---
name: proximity-before-volume
description: "Load when the user hands over gathered competitor or market signals plus their own position (what they sell, to whom, their bet) and wants a watch brief for a strategy, product or executive reader. Returns a top-line read naming the lead signal and the louder one demoted, framed core signals with implication tags (thesis relationships only when a thesis was given), adjacent items, open questions with a resolution path, and a synthesis. Not for battlecards, feature matrices, or gathering signals."
---

# Reading a Signal Batch

Someone is going to read this and then act. That is the only test it has to pass — not completeness, not coverage, but whether the person holding it knows what to do differently by the end.

Two rules govern the whole procedure. **The supplied signals are the entire evidence** — what you know about these companies helps you interpret wording and never adds a fact, a motive or an event. And **the brief's shape follows the inputs**: sections exist when the inputs support them and are absent when they do not. A section filled with plausible-sounding content the inputs cannot carry is worse than no section.

## Which brief is this

If no position is in the message, ask for it in one line before writing. If the user declines, proceed as below.

Before writing anything, answer two questions about the inputs and let the answers bind you.

**Is a reader position supplied?** If no company, product, buyer or strategic bet is stated, there is no vantage point from which signals mean anything for anyone. The brief then carries signals with their implications and an open questions log, and the synthesis section does not appear. Say in one line that synthesis is withheld because no position was supplied.

**Is a prior thesis supplied?** A thesis is an explicit claim the reader currently believes about their market or their advantage. This question has two binding answers and no middle.

If a thesis **is** stated, every core signal carries an explicit relationship to it — confirms, complicates or contradicts — named in the entry itself. Leaving a core signal untagged when a thesis exists wastes the most decision-relevant judgment in the brief.

If no thesis is stated, do not tag against one: no confirms, complicates or contradicts, and no empty tag field.

Neither absence is a defect in the inputs. Each is a legitimate state that changes what the brief contains, and recognising it unprompted is part of the work.

## Lead times and what they cost to fake

Different moves carry different lead times and different reliability, and a brief that treats them alike wastes the reader's attention.

Hiring and capacity commitments lead product by a long way — a role opened is a decision already taken and usually a direction visible months before anything ships. Read the function, the seniority and the segment named in the posting rather than the count. Certification, compliance work and infrastructure build-out are the same kind of evidence: slow, expensive, and therefore honest about intent.

Announcements, roadmaps and blog posts are free, and their information is about positioning rather than capability. What is announced may not exist; what exists may not be announced. A shipped feature with customers on it and a named roadmap item are not the same evidence and should not sit in the same tier.

Pricing and packaging changes are the fastest-moving signal and the most often misread. A repackaging tells you who a company has decided to sell to next — the buyer they are optimising for — which frequently contradicts what their marketing still says. Read a tier rename as a statement about target buyer, not about product.

Departures, reorganisations and quiet deprecations are negative signals and are usually under-weighted because nothing was announced. A capability going unmaintained tells you where a competitor has stopped investing, which is often more useful than where they started.

## Start from the bet, not the batch

Read the reader's position first, and extract the thing that decides everything after it: what they sell, to whom, and what they are betting on. That bet is their exposure. A signal matters in proportion to how directly it touches it.

Do this first because reading signals cold invites ranking them by size — the largest raise, the loudest launch — and size is not proximity. A modest hire into the exact capability the reader claims as their moat outranks a large raise by a company selling to a different buyer.

## Three questions before any label

Before tiering anything, say what the actor did. Ask three questions of every signal, in this order, and let the answers rather than any keyword decide the classification.

**What changed, concretely?** Distinguish a change in how something is sold — packaging, pricing, channel, target buyer, positioning — from a change in what is being built. Both may appear in a single announcement; separate them. A page edit that renames tiers is a go-to-market move even when the product it prices is unchanged.

**What did it cost them?** Costly moves reveal intent more reliably than free ones. Hiring, certification, acquisition and infrastructure commit money and time; announcements, roadmap statements and blog posts commit neither. A capacity signal — who is being hired, into what function — often reveals a direction six months before any product does, which is why quiet signals frequently outrank loud ones.

