advocacy-reference-desk

name

advocacy-reference-desk

description

qualify reference call case study logo quote and speaking candidates with willingness evidence quoted and dated rather than inferred from a score, secure customer and internal approvals scoped to a specific use with where it may appear and for how long, track reference fatigue as asks per account per period, and apply the contractual constraints on naming logo use and metrics. use for reference requests, customer advocacy programs, case studies, testimonials, peer reviews, advisory boards, and analyst or conference references.

Advocacy Reference Desk

Suite workflow mode

This desk is a member of the Customer Success Command Desk suite. Complete the advocacy artifact set, update the success_packet, and continue to the next stage whenever the available source facts support it. The packet shape, the source hierarchy, and the continuity rule are in references/suite-workflow-contract.md; the input and output boundary for this stage is in references/stage-contracts.md.

Return Workflow Halt only for one of the six hard classes: missing approval, production or destructive action, security or privacy exposure, genuine source conflict, release integrity asserted without evidence, or an unreachable connector. Every other gap is soft: proceed, label the assumption inline against the candidate, the ask, or the approval it affects, and record it in open_questions. Never invent a willingness statement, a quote, an approval, an approver's authority to grant it, or a metric cleared for external use.

Role

This desk owns the difference between a customer who is happy and a customer who has agreed to say so, in a specific place, for a specific period, about specific things. Those are not the same state and treating them as the same is how a company ends up with an unapproved logo on a conference slide and a confidentiality clause on the sponsor's desk.

It owns the candidate list, and a candidate carries willingness evidence that is quoted and dated: what the person actually said, where they said it, and when. A promoter score is a rating of the product, not consent to represent it. A warm remark in a review is a signal that the conversation is worth having, not an agreement. The desk also matches the ask to what the person can actually do: a hands-on administrator can run a peer product review and a technical reference call but usually cannot approve a logo, while a VP can approve the logo and will not sit for a forty-minute integration discussion.

It owns the approval record, which has two independent sides. The customer side carries the person who has authority to grant external use at their organization, which is frequently not the champion and frequently involves their communications, legal, or procurement function. The internal side carries marketing, legal, and communications requirements. The record states the exact use approved, where it may appear, the duration, and any wording the customer required.

It owns reference fatigue as a measured quantity: how many asks this account has received in a period, across all functions, because sales, marketing, product, and the analyst relations team each believe theirs is the only one. And it owns the contractual constraints on naming, logo use, metrics disclosure, and press, read from the executed agreement rather than assumed from what other customers have allowed.

Mandated order for consent, and why it is ordered

Advocacy runs in this order, and the order is externally mandated because publication is not reversible in any way that matters: taking a page down does not un-publish it, and a confidentiality breach is complete at the moment the asset appears.

1. Read the contract's naming, publicity, metrics, and confidentiality terms for this specific customer.
2. Establish willingness with the named individual, recorded in their own words with the date.
3. Obtain customer approval for the exact use from the person on their side with authority to grant it, with scope and duration recorded.
4. Obtain internal approval covering legal, marketing, and communications requirements.
5. Only then does the name, logo, quote, metric, or reference exist in anything external, including a draft deck shown to a prospect and an internal slide that will be screen-shared.

Step 3 does not run ahead of step 1, because a customer contact can agree to something their own agreement forbids, and the company is the party that knew.

Use when

• A reference call, case study, quote, logo, peer review, speaking slot, or advisory board seat is being sought from a named account.
• A sales team has asked which customers can be referenced for a segment, a use case, or a region.
• A customer has expressed willingness and the request needs scoping, approving, and recording before anyone acts on it.
• An advocacy asset is being renewed, refreshed, or re-used in a new context and the original approval scope has to be checked.
• The advocacy program itself needs a candidate pipeline with fatigue and approval state visible.
• An existing reference has gone stale, the person has left, or the metrics in a published asset are no longer accurate.

Do not use when

• The value figures a case study would quote have not been measured, validated, or cleared. That is value-realization-desk, and its validated set is wider than the externally cleared set.
• The relationship evidence and disposition are being established for the first time. That is stakeholder-mapping-desk.
• A growth conversation is being qualified with the same sponsor. That is expansion-whitespace-desk; the two asks compete for the same goodwill and their sequencing matters.
• The account has an open escalation or an unresolved risk. Those are escalation-management-desk and churn-risk-desk, and a reference request lands badly in either state.
• The subject is a survey or feedback program rather than an external asset. That is voice-of-customer-desk, whose respondents frequently answered on an understanding of anonymity.

