measured constant, but the order of magnitude holds. So the ratio is: how much money goes to the people searching today, and how much to the far larger group that will search in two years and have one name ready.

The work that follows isn’t “replace SEO with GEO”. It’s: the sentence that sticks, the three to five places your buyers actually are, and arranging your texts and sources so that you appear both in the search result and in the answer built on top of it. Cut one of the two in response to the death notice and you halve your visibility and call it focus. That’s the core of Spark. Discover.

Choosing. The two Gartner numbers aren’t a contradiction but an instruction. Buyers want to work it out themselves and have someone confirm what they found. That means two kinds of proof, for two different moments. Material that does its job when nobody is there, and material a salesperson reaches for at the moment the buyer hesitates. Most companies have the first half-finished and the second not at all.

And because a B2B company with fifty quotes a year has no volume to test with, you can’t calculate those two. You put them in front of real buyers, including the ones who walked away, and adjust to what you hear. That’s judgement, and judgement is exactly what a death notice tries to take out of your hands. That’s the core of Fuel. Choose.

Staying. Nike is the lesson here. Selling direct doesn’t remove the work, it relocates it and changes who pays for it. Every death notice about channels is really a question about who carries which part of the work, at what cost, per customer and per product.

The work that follows is prosaic and pays back fastest: three years of revenue per customer side by side, so you can see by name and amount who is shrinking and who has gone quiet. The customers taking one product line while comparable customers take three. And a legitimate reason to call at all the moments when there is no problem and no renewal. That’s the core of Lift. Stay.

Four questions for the next proposal

Judging a proposal doesn’t take a day of research. Four questions are enough, and they work on any piece of marketing advice that arrives in the boardroom with a number on a slide.

1. Which study is this number from, what year, and which market does it cover? Three of the four examples above already fell over at this question.
2. What else was measured in that same study that isn’t here? At Gartner, the second number was two press releases further on.
3. Which track does this proposal point at, and does that match where we’re actually stuck? Do we know that from our own numbers, or from somebody’s post?
4. If the answer is both, in what ratio? And what does it cost us if we have that ratio ten percent wrong?

Question four is the only one that forces anyone to do the arithmetic. Which is why it gets skipped most often, and why it’s the only one that produces a decision instead of a cut.

This is where strategic marketing belongs. Not in the question of whether to abolish something, but in the question of which track you’re stuck on, what that costs you per year, and what ratio the budget should sit in against it. And who holds that ratio when the next death notice comes past.

Further reading: Marketing at the board table, on where these trade-offs belong and where they don’t. And Five marketing questions for the board, for the questions that belong on the table when there is no proposal.