What business protocols are
A business protocol is a repeatable pattern of behaviour for a recurring business situation — a negotiation, a hiring conversation, a client escalation, a board report, a vendor review, a crisis communication. Unlike an SOP, which governs a technical or operational procedure, a business protocol governs a human interaction: what to say, in what order, to whom, and what constitutes a successful outcome.
Distinction
SOP — governs how a task is executed. Steps are deterministic; correct execution is verifiable by outcome metrics.
Business Protocol — governs how an interaction is conducted. Steps involve judgement; success is assessed against relationship and strategic outcomes, not just immediate task completion.
Every high-performing organisation has business protocols. Most of them exist only as tribal knowledge — in the heads of senior people, transmitted by observation and informal coaching, lost when those people leave. A founder knows intuitively how to handle a customer complaint that threatens to go public. A veteran sales director knows exactly when to bring in the CEO and when to hold the line alone. A CFO knows what level of detail the board actually wants in a financial report versus what they say they want.
This knowledge is enormously valuable and almost never written down. CLN is the format that makes it writable — without killing the nuance, the branching logic, and the embedded risk awareness that make it useful.
Why protocols stay tribal — the documentation trap
The reason business protocols are rarely documented is that the formats available for documentation are all wrong for the content.
Prose buries the decision logic in narrative. You can read a two-page description of "how we handle escalated client complaints" and still not know whether to involve legal counsel before or after the first call with the client. The decision point is in there somewhere — but it is dressed as context, not flagged as a branch.
Bullet lists flatten conditional logic. "Step 4: assess severity" tells you nothing about what to do with the assessment. The branch — "if critical, escalate to CEO; if moderate, handle at account manager level; if cosmetic, log and schedule" — is implicit at best, absent at worst.
Flowcharts capture branching but obscure risk, ownership, and rationale. A diamond box that says "LEGAL REVIEW NEEDED?" contains none of the information needed to answer the question.
Most organisations don't lack business knowledge. They lack a notation capable of holding it — one that captures steps, branches, risks, ownership, and rationale without requiring a separate document for each.
— CLN positioning, business context
What CLN adds to business protocols
CLN's operator set was designed for exactly this kind of content. Every element that makes a business protocol useful — and currently undocumentable — has a direct operator mapping:
| Element | Op | Business meaning |
| Goal / purpose |
= |
What success looks like for this interaction. Not the task — the strategic outcome. |
| Action |
. |
Concrete behaviour: say this, send this, arrange this. One action per line. Verb-first. |
| Risk / avoid |
! |
What can damage the relationship, the deal, or the company's position if handled wrong. |
| Hard prohibition |
✗ |
Absolute constraints — legal, ethical, regulatory, or reputational. No exceptions. |
| Expected outcome |
→ |
What the other party's response should look like if this step was handled correctly. |
| Decision point |
? |
Where the protocol branches based on the other party's behaviour or an internal condition. |
| Ownership / timing |
@ |
Who does this step, by when, in which channel, with what authority level. |
| Resource / leverage |
| |
What you have available: data, relationships, budget authority, product capability. |
| Context / rationale |
* |
Why this step matters. What a senior person would tell a junior after the fact. |
| Tension / tradeoff |
~ |
Where two valid objectives pull in opposite directions and the protocol cannot resolve it universally. |
| Advantage |
+ |
What your position, offer, or approach has working in its favour. |
| Disadvantage |
− |
What is working against you — and must be addressed, not avoided. |
Protocol types and where CLN fits each
🤝
Negotiation Protocol
Deal structure, concession sequencing, BATNA management, walk-away triggers, closing moves.
Key operators: = objective · ~ tradeoffs · ! red lines · ? branches
🔺
Escalation Protocol
Who gets pulled in at what threshold, in what sequence, with what framing. Prevents both under- and over-escalation.
Key operators: @ owner · ! thresholds · → expected resolution
💼
Client Communication Protocol
Tone, channel, frequency, escalation path, and recovery playbook for every stage of the client relationship.
Key operators: . actions · ! never-say · → confirmation signal
📋
Hiring Protocol
Scorecard criteria, structured interview questions, debrief format, offer sequencing, and rejection handling.
Key operators: = criteria · +/− signal weighting · ! disqualifiers
📊
Board / Investor Reporting Protocol
What to lead with, what to bury, how to frame bad news, what questions to pre-empt, what to never present without context.
Key operators: @ audience · ! landmines · → outcome desired
🚨
Crisis Communication Protocol
Holding statement, spokesperson rules, media response sequencing, internal vs external message delta management.
Key operators: ✗ prohibitions · @ spokesperson · . timed actions
🤝
Vendor / Partnership Protocol
Qualification criteria, commercial terms, SLA negotiation, renewal triggers, and exit conditions.
Key operators: | leverage · ~ dependency tensions · ! lock-in risks
🏁
Onboarding Protocol
What a new employee, client, or partner must understand in what sequence to become independently effective.
Key operators: . milestones · → readiness signals · ? assessment gates
Negotiation protocol — full example
Negotiation is the business protocol where tribal knowledge is most expensive to lose and hardest to transmit. A junior account executive watching a senior deal with a procurement team learns by osmosis — which concessions came first, which were held back, when the tone shifted, when silence was used as a tool. CLN can encode this.
PROTOCOL-DEAL-001 · Enterprise Contract Negotiation
NEGOTIATION: PROTOCOL-DEAL-001 — Enterprise Contract
@ owner: account executive + VP Sales (deals > $200K)
@ trigger: prospect signals intent to buy · procurement engaged
@ success: signed contract within approved commercial terms · relationship intact
= goal: close at list price or within pre-approved discount band
= non-goal: win the argument · out-negotiate the buyer's team
* they will work with us for 3+ years. they need to feel this was fair.
| leverage: reference customers in their sector · technical differentiation · timeline pressure
| authority: AE up to 10% · VP up to 20% · CEO required above 20%
! never lead with discount — it signals room exists before they've asked !!
! do not negotiate against yourself — wait for their counter before moving !!
⟶ Phase 1 · Anchor
. present full commercial proposal — no pre-emptive discount
* framing: "here is what we've designed for your use case"
→ they counter, ask for discount, or raise procurement objections
? what is their stated objection?
→ price too high: proceed to §price
→ scope uncertainty: proceed to §scope
→ legal/compliance: proceed to §legal
// ##price Price negotiation
. acknowledge the ask — do not concede immediately
* say: "let me understand what you're trying to solve" — buy time, gather intel
. trade concession for concession — never give something for nothing
* examples: faster payment terms, longer commitment, expanded scope
~ give on price to close fast vs hold to protect margin for renewal baseline
* the price you set today is the floor they'll start from at renewal
! if they name a number below your floor: pause · do not respond immediately !!
? are they within approved discount band?
→ yes: AE can close · confirm VP is aware
→ no: escalate to VP before responding · do not promise what you can't deliver
// ##scope Scope objections
. clarify: which capability is uncertain for them?
. offer: scoped pilot or phased rollout to reduce their risk
→ scope is defined · risk is owned · price re-anchored to reduced scope
! do not expand scope without repricing — scope creep starts here !!
// ##legal Legal & compliance review
. introduce legal to legal — do not proxy legal questions through AE
* "let me connect you with our counsel directly"
✗ AE never makes legal commitments — refer all to contracts team
. set timeline: "our standard turnaround on redlines is 5 business days"
→ legal aligned · commercial terms unaffected
⟶ Phase 2 · Close
. summarise agreed terms in writing before DocuSign
* prevents last-minute re-opens; creates shared record
. set signature deadline — create mild urgency without pressure
* "we can hold these terms until [date]"
→ signed contract · kickoff scheduled · champion thanked
@ post-close: AE logs deal terms · flags any non-standard clauses to CS team
Client escalation — the hardest protocol to get right
Client escalations are the protocol most organisations handle the worst — and the one where getting it wrong is most visible. The failure modes are symmetric: under-escalate and the client feels unheard; over-escalate and you signal the account is unstable, burning senior resources on manageable problems.
✕ Typical escalation handling (tribal)
CSM emails account manager:
"Hey, Acme is unhappy about the
integration delay. What should I do?"
Account manager forwards to VP:
"FYI, Acme issue."
VP escalates to CEO before anyone
has assessed severity or proposed
a resolution path.
Acme receives three overlapping
calls from different seniority levels
with contradictory messages.
✓ CLN escalation protocol
@ owner: CSM
. assess: severity tier ☐
? tier?
→ P1 (churn risk): §exec
→ P2 (satisfaction): §senior
→ P3 (friction): CSM handles
// ##exec Executive involvement
. brief VP before contacting client ☐
. align on message — one voice ☐
! client hears from ONE person !!
PROTOCOL-CS-003 · Client Escalation
ESCALATION: PROTOCOL-CS-003 — Client Issue Response
@ owner: Customer Success Manager (first responder)
@ trigger: client raises issue · or CSAT drops below threshold
= goal: resolve the issue · protect the relationship · prevent churn
= non-goal: win the blame argument · prove the client is wrong
⟶ Step 1 · Assess severity (before any external communication)
. classify against severity matrix ☐
? severity?
→ P1 · churn risk or reputational: proceed to §exec-track
→ P2 · satisfaction risk, recoverable: proceed to §senior-track
→ P3 · friction, low business impact: CSM owns · resolve in 5 days
! never tell client "I'll escalate this" before you have a resolution path !!
* "escalation" signals systemic failure · it raises their anxiety without a payoff
// ##exec-track P1 Executive Track
. brief VP/CRO BEFORE contacting client — 15 min call ☐
. agree: single spokesperson · key message · concrete commitment
. VP/CRO contacts client executive-to-executive within 2 hours ☐
* tone: "I'm personally taking ownership of this"
! do not bring engineering on this call unless client insists !!
✗ VP and CSM do not both call the client — one voice, always
. within 24h: written resolution plan with owner names and dates ☐
→ client confirms they have a plan · tone shifts from complaint to monitoring
// ##senior-track P2 Senior Track
. CSM proposes resolution path before calling ☐
* "I've already spoken to [team] and here's what we're doing" > "I'm looking into it"
. CSM holds the call — account manager joins if client requests ☐
. follow up in writing within 4 hours of call ☐
→ client feels heard · has something in writing · issue tracked
⟶ Closure — all tracks
. post-resolution check-in: 48–72 hours after fix ☐
* do not wait for them to confirm — proactively close the loop
. log: root cause · response actions · what to change in product/process ☐
? did the root cause reveal a product gap?
→ yes: open product ticket · link to escalation record · notify client of roadmap intent
→ no: close · update account health score
Hiring protocol — making the invisible scorecard explicit
Most hiring decisions are made by pattern-matching against an unspoken mental model. The experienced interviewer "knows" when a candidate is right — but cannot articulate the criteria precisely enough to train the next interviewer. CLN encodes the model.
PROTOCOL-HR-002 · Senior Engineer Hiring
HIRING: PROTOCOL-HR-002 — Senior Engineer
@ owner: hiring manager (loop design) · each interviewer (their domain)
@ trigger: JD approved · sourcing active
= goal: identify engineers who are both technically excellent AND will thrive here
= non-goal: the most impressive CV · the most impressive whiteboard
⟶ Screening · Resume + async assessment
. apply must-have filter first — saves everyone time ☐
! must-haves are binary: present or absent · do not "compensate" !!
+ signals: shipped something used by real people · wrote about it · open-source contributions
− anti-signals: CV optimised for keywords · 6-month average tenure without explanation
? passes screening?
→ yes: async technical problem (2 hours) · evaluate before phone screen
→ no: decline within 3 business days · personalised one sentence
⟶ Loop design
= each interviewer owns ONE dimension — no redundant coverage
@ interview 1: technical depth · interviewer: senior IC
@ interview 2: system design / architecture · interviewer: staff engineer
@ interview 3: collaboration / communication · interviewer: cross-functional partner
@ interview 4: values + growth · interviewer: hiring manager
! do NOT ask the same question twice across the loop — it signals disorganisation !!
⟶ Debrief protocol
. each interviewer submits written scorecard BEFORE debrief ☐
! no verbal sharing before scorecards submitted — anchoring bias !!
. debrief: strongest voice presents first, then opens for dissent ☐
~ "culture fit" framing vs evidence-based behavioural assessment
* "culture fit" is a bias amplifier · ask for specific behavioural evidence instead
? unanimous strong hire?
→ yes: move to offer within 24h · speed signals respect
→ split: document specific disagreement · hiring manager decides with reasoning on record
→ no hire: decline within 48h · always offer call for feedback if senior candidate
⟶ Offer & close
. offer conversation: hiring manager calls — not recruiter ☐
* "I want to personally tell you we want you on the team"
! never lowball — it poisons the well before day 1 !!
✗ do not ask "what are you currently making" in any jurisdiction where prohibited
. give 5 business days to decide · one extension if requested ☐
→ signed offer letter · start date confirmed · welcome message sent same day
Board reporting protocol — what boards actually want
Board reports are among the most consequential recurring business communications and among the most poorly designed. Most boards receive decks that are too long, lead with metrics the board already knows, bury the decisions that require their input, and present problems without proposed resolutions.
