The Verbs Behind NANOG’s Executive Office

Summary

• NANOG’s budget, reserve, audit and Workshop records show a legible division between Board authority, executