A Client's "Yes" Has a Half-Life

Why the opportunities you worked so hard to create still dissipate — and the system that preserves them

This is Part 4 of a series. Part 1, "Locally Rational, Globally Irrational," explained how intelligent people choose defensible work over the work that changes their trajectory. Part 2, "Median Logic in a Tail-Driven World," explained why the avoided work matters disproportionately — it creates collisions in domains where rare outcomes dominate. Part 3, "Needle Architecture," built the system for producing those collisions on purpose. You can read this one on its own — but it completes something the first three leave open.

Part 3 ended with a single operating rule: create enough qualified collision points for the tail to find you. Make the call. Send the message. Publish the work. Enter the rooms where reality can answer back.

But a collision is not an outcome. It's the start of a new chain. A prospect replies, takes the call, leans in, says "this is really interesting — let's do it." The tail event you were playing for has finally arrived. And then, three weeks later, it's gone. No rejection, no competitor, no dramatic objection. Just silence that hardened into a no.

The first three essays were about how hard it is to create a valuable collision. This one is about a quieter, more painful failure: a collision succeeds, and you let its value leak away afterward. You did the hard part — you got into the collision space, you played past the median, the rare event fired — and then the opportunity dissipated in the ordinary mechanics of what came next.

Creating an opportunity and preserving one are two different systems. The series so far has been entirely about the first. This is the second.

The fourth failure

Across the series, the same underlying mechanic — locally rational choices compounding into globally irrational outcomes — keeps showing up wearing different clothes. By now it has produced four distinct failure modes, and it's worth seeing them in a row:

• Selection failure (Part 1): you choose work with predictable local rewards over the work that changes future possibilities.

• Exposure failure (Part 2): you stay inside the building instead of entering the collision space.

• Persistence failure (Part 2 → 3): you judge each attempt by its median result and stop before the tail can appear.

• Propagation failure (this piece): a valuable collision finally occurs — and its energy fails to survive the chain between interest and outcome.

That fourth one is the subject here. Call it the opportunity propagation problem. A collision creates potential value. Everything that happens after it determines how much of that value survives to completion. And most of it doesn't survive — not because the opportunity was bad, but because nobody understood the system it now had to travel through.

The core error: treating interest as a decision, not a state

Here's the mistake underneath the propagation failure. We talk about a prospect "deciding" to buy, as if interest were a switch that flips and stays flipped. Once they've decided, the deal is in the bag; the rest is logistics.

Interest doesn't work like a switch. Interest is a state, and states decay. A prospect's "yes" is not a stored decision — it's a temporary level of conviction that tends to weaken without reinforcement after the call ends. And it weakens in a specific shape.

This is where the word half-life earns its place, and isn't just a metaphor. A half-life is what you get from exponential decay — a quantity that loses a fixed proportion of itself per unit of time, not a fixed amount. Conviction doesn't bleed off in a straight line. It drops fast at first, then slower, then slower still, flattening as it approaches the floor. A prospect at 80% conviction might sit near 40% a week later, and near 20% the week after that — each interval roughly halving what remained. That curved, front-loaded decline is the signature of exponential decay, and it has three consequences that ordinary "interest fades" never tells you:

First, the early days are worth far more than the later ones. Most of the conviction you lose, you lose in the first stretch after the yes — precisely when most people are still "giving them space." Waiting a week to follow up isn't a small delay; on an exponential curve it can be half your probability.

Second, a deal can look alive and be mostly gone. Conviction flattens near the bottom, so a prospect who's technically "still interested" three weeks later may be sitting at 15% — warm to the touch, structurally dead. The flat tail is what makes cold deals feel recoverable when they aren't.

Third — and this is the variable that explains why generic follow-up advice fails — every deal decays at its own rate. Call that rate λ (lambda): how fast the proportion bleeds off. A hot inbound lead has a steep λ — it halves in days. An enterprise relationship has a shallow λ — it halves over months. The same follow-up cadence that keeps a slow-λ deal warm will lose a fast-λ deal entirely, and the same cadence that smothers a slow deal is barely enough for a fast one. There is no universal follow-up schedule. There is only a schedule matched to a decay rate.

Readers of the earlier essays will recognize the floor this curve falls toward. In "Why Interest So Rarely Becomes Action," I described the cloud — the undifferentiated mass of tasks and half-formed problems that fills a busy person's attention, where most pitches go not to be rejected but to be never evaluated at all. Isolating a problem from that cloud is the first gate. The half-life is what happens next: a problem that briefly isolated itself sinks back into the cloud unless something holds it in the foreground. The cloud has gravity. Interest is the brief escape from it. Decay is the rate of the fall back in.

This is the distinction worth naming precisely, because it separates this essay from the last one. "Why Interest So Rarely Becomes Action" asked: why does recognized value fail to produce a first action? This asks: why does expressed intention fail to survive into a completed one? The earlier essay examined the distance between value and action. This one examines the decay that happens while the buyer is trying to cross that distance.

A deal dies two ways — and they are not two numbers

A deal is a chain of steps: reply, call, terms, signature, payment. To win, every step has to clear. And there are two distinct ways the chain bleeds value, operating at once:

Conversion loss — the deal leaks while moving through a step. Friction in the step itself: the prospect won't take the call, balks at terms, ghosts at the contract.

Momentum loss — the deal cools while waiting between steps. The half-life doing its work in the gaps. Even a step that would have cleared can fail because conviction decayed in the idle time before you got to it.

A note on the math, because this series trades on being careful with it. These two are not two separate numbers you multiply together. If you look at the stage-conversion rates in any real pipeline, the deals that died while waiting are already inside those numbers — the loss is blended. So the right way to hold this is as a diagnostic decomposition, not a formula to compute:

Opportunity survival has two hidden components: step survival (surviving the required actions) and time survival (surviving the elapsed time between them).

Ordinary stage-conversion data usually blends both into one figure. You separate them not to multiply them, but to know which one is hurting you — because they require different interventions. Conversion loss is fixed by reducing friction at the step: a cleaner proposal, a simpler contract, a smaller first commitment. Momentum loss is fixed by reducing or refreshing the wait: faster turnaround, a well-timed touch. They often reinforce each other — a simpler proposal lowers friction and shortens the wait — but if you mistake one for the other, you'll polish your pitch deck while the deal is actually dying in the silence afterward.

Conversion loss is friction at the step. Momentum loss is decay in the gaps.

Why "all my steps look fine" deals still collapse

This is the same compounding I described as derivative chains in "Why Interest So Rarely Becomes Action," now operating on a single live deal — and it fools experienced people every time.

Say your pipeline has five steps, each converting at a healthy 80%. Eighty percent feels strong. But all five must clear in sequence, so the survival probabilities multiply:

0.8 × 0.8 × 0.8 × 0.8 × 0.8 ≈ 0.33

Thirty-three percent. Five individually-healthy steps produce a pipeline that loses two of every three deals. The losses don't announce themselves at one broken step — there usually isn't one. Each step is "fine"; the product is not.

It cuts the other way too, which is the good news. Lifting just three of the five steps from 80% to 88% — leaving the other two untouched — takes the chain from roughly 33% to about 44%. A third more closed deals, from three modest improvements, most of the pipeline unchanged. The murder happens in the multiplication, and so does the rescue.

Not every deal deserves equal rescue

Before the operating rules, one inheritance from Part 2 that changes how you apply all of them. Part 2's whole argument is that outcomes are tail-distributed — a small number of attempts produce most of the value. That logic doesn't switch off once a collision lands. It applies just as hard inside your pipeline: a few live opportunities carry most of the expected value, and the rest, however warm, carry little.

So the question is never simply "is this deal warm or cold?" It's "does the value recoverable from renewed attention justify the effort it costs?" One unusually large opportunity may deserve a level of patience, enablement, and precise timing that ten ordinary ones do not. Spreading equal preservation effort across every cooling deal is its own kind of median logic — treating a tail-distributed pipeline as if it were uniform. The effort you spend holding a deal alive should track the value at stake, not just the temperature on the thermometer.

Hold that underneath everything that follows.

The Closed-Door Rule: don't push a lock

Some pauses are not caused by neglect. They're caused by dependencies.

When a prospect says "I love it — I just need our partner meeting next Tuesday" or "this moves once the new budget opens," the deal isn't cooling from neglect. It's waiting on an external event you don't control. The decision physically cannot advance until that event happens.

This is the same mechanism I called an external trigger in the earlier essay — the deadline, the board meeting, the competitor move that forces a latent problem into the foreground. Here it works in your favor: the trigger is what unlocks the deal, and it arrives on a schedule the prospect usually tells you for free.

In decay terms, a gate does something specific: it pauses the part of the curve you can influence. Before the gate, no follow-up can lift conviction toward a decision, because the decision isn't available to be made — the reset has nothing to grab. The deal can still erode in other ways while it waits (a competitor appears, enthusiasm fades, the details blur), but ordinary closing pressure has zero leverage on a curve that's frozen against you. Rattling a locked door with "have you decided yet?" can't move the curve up; it can only annoy. So separate two kinds of touch:

• Decision follow-up — "are we moving forward?" Premature before the event. The decision can't be made yet; asking just spends goodwill.

• Enablement follow-up — something that prepares the moment. A one-page summary your champion can forward into the partner meeting. An answer to the objection you know the CFO will raise. This can be highly valuable before the gate, because it arms the person who'll be in the room when the door opens.

The Closed-Door Rule: Don't ask a gated deal to move before it can move. Arm the champion if you can, record the gating date, and follow up the moment the door opens.

Most operators don't write the date down, send pointless decision-pressure beforehand, then forget to reach out right when the door actually opens. The discipline is the inverse: no closing pressure before the date, enablement only if it's useful, and first-in-the-inbox after. One precisely-timed touch beats five mistimed ones.

Follow-ups are resets against the curve

Now put the follow-up back on the decay curve, because that's what a follow-up is: a discrete upward jump against a continuous downward slide.

Left alone, conviction follows its exponential fall. A follow-up interrupts it — knocks the curve back up. Then the decay resumes from the new, higher point. Follow up again, another jump, another slide. The picture isn't a smooth line; it's a sawtooth — a falling curve repeatedly caught and lifted. A well-timed sequence of touches keeps the sawtooth riding high; no touches, and it's just the bare decay curve sliding into the cloud.

But the resets themselves decay. The first follow-up jumps the curve up substantially. The second lifts it less. By the fifth or sixth "just checking in," the jump is negligible — and then it goes negative, because an empty touch now signals you need the deal more than they do. The ceiling the resets reach is itself sliding downward. So the sawtooth doesn't climb forever; its peaks descend. There's a point where the next reset lands lower than the curve would have sat anyway, and past that point the touch is worse than silence.

What lifts the ceiling back up is not another touch — it's a touch carrying something new. An empty nudge is the same stimulus repeated, and repeated stimulus habituates. A touch that moves the decision is new input, and new input resets the ceiling. So the test before sending anything is: does this follow-up give them a reason, or just a reminder? A reminder reappears in the inbox. A reason changes the buyer's position. There are five things a real reason can do:

• Raise confidence — a relevant proof point, a comparable result.

• Lower effort — a smaller pilot, a simpler path, less to commit to.

• Create urgency — a real deadline or constraint, not a manufactured one.

• Resolve uncertainty — answer the specific question or objection actually blocking them.

• Land at a newly relevant moment — right after their external trigger, or after something changed on their side.

Reminders habituate fast; the ceiling drops under them. Reasons don't; they reset it. The job isn't to follow up more, or less. It's to not follow up empty.

When pursuit stops paying: from active pursuit to passive presence

You've followed up well, and the deal still hasn't closed — but it hasn't died either. It's gone quiet in a way more chasing won't fix. You've reached the point where another personal follow-up is more likely to irritate than advance.

The instinct is binary: keep nagging, or write it off. Both are wrong. The right move is to shift the deal from active pursuit to passive presence.

• Active pursuit is high-effort, personalized, one-to-one communication aimed at moving a current decision. It's expensive, and it's where you've been operating.

• Passive presence is low-effort, scalable, one-to-many communication aimed at staying in memory until the prospect's circumstances change. A newsletter is the obvious instance, but it's anything that keeps you quietly visible at low marginal cost — periodic useful updates, the occasional relevant case study, a light quarterly check-in.

In decay terms, this is a deliberate change of regime. Active pursuit is fighting a steep λ with expensive resets. Passive presence accepts that the conviction curve has flattened near the floor — that there's no high ceiling left to reset toward — and switches to the only move that still pays: keeping a faint, cheap signal alive so the deal doesn't fully vanish into the cloud. You're no longer trying to lift the curve. You're keeping it from hitting zero, at near-zero cost, while you wait for something on their side to redraw it.

You stop spending scarce personal attention on a deal that isn't ready, without losing it. And here's why that matters, and why it connects straight back to Part 2: a meaningful share of deals close for reasons that have nothing to do with you. A budget frees up. An incumbent fumbles. A mandate lands from above. These are collisions too — external triggers firing on the buyer's side — and you can't predict or cause them. But when one fires, you want to be the name already in the room. Passive presence is how you hold that position across dozens of cold deals at once, without giving each individual attention, until one reheats.

The signal to pull a deal back into active pursuit isn't a calendar reminder — it's fresh evidence something changed: they click, they reply, they revisit your pricing, their situation visibly shifts. That's the door reopening, and the deal is worth your personal attention again, now with a high ceiling because there's genuine new relevance.

And a deal doesn't cycle forever. Some genuinely end — to Won, to Lost (they chose someone else, the need vanished), or Disqualified (the economics never made sense, they were never a fit). A real system marks those exits cleanly instead of nursing dead deals in passive presence out of optimism. Passive presence is for deals that are cold but alive — not deals that are done.

This is also what makes a high-volume pipeline survivable for a small team or a solo operator. You cannot personally pursue a hundred cold deals. But you can keep all hundred in passive presence without giving each individual attention, and let that channel watch for the handful that reheat — turning an unmanageable chase into a manageable system.

The whole arc

Step back, and the four essays form one continuous argument about how rare opportunities are made and lost:

First, choose the work that can actually change the future — not the defensible work that merely feels productive.

Then create enough contact with reality for the tail to find you — because the outcomes that matter are rare, unpredictable, and only available in the collision space.

Stay in the game long enough for the tail to appear — because the median attempt disappoints, and judging by the median makes you quit one attempt too early.

And when reality finally answers — when a collision lands and a prospect says yes — do not mistake the collision for the outcome. The opportunity still has to survive the chain. It can leak at any step, cool in any gap, and sink back into the cloud it briefly escaped.

Rare opportunities are hard enough to create. The final failure is letting them dissipate after they arrive.

Your "sure thing" deals were never dying of bad luck. They were decaying on a curve you couldn't see. Now you can.