Skip to content
Final StateBy the Time Anyone Looked, the Option Had Closed.
VOL. I  ·  NODE 001▢  ATLAS

AFTER CLOSING

By the Time Anyone Looked, the Option Had Closed.

A field guide to building a recognition engine without flooding the owner: scan named sources, carry the deadline and base rate, prepare the move, then hand back the decision.

BY-PRODUCTS

Resources can hide a candidate move. Recognition is only the first gate.

Sequence from an unrecognised candidate move through recognition and securing investment to a real option that can be held or exercised.The figure separates the stages: resources contain a candidate move, recognition identifies it, and investment secures preferential access before a real option exists.CANDIDATE MOVE → REAL OPTION1 · RESOURCECANDIDATE2 · RECOGNISEIDENTIFY MOVE3 · SECUREINVESTPRIORITY ACCESS4 · OPTIONWAIT · EXERCISEOR ABANDONEXPIRY BEFORE RECOGNITIONCANDIDATE CLOSES UNCONSIDEREDRECOGNITION FINDS · INVESTMENT CREATES
  • Bowman and Hurry, 1993: a shadow option is a possible move awaiting recognition
  • Recognition identifies the candidate but does not create a right
  • A securing investment creates the real option before waiting or exercise

In Bowman and Hurry's sequence, pp. 763–765, resources contain candidate moves. Recognition identifies one; a usually small investment then secures preferential access and creates the real option that can be held, exercised, or abandoned. This article's extension is that recognition can be an assignable task.

An option unseen at expiry may already be gone.

A grant deadline, a lease signing, a rival's sale: some rights close on someone else's clock. The same logic hides options inside a menu drawn before you arrived.

CENTRE AND EDGE

Attention pools at the centre. Some value waits at the edge.

Watched center split from an unwatched edge holding options and exposures.The figure contrasts routine attention at the center with edge cases such as lease changes, successor gaps, and key-person exposure.CENTRE AND EDGEWATCHEDUNWATCHEDTILLROTAACCOUNTSDAILY ROUTINELEASESTREET SHIFTRIVAL EXITNO SUCCESSORKEY PERSONSINGLE COPYEASY TO MISSROUTINE SHOUTSTHE EDGE WHISPERS
  • Watched: the till, the rota, the day's accounts
  • Watched: whatever shouts loudest today
  • Unwatched: the street changing around your lease
  • Unwatched: a rival retiring with no successor
  • Unwatched: half the operation living in one head

That last one is a latent exposure: operating logic concentrated in one person, still cheap to protect before it walks out the door.

THE QUADRANT

A triage grid can show where attention goes missing.

Neutral frequency-impact triage grid highlighting a rare, high-consequence attention gap.The grid locates where infrequent consequential events may lack routine attention; it deliberately assigns no automatic response, price, or decision.TRIAGE MAPRAREFREQUENTFREQUENCY →HIGHLOWCONSEQUENCE →RARE · HIGHLITTLE ROUTINEMEMORYFREQUENT · HIGHREPEATEDEXPERIENCERARE · LOWSPARSE NOTICEFREQUENT · LOWROUTINEVISIBILITYATTENTION GAPLOCATES · DOES NOT DECIDE
  • Frequent events leave memory and routines
  • Rare events do not train the operator
  • The grid locates a blind spot; it does not price or decide it

Use the grid here only to mark an attention gap. The separate question of what makes that boundary move belongs to the quadrant's price.

THE CORPUS

Rare for you. Common for the corpus.

Tail bar chart showing the top tenth of innovation cases carrying much of the value.The exhibit visualizes Scherer and Harhoff's highly skewed return samples as a reason to test the search, not a promise that every rare event pays.RETURN SAMPLESTOP 10% SHARE OF TOTAL RETURNSTOP 10%48%OTHER 90%52%LOW ENDTOP 10%93%7%HIGH ENDTOP 10%: 48-93%EIGHT DATASETS · 2000SEARCH EVIDENCE, NOT EVERY ALERT
Source: F. M. Scherer and Dietmar Harhoff, Research Policy 29 (2000), 559-566, doi:10.1016/S0048-7333(99)00089-X.

Rare for you, documented elsewhere can supply a reference class. Truly novel cases cannot, and a fluent model does not make them familiar.

  • Eight datasets on inventions and innovations from firms and universities
  • Scherer and Harhoff, 2000: the top 10% captured 48 to 93% of total returns within the samples
  • Highly skewed returns justify testing the search, not trusting every alert

The blind spot is a price, not a law.

AI can lower the cost of a first look, but false alarms return that cost through the owner's attention. Price the whole watch: scanning, checking, process, and interruption.

THE ENGINE

Build a watchlist, not another dashboard.

  1. 01Scan named sources; return the original record and deadline
  2. 02Triage with comparable cases or a base rate; suppress weak flags
  3. 03Prepare the move, then hand the accountable decision back

A recognition engine earns attention only when every flag carries its evidence, clock, reference class, and named owner.

Recognition engine turning named sources into an illustrative dossier with source, deadline, reference class, base rate, prepared move, and named owner.The figure shows triage suppressing weak signals, preparing a draft or application, and passing only a traceable dossier to the accountable owner.RECOGNITION ENGINENAMED SOURCESPLANNINGREGISTERGRANTPORTALRENEWALDATESTRIAGECHECK · SUPPRESSNO SOURCE · NO ALERTREQUIRED FLAG DOSSIERORIGINAL SOURCEDEADLINEREFERENCE CLASSBASE RATEPREPARED MOVENAMED OWNEREVIDENCE IN · OWNER OUT

AMBIENT LIGHT

Every alert spends trust.

  • Admission rule: original source, clock, and reference class
  • Silence weak flags before they become interruptions
  • Escalate only while a cheap response still exists

Ambient visibility is useful only when the quiet stays quiet. A latent exposure earns interruption when the evidence is traceable and action is still cheap.

WHAT REMAINS

Noticing comes before sizing.

Critical sequence from surfaced position through admission checklist to sizing handoff, with a compact owner-only remainder.The closing diagram lands one sequence: surface the position, require complete evidence, then size downside and upside; novelty, values, and trust remain a compact owner-only exception.NOTICE FIRST · THEN SIZE1 · POSITION SURFACEDSOURCECLOCKMOVEIN TIME2 · ADMIT: COMPLETE OR STOPSOURCETRIGGERDEADLINEREF CLASSBASE RATEOWNER3 · SIZING HANDOFFCAP DOWNSIDE · KEEP UPSIDE OPENPRICE · TIME · CONTINUE TO 002OWNER-ONLY JUDGMENTNOVELTY · VALUES · TRUSTKEEP JUDGMENT HUMAN
  • Operator test: what can expire before the next meeting?
  • For each watch: source, trigger, deadline, base rate, owner
  • Next: size the positions worth holding

First see the position in time. Then ask whether its downside is capped and its upside remains open. Continue to how to size what noticing finds.

Read the transcript

01 · AFTER CLOSING

It is late. The shop is shut. Inside, one light burns over the till and the owner runs the day's takings. Outside, the street goes about its business in the dark. A neighbour packs up a failing trade. A lease two doors down changes hands. A grant window opens and will close. None of it hides. It simply goes unwatched.

02 · BY-PRODUCTS

Existing resources can contain candidate moves nobody has recognised: a lease may support a new use, a relationship may create unusual access, a licence may make an adjacent service possible. In 1993, Edward Bowman and Dileep Hurry called such candidates shadow options, options awaiting recognition. The order matters. Recognition identifies the move, but it does not yet create a right. A usually small securing investment must establish preferential access or capability before the candidate becomes a real option. Only then can the firm wait, learn, exercise, or abandon.

03 · THE CLOCK

Some options carry a hard date. Others close when somebody else acts. A grant window shuts. A lease is signed. A retiring rival sells the book. The clock runs whether you watch it or not. Miss that moment and the right may be gone, as gone as if you had bid in an auction and lost. Except there was no auction. You were never in the room.

04 · CENTRE AND EDGE

Picture an operator running several small, hands-on businesses at once. Keeping them open swallows the day. Attention pools at the centre, on the till and the rota and the accounts, and the edge goes dark. Yet consequential changes often begin there: the street shifting around a leased site, a rival retiring with no heir, a supplier whose emails have cooled, an employee who carries half the operation in his head.

05 · THE QUADRANT

A frequency-impact grid gives us a map, not a mechanism. Frequent events leave memory, practice, and routines. Rare events do not train the operator, so the consequential ones can arrive without rehearsal and land on the owner. Use the grid for one job here: mark the attention gap. Events too infrequent to form a routine, but serious enough to matter. Whether that gap is worth watching is a different question.

06 · THE CORPUS

Now ask whether a machine can see what the owner cannot. Only sometimes. Where the same kind of event is documented elsewhere, the corpus may supply comparable cases. Where the event is genuinely new, there is nothing to pool. The payoff can also be sharply skewed. In 2000, economists Frederic Scherer and Dietmar Harhoff examined eight data sets on inventions and innovations from firms and universities. The top tenth captured between forty-eight and ninety-three percent of total returns within those samples. Their paper does not prove a model can find your opportunity. It shows why a search can be worth testing when a few outcomes carry most of the return.

07 · A PRICE, NOT A LAW

Here is the turn. The blind spot is not a law of nature. It is a bargain struck under limited attention. Artificial intelligence can cut the cost of a first look. But false alarms send the cost straight back through the owner's attention, and checking each flag adds a second bill. So price the whole watch: scanning, verification, process, and interruption. If triage costs more than the signals return, do not build it.

08 · Advertisement · Twain GTM Agents

A good outbound message should know why it is being sent. Twain GTM agents research accounts, qualify contacts, and turn live signals into relevant outreach. They work with your CRM, personas, and GTM stack, so the message is not just personalized. It is backed by a specific reason to start the conversation.

09 · THE ENGINE

Build no dashboard. Give the engine a watchlist: a planning register, a procurement bulletin, a grant portal, renewal dates. It scans those named sources. It suppresses weak flags. What survives comes back with the original record, a deadline, and comparable cases or a base rate. A grant opens; the engine checks eligibility, marks the closing date, and drafts the application. Then it hands the prepared move, and the accountable decision, back to the owner. A fluent alert without a source is not recognition. It is noise dressed as urgency.

10 · AMBIENT LIGHT

Return to the dark field. One lit shopfront at its centre, the owner's attention fixed on the work. The engine does not widen that human beam. It raises the ambient light around it. Named sources become a watchlist. Sourced, time-bounded flags rise into view. The rest stay quiet. That is the promise, and the boundary: make the documented edge legible in time, without pretending the truly new has become routine.

11 · WHAT REMAINS

On Monday, ask one question: what could expire before the next meeting? For each candidate, write down the source, the trigger or deadline, the comparable cases or base rate, and the person who owns the call. If one of those is blank, do not automate the alert yet. The model can surface a documented pattern. It cannot settle the genuinely new decision, the value-laden fork, or the trust a person still has to repair. For this class of option, capital was not the first constraint. Noticing was. Once you have seen the position, the next question is how much of it you can afford to hold.

01 / 11 · AFTER CLOSING0:00 / 6:33