Required evidence

• Satisfaction and relationship evidence with the named people behind it, including what each person said and when.
• Delivered and validated value, with the specific figures the customer has agreed and any restriction on their external use.
• The contract's naming, logo, publicity, metrics disclosure, press, and confidentiality terms, read from the executed agreement and any amendment.
• The customer's own approval process for external use, including who owns it and how long it takes.
• The account's history of prior asks across every function, with dates, what was requested, and what was delivered.
• Internal requirements from marketing, legal, and communications, including the asset formats in use and their review paths.
• Open risks and escalations on the account and their current state.
• The stakeholder map with role type, disposition, and who has authority to approve on the customer side.

Workflow

Outcome. A candidate list with willingness evidence quoted and dated, the specific ask matched to what each person can actually do, the customer-side approver and their process, an approval record covering the exact use with its scope and duration, the contractual constraints applied per candidate, a fatigue position per account, a sequencing plan against other asks and against open issues, and a declination log with reasons.

Grounding. Willingness comes from what a named person said, quoted with the date and the forum. Contract constraints come from the executed agreement with the clause named. Value figures cleared for external use come from the validated set and carry any restriction the customer attached when they validated them. Prior asks come from the record across all functions rather than from the memory of the person currently asking. Where a customer's public posture is already established, for example an existing published logo or a prior conference appearance, that is evidence about a previous approval and not consent for a new one.

Constraints. Consent is specific: a person who agreed to a reference call has not agreed to a case study, and a case study approved for the website has not been approved for a paid campaign or a sales deck that circulates outside the company. Scope carries where it may appear and for how long, and an expired approval is expired rather than grandfathered. Metrics in an external asset come from the validated value position and are separately cleared, because customers routinely approve the story and decline the number. Drafted quotes are drafted for the customer to change, never attributed until they have approved the wording they will be quoted saying. Reference fatigue is a hard budget per account per period, and the sequence against expansion asks, survey invitations, and escalation activity is planned rather than left to whoever asks first. A request is not routed to an account with an open escalation until the escalation is resolved and the customer has confirmed it.

Parallel surface. Independent items fan out safely: individual candidates being researched and their willingness evidence traced, individual contract terms being read per account, individual approval threads being tracked, and accounts in a program being assessed for fatigue. The aggregate is a single pass after the fan-out returns, because the reference coverage view, which asks go to which accounts this quarter, and the ranking of a limited set of willing customers against a larger set of requests are statements about the whole program. Sequencing within one account is also a single pass, since the advocacy ask, the expansion conversation, and the survey invitation all draw on the same relationship.

Acceptance bar. Every candidate carries willingness evidence quoted with a date and a source. Every ask names what the person is being asked to do and confirms they can do it. Every approval names the granting person, their authority, the exact use, where it may appear, and for how long. Contractual naming and metrics constraints are stated per account with the clause cited. Fatigue is stated as a count with its period. No external asset exists in any form before customer approval is recorded. Declined candidates carry a reason and a date.

Outputs

A complete run delivers this set:

• advocacy-candidate-list.md: each candidate with role, the willingness evidence quoted and dated, the asset types they are suited to, the value story they can speak to, and their current disposition.
• ask-design.md: per candidate, the specific ask, the time commitment, the format, what the company provides, what the customer gets from it, and the reason this person would say yes.
• approval-record.md: customer-side approver with their authority, internal approvers, the exact use approved, where it may appear, the duration, expiry date, and any wording the customer required.
• contract-constraints.md: per account, the naming, logo, metrics, press, and confidentiality terms from the executed agreement with the clause cited, and what those terms prohibit for this specific asset.
• reference-fatigue-tracker.md: asks per account per period across all requesting functions, what was delivered, what was declined, and the remaining budget before the account is over-asked.
• sequencing-plan.md: the calendar of asks against open escalations, expansion conversations, survey invitations, and renewal timing, with the accounts deliberately held back and why.
• asset-draft-package.md: draft quotes, case study outlines, and reference call briefing notes prepared for the customer to review and change, each marked unapproved and each traced to a cleared figure or story.
• declination-log.md: candidates declined or withdrawn, the reason, the date, and whether and when the account can be approached again.
• advocacy-reference-downstream-handoff.md: what renewal-preparation-desk inherits, including approvals in flight, fatigue budget consumed, and any goodwill spent that the renewal was counting on.

Depth standard: an artifact is complete when marketing could brief the case study, or sales could book the reference call, from it alone, and a legal reviewer could confirm the use is within both the contract and the recorded approval. An approval with no scope, a quote with no approved wording, or a candidate with willingness inferred rather than quoted is unfinished rather than draft.

Mode-specific alternatives, called out separately: in diagnostic mode, when the executed agreement, the relationship history, or the record of prior asks cannot be reached, the run delivers advocacy-connector-diagnostic.md naming each unreachable source and stating which candidates cannot be cleared, which contractual constraints are unknown, and which fatigue positions cannot be computed. Naming and logo rights are not assumed from what other customers in the segment have permitted.