**Who is speaking, and how do they know?** A company describing itself is evidence of what it wants believed. A customer, regulator or third party describing it is evidence of what happened. A single unattributed report is neither until corroborated.

Novel signal types will arrive that fit no familiar category — a regulator acting, a partnership forming, a standards body moving, a signal that is two things at once. Do not force these into the nearest label. Run the same three questions: what changed, what it cost, who says so. Then state plainly what kind of move it is in your own words and why it matters here. A signal described accurately in unfamiliar terms is worth more than one filed neatly under the wrong heading.

## When the actor is not a competitor

Much of what arrives is not one competitor attacking another, and reading it that way produces the wrong brief.

When two companies partner, integrate or embed one another, ask whose distribution changed and whose product did not. A platform folding a capability into its own surface changes where buyers encounter that capability — it is a distribution and default-choice signal, not a head-to-head product threat, and its danger is that the reader's category quietly becomes a feature of something else. A supplier or infrastructure partnership usually reveals what a competitor has decided not to build themselves, which tells you where their engineering is going instead.

When a regulator, standards body or large buyer acts, the move is rarely aimed at any one vendor, and the question is who it advantages. A requirement that is expensive to meet favours whoever already meets it and raises the wall against everyone else. Ask whether the reader is on the near or far side of that wall, and whether their claimed advantage just became table stakes for the whole market.

Classifying such a signal correctly is only half the work, and the half that is worth nothing on its own. Finish by naming what it does to **this reader's** stated advantage: whether the capability they sell is becoming a default feature of someone else's platform, whether the channel they rely on now has an incumbent sitting in front of it, or whether the thing they charge for is on its way to being free infrastructure. A brief that identifies an infrastructure move and stops has told the reader a fact; one that says which part of their moat it dissolves has told them something they can act on.

Some signals are two things at once — a partnership that is also a product launch, a hire that is also a market entry. Say so and carry both readings rather than picking the tidier one.

## Nearness beats loudness

Analyst or category-level commentary about the market as a whole is adjacent at most: it is not a competitor's move. It can inform the synthesis but never takes a core row.

Sort signals into those close to the reader's bet, those in their market but not touching it, and those that are context only.

The test is not how big the news is. It is: if this continues, what does the reader lose, and how soon? A competitor building the capability the reader calls their moat is core however undramatic the evidence. A large raise by a company selling to a different buyer is adjacent. Genuine background is noise, and saying so is useful — it tells the reader what they can stop tracking.

Some cycles contain no core signal, and the core section is then **empty**. Leave it empty, say plainly that this is a quiet cycle, and let the top line say what remains worth watching. Promoting the least-weak signal to core to avoid an empty section is the single most common way a brief misleads: it tells the reader something crossed the threshold when nothing did. Where the evidence this cycle is uniformly thin or unverified, every signal belongs in open questions and core stays empty — that is a finding, not a failure to analyse.

## Where a stated claim stands after this

Where the reader states a prior thesis, each core signal stands in one of three relationships to it, and the choice is evidential, not diplomatic.

**Confirms** — the signal is what the thesis predicts. **Complicates** — the thesis still holds but a condition it relied on is weaker, narrower or more contested than assumed. **Contradicts** — the signal is evidence the thesis is wrong about something it asserts.

Decide between the middle and the last by a mechanical test rather than by feel. State what the thesis asserts as a plain proposition. If a signal shows that proposition is false — the rival doing what the thesis says they do not, entering the segment it says they avoid, committing now where it placed them years away — that is **contradicts**, however unwelcome. Reserve **complicates** for a proposition that survives intact but narrower, later or more conditional than assumed. When you find yourself reaching for complicates, check whether the thesis actually survives the signal or whether you are protecting the reader from a conclusion they need. A reader who learns this from someone else first will not trust the next brief.

Where the reader supplies no thesis, this step does not exist. Do not invent a thesis, do not tag against an implied one, and do not leave the labels in place unfilled. The absence is a legitimate input state, not a gap to paper over: the brief simply reports implications without thesis relationships.

## Items that stay questions

A signal is not resolvable when it rests on one unverified report, when its source is anonymous or interested and uncorroborated, when it admits two readings that lead to different decisions, or when it is a single early data point that may be the start of something or noise.