PROTOCOL-EXEC-005 · Board Meeting Report
REPORTING: PROTOCOL-EXEC-005 — Board Meeting
@ owner: CEO (overall) · CFO (financials) · functional leads (their sections)
@ trigger: 10 days before board meeting · materials due 5 days before
= goal: board understands company state · makes required decisions · leaves aligned
= non-goal: impress the board · demonstrate busyness · avoid hard topics
! never present a problem without a proposed resolution path !!
* "here's a problem, what should we do?" is an abdication, not a governance item
! never surprise a board member in the meeting — pre-read all landmines 1:1 beforehand !!
✗ do not present vanity metrics — engagement, followers, impressions — without business context
⟶ Pre-meeting (5–7 days before)
. CEO: 30-min 1:1 with each board member — surface concerns before the room ☐
* ask: "is there anything you want to make sure we cover or discuss?"
. materials distributed with 5-day lead — not the night before ☐
→ board arrives having read the deck · meeting is discussion, not presentation
⟶ Report structure (maximum 15 slides / equivalent)
. slide 1: one-page state of the company — 4 metrics, 2 sentences each ☐
* green/yellow/red for each · board can assess health in 90 seconds
. slides 2–4: progress on last board's commitments — what we said, what happened ☐
! every missed commitment needs an explanation · silence is worse !!
. slides 5–8: decisions required from this board ☐
* format: context (1 slide) · options (pros/cons) · recommendation · ask
~ present your recommendation vs leave it open for discussion
* boards prefer a recommended option to debate from — easier to refine than to generate
. slides 9–12: functional deep dive (one function per meeting, rotating) ☐
. slides 13–15: financials · 13-week cash · variance to plan ☐
! CEO must understand every number in the financial section — no CFO-only slides !!
⟶ In the meeting
. CEO opens: "here are the 3 things I need from this board today" ☐
. present slides 1–4 in 20 minutes — board can read; meeting is for dialogue ☐
. reserve 40% of meeting time for unstructured board questions ☐
! do not fill every minute with slides — silence and questions are valuable !!
. close: read back decisions made · confirm owners · set next meeting date ☐
→ every board member leaves knowing what was decided and by whom
⟶ Post-meeting (within 48 hours)
. send written summary: decisions + owners + timelines ☐
. update commitment tracker — becomes the opening of next meeting ☐
Crisis communication — when the protocol matters most
Crisis communication protocols fail most often because the people executing them are under maximum stress, minimum time, and conflicting pressures from multiple stakeholders. CLN's design — cognitive load at write time, near-zero load at execution time — is precisely optimised for this condition.
PROTOCOL-COMMS-001 · Crisis Communication (first 4 hours)
CRISIS: PROTOCOL-COMMS-001 — External Crisis Response
@ owner: CEO or designated spokesperson
@ trigger: incident with external visibility · regulatory contact · media inquiry
@ decision threshold: any situation where public silence > 4 hours is damaging
✗ no external statement without CEO + Legal approval — zero exceptions
✗ do not speculate on cause in public until root cause confirmed
✗ do not say "no comment" — it implies guilt · use holding statement instead
✗ do not attack, minimise, or deflect — acknowledge first, always
⟶ Hour 0–1 · Internal assembly
. assemble crisis team: CEO · Legal · Comms · functional lead ☐
@ single comms lead designated — all external messages route through them
. establish facts: what happened · confirmed scope · what is still unknown ☐
! separate confirmed facts from speculation · mark clearly in all internal notes !!
. draft holding statement ☐
* structure: acknowledge · ownership · action · timeline for update
* example: "We are aware of [X]. We take this seriously. We are [doing Y]. We will provide an update by [time]."
! do NOT include apologies for events not yet confirmed — legal liability !!
⟶ Hour 1–2 · First external communication
. publish holding statement: website + social + direct to affected parties ☐
. notify key stakeholders before public — investors, major clients, regulators ☐
* "I wanted you to hear this from us before you see it elsewhere"
? media contact received?
→ yes: acknowledge receipt · "we will respond by [time]" · no interview without prep
→ no: monitor · set 30-min check-in cadence
⟶ Hour 2–4 · Full statement
. draft full statement once root cause is sufficiently understood ☐
~ speed vs accuracy — incomplete statement risks second crisis if corrected
* guideline: 80% certainty is enough to publish with explicit unknowns acknowledged
. CEO review + Legal sign-off before publish ☐
. publish + brief media spokesperson (if press conference needed) ☐
→ statement live · spokesperson briefed · update cadence set (every 4h or on change)
⟶ Ongoing
. internal update every 2 hours to crisis team — even if no change ☐
* silence breeds rumour internally and externally
. post-crisis debrief within 5 days: protocol review · what to update ☐
Building a protocol library
Individual protocols are useful. A protocol library — a versioned, linked collection of CLN protocols covering all major recurring interactions — is transformative. It is the institutional knowledge system that most organisations try and fail to build with wikis, Notion databases, and policy handbooks.
Protocol library design principles
One protocol per situation. Each protocol covers exactly one recurring interaction type. No omnibus documents. A "Client Communication Protocol" that covers onboarding, QBRs, escalations, and renewals is four protocols merged into one unnavigable document.
Version-controlled. Every protocol has a version number, a last-reviewed date, and a change log. Use @ metadata. Outdated protocols are worse than no protocols — they create confident incorrect behaviour.
Cross-linked. CLN's »[text]##anchor addressing lets protocols reference each other. The Escalation Protocol links to the Crisis Communication Protocol. The Hiring Protocol links to the Onboarding Protocol. The library is a graph, not a list.
AI-queryable. A library of CLN protocols is directly queryable by any LLM. Because the semantic type of every line is pre-classified, an AI can answer "what are our prohibitions in client escalation?" by scanning for ✗ and !! lines in the relevant protocol — without natural language parsing.
protocol-library-index.cln — example structure
LIBRARY: Business Protocol Index
@ owner: COO
@ review: quarterly · all protocol owners
⟶ Commercial
→ »[Enterprise Negotiation]##PROTOCOL-DEAL-001 v2.1 · reviewed 2026-01
→ »[SMB Negotiation]##PROTOCOL-DEAL-002 v1.3 · reviewed 2025-11
→ »[Partnership / BD]##PROTOCOL-DEAL-003 v1.0 · reviewed 2025-09 ? due review
⟶ Customer Success
→ »[Client Escalation]##PROTOCOL-CS-003 v3.0 · reviewed 2026-02
→ »[QBR Protocol]##PROTOCOL-CS-004 v2.2 · reviewed 2026-01
→ »[Churn Recovery]##PROTOCOL-CS-007 v1.1 · reviewed 2025-10 ? due review
⟶ People
→ »[Senior Engineer Hiring]##PROTOCOL-HR-002 v2.3 · reviewed 2026-02
→ »[Executive Hiring]##PROTOCOL-HR-005 v1.2 · reviewed 2025-12
→ »[Performance Improvement]##PROTOCOL-HR-009 v1.0 · reviewed 2025-06 ! overdue
⟶ Executive
→ »[Board Meeting]##PROTOCOL-EXEC-005 v2.0 · reviewed 2026-01
→ »[Investor Update]##PROTOCOL-EXEC-006 v1.1 · reviewed 2025-11
→ »[Crisis Communication]##PROTOCOL-COMMS-001 v3.1 · reviewed 2026-02
⟶ Library health
? protocols with review > 6 months ago?
→ PROTOCOL-DEAL-003 · PROTOCOL-CS-007 · PROTOCOL-HR-009
! PROTOCOL-HR-009 — performance process changed · do not use until updated !!
From tribal to institutional: the transfer protocol
The practical question is how to extract the tacit knowledge from senior people and encode it in CLN without requiring them to write documentation — an activity most senior people actively resist.
The most effective method is the observation-annotation loop: a junior team member observes a senior executing a protocol, then writes a CLN draft of what they saw. The senior reviews the draft — not to write, but to react. Reactions are fast: "that's wrong," "add a risk here," "this step only applies if X." The CLN draft becomes the anchor for a 30-minute conversation that captures years of tacit knowledge.
This works because CLN's structure makes the gaps in documentation immediately visible. If a senior person looks at a CLN protocol and sees a ? branch with no targets, or a ! with no consequence, they immediately know something is missing — without having to read the document as prose first.
The institutional memory problem is not that knowledge is impossible to capture. It is that the available formats make capturing it feel harder than it is worth. CLN shifts the burden from the knowledge-holder (write a document) to the observer (draft a CLN, then react). The senior person edits; they don't author. The editing is fast. The result is institutional.