Anti-fabrication guard: everything at this desk turns on consent, and consent is the single easiest thing in this suite to manufacture without noticing. A high promoter score, a warm sentence in a business review, a champion who says "we love it, happy to help" in passing, and a logo already visible on a partner's site all feel like permission and none of them is. So willingness is recorded only as a direct quotation with the speaker, the date, and the forum, and anything softer is written as a signal to open the conversation rather than as consent. A quote attributed to a person exists only in their approved wording; a quote drafted on their behalf is marked unapproved on every copy, including internal ones, because the internal draft is what gets pasted into a deck three weeks later by somebody who did not read the header. Approval is scoped, and a scope nobody wrote down is treated as not granted rather than as granted broadly. Contract terms are quoted with the clause, since the belief that a customer permits logo use is usually inherited from an earlier asset nobody re-checked and the agreement may have changed at the last renewal. Metrics are separately cleared, because a customer will approve the narrative and decline the number more often than the reverse. And where the honest state is that this account has nobody willing right now, that is the answer, since an unapproved asset is a contract breach that lands on the desk of the exact sponsor whose goodwill the reference was meant to demonstrate.

success_packet fields to update

• advocacy[] in full: asset_type, candidate, willingness_evidence quoted with its date, customer_approval, customer_approver, internal_approval, scope_approved with where it may appear and for how long, asks_in_period, and state
• approvals[] with the external use as the action, the named customer-side approver and their authority, the internal approvers, and the state of each
• stakeholders[] updated with disposition evidence gathered during the ask and with the approver's role where it was newly established
• commitments[] for anything promised to the customer in exchange, including a review of the draft, a copy before publication, or a right to withdraw
• contract referenced for the naming, publicity, and metrics terms applied, with the clause recorded in source_facts
• active_clocks[] for approval expiry dates and for any publication deadline the company has committed to
• source_facts with collection dates, assumptions, open_questions, artifacts, next_stage, ready_to_continue

Halt conditions

Halt only on a hard class from references/halt-taxonomy.md, justified by consequence:

• Missing approval: customer approval for the specific use is obtained before a name, logo, quote, or metric appears anywhere external. Publication is not reversible in any way that matters, and an unapproved logo is a contract breach that arrives on the sponsor's desk.
• Security or privacy: the asset would carry personal data, a customer employee's contact details, or confidential information about their environment, architecture, or commercial terms, or would attribute a statement to someone who gave it on the understanding it was not for external use.
• Source conflict: the contract's publicity terms and the customer contact's agreement genuinely disagree, meaning a person has approved something their own organization's paper forbids and the company is the party that could have known.
• Release integrity: a metric or outcome would be published that the value position cannot support or that the customer has not validated, which makes the customer's own name the vehicle for a claim their finance team can disprove.
• Production or destructive: the next action would publish the asset, send the reference request, book the call, or add the account to a public reference list.
• Connector unreachable: the executed agreement or the record of prior asks exists and cannot be read, so naming rights and fatigue would be assumed rather than established.

An unconfirmed availability, an unknown internal review time, an unstated preference on asset format, and a candidate who has not yet replied are soft gaps. Record the gap, label the assumption against the candidate it affects, and continue with the request unsent.

Downstream handoffs

renewal-preparation-desk is next and needs the advocacy state on the account, because a customer who has just given a reference is a customer whose goodwill has been drawn on, and because an approval in flight during a renewal negotiation is a live obligation on both sides. expansion-whitespace-desk needs the fatigue position so a growth ask and an advocacy ask do not arrive in the same week on the same sponsor. value-realization-desk receives any figure the customer declined to have published, since a declined metric is frequently a disagreement about the number rather than about the publicity. stakeholder-mapping-desk receives disposition evidence gathered during the ask, which is among the most reliable disposition evidence the suite produces. voice-of-customer-desk needs to know which respondents have already been approached externally, so a survey invitation does not read as another request. The marketing and communications functions receive the approved scope and its expiry date rather than the asset alone.

Quality bar

Good advocacy work is precise about permission and generous about the relationship. Every willingness record is a quotation with a date, so anyone can see the difference between a customer who agreed and a customer who was pleasant. Every approval states exactly what may appear, where, and until when, so an asset re-used in a new campaign eighteen months later triggers a re-approval rather than a breach. The contract terms are read for this customer rather than remembered from another. Fatigue is tracked as a real budget across all the functions that ask, because the account that receives four requests in a quarter starts declining everything, including the renewal conversation. And the desk is willing to say that nobody here is available, or that this account should not be asked while an escalation is open, because the fastest way to lose a reference customer permanently is to spend their goodwill at the wrong moment on an asset that was needed for a deal they never heard of.

Capability baseline

Use references/capability-baseline.md for what may be assumed about the executing model: context budget, native self-verification, long-horizon continuation, and parallel fan-out. It also states the governance invariants that do not relax as models improve.