These go to open questions, not into the core section with a confident implication attached, and not into vague hedging that leaves the reader nothing. Write each as a question that could actually be answered, and name what would answer it — a filing, a job posting, a customer conversation, next quarter's disclosure, a specific person to ask. A question with no resolution path is an anxiety, not an open question.

Everything else follows from this: hedging language belongs to claims the evidence genuinely underdetermines, and confident language belongs to claims it supports. Applying either uniformly destroys the brief's usefulness.

## The pattern none of them shows alone

Synthesis is the part a reader cannot do themselves, and it is not a summary. Restating each signal in order is not synthesis. The question is what several signals **collectively** imply that none implies alone — a pattern, a convergence, a shift in where the market's attention is moving.

Name the signals the claim rests on so the reader can check it. Stay inside them: if the pattern requires an assumption the signals do not supply, say that the pattern is provisional and what would confirm it. Then end where the reader needs you to — the decision this forces, the thing to do now, or the question to resolve before the next cycle. Synthesis that stops at insight leaves the work unfinished.

Where the inputs supply no reader position, there is no vantage point from which anything means something, and this section is omitted. Report the signals with their implications and the open questions, and say plainly that synthesis is withheld for want of a stated position. Inventing a position to write toward is the worst available move: it produces confident advice aimed at a reader who does not exist.

## Putting it in front of the reader

Open with a single line naming the reader and the cycle, then the read itself.

The read is where most of the brief's value has to land, because it is what gets read. Say which signal is the lead **and which apparently bigger one is not** — naming the thing you demoted is what shows the judgment. Then argue the two or three calls that matter most, including their thesis relationships where a thesis exists, right there in the read rather than deferring them to entries below. Someone who stops after the read should already know what changed, what it means, and what they must decide.

Write the entries as analysis, never as a form. Do not print your reasoning steps as labelled fields — no "What changed:", no "Source:", no "Cost:", no "Implication:". Those are the questions you asked yourself, not the answer the reader wants, and they consume the space judgment should occupy. Each entry is a short paragraph that names the change, attributes it, and says what it does to this reader, with its tag stated inline in the same breath.

If nothing crossed the bar, say that plainly in the read instead. Then the core signals — empty if nothing qualifies — each with what changed, its source, its implication for this reader, and, only where a thesis was actually supplied, its relationship to that thesis. Then the adjacent ones more briefly, then what the evidence cannot settle, then anything worth noting but not tracking.

Finish every section the inputs call for. Finish every section the inputs call for; budget before you write rather than trimming after: two or three sentences per core signal, one per adjacent one, a line each for noise. Depth belongs to the core signals and the synthesis; everything else is there to be scanned. Precision earns more than length — every sentence should carry a fact, an implication or a decision. Cut process narration, restatements of the assignment, and caveats that apply to everything equally.

## One cycle, written out

Invented from end to end. The point is the density of the opening, not the layout.

> **HOLLOWAY — COMPETITIVE WATCH BRIEF | 9 June 2027**
> *Six items in; three framed, two questions, one set aside. Items 2 and 4 are the same development.*
>
> **THE READ**
>
> Holloway sells ticketing and membership software to independent museums, on the strength of handling gift-aid claims better than anyone. The item that touches that is four lines in a release note: Vantage, the card terminal supplier most of these museums already use, now files gift-aid claims itself, included (item 1). Vantage is not a competitor and has no view about Holloway — which is the point. The capability Holloway charges for is now bundled beneath it, and every rival on Vantage terminals gets it too, so the moat clause in the standing thesis does not survive this intact. Alongside that, Thorne has withdrawn its single-site plan (item 2) and three single-site museums have written to Holloway this fortnight asking what switching involves (item 4); those are one event seen from outside and inside, and the enquiries are what date it. Meanwhile Cassin's £22m raise (item 5) is the largest number here and the least relevant — the stated use is local-authority estates, a buyer Holloway does not sell to, so it sits adjacent. An unattributed post claiming Vantage will charge for claims from next year (item 6) is held below; if it were true it would reverse most of the above, which is exactly why nothing above rests on it. The decision this forces is whether the next quarter is spent moving the pitch from filing claims to reconciling and auditing them — the part Vantage has not touched — while Thorne's single-site museums are still in motion.
>
> **WHAT IT MEANS FOR HOLLOWAY**
>
> Items 1 and 2 together say the ground has moved twice in one month: what Holloway sold is now included elsewhere, and the neighbouring segment is loose at the same moment. Those pull in opposite directions — one erodes the reason to buy, the other creates a short queue of buyers — and the queue is the perishable one.
>
> **FRAMED**
>
> **Item 1 · Vantage files gift-aid claims · dependency exposure · contradicts the moat clause.** Four lines in a release note, live now, free. Reconciliation and audit trails remain above the line; the filing itself no longer distinguishes anything.
>
> **Items 2 + 4 · Thorne withdraws single-site plan · go-to-market · complicates.** Withdrawal alone is ambiguous. Three enquiries in a fortnight make it live and dated. Whether Thorne is repricing or leaving is unresolved and changes how long the queue lasts.
>
> **ADJACENT**
>
> Cassin's £22m — local-authority estates, not Holloway's buyer. Revisit if the stated use changes.
>
> **QUESTIONS**
>
> Will Vantage charge for claim filing? One anonymous post. Ask the two museums on Vantage's beta what their terms say, and check the terminal pricing page at quarter end. A yes restores the moat; a no makes the pitch change urgent.
>
> Is Thorne repricing or exiting? Ask the three enquirers what they were told. A repricing closes the window within weeks.

## The thin cycle, miniature

Also invented. This is what an honest brief looks like when almost nothing clears the bar.

> **LEDGERWAY — COMPETITIVE WATCH BRIEF | Week of 2027-05-02**
> *Position on file, no prior thesis supplied; five signals, none framed; this is a quiet cycle and the questions are the product.*
>
> **TOP-LINE READ**
>
> Nothing this cycle clears the bar for a framed signal, and that is the finding rather than a gap in the work. The two items that would matter most if true — a claim that Fenwick is pulling out of single-practice conveyancing, and a report of Aldermoor cutting its per-matter price — each rest on one unverified account, and both readings of either would point Ledgerway in opposite directions. Nothing here should change what Ledgerway does this month. What would change it is named below, with the check that would settle it, and two of the four can be resolved inside a fortnight without waiting for the next cycle.
>
> **CORE SIGNALS**
>
> *[Empty. No signal met the bar this cycle — see the questions below.]*
>
> **WHAT THIS MEANS FOR US**
>
> *[Withheld. A cross-signal reading built on one forum post and one unattributed newsletter line would assert more than this batch supports.]*
>
> **OPEN QUESTIONS**
>
> Is Fenwick leaving single-practice conveyancing? One forum post from an account claiming to be a former employee, uncorroborated. Their pricing page still lists the single-practice tier — check whether it survives the quarter, and ask the two Fenwick customers who spoke to us in March what their renewal terms look like. An exit opens a segment Ledgerway already serves; a repricing closes it.
>
> Has Aldermoor cut per-matter pricing? A newsletter line with no source and no figure. Their public pricing is unchanged as of this week. Worth one look at the page each Friday for a month; if it moves, it resets what Ledgerway can charge new practices.
>
> Did the Law Society consultation on matter-file retention close with a fixed retention period? The consultation is real and cited, the outcome is not published. The published outcome decides whether Ledgerway's archive feature becomes a requirement or stays a preference — check the Society's publications page after the stated close date.
>
> Who is behind the three anonymous job listings for conveyancing product roles? The listings are real, the employer is masked. If it is Fenwick they are not leaving; if it is a new entrant it is the first sign of one. A recruiter contact can usually identify a masked listing in a week.
>
> **ADJACENT / NOTED**
>
> Cranmore's office move — no bearing on product, buyer or price; noted so it is not raised again next cycle.

## Last pass

Every signal in the input is accounted for somewhere. No signal, company, event or figure appears that the input did not supply. No reader position or thesis has been assumed. Each implication names the signal it rests on. Ambiguous items are questions with resolution paths, not softened assertions. The thesis tags are the ones the evidence supports, including the uncomfortable ones. Sections absent from the brief are absent because the inputs did not support them, and the brief says which and why. If no position was supplied, confirm no synthesis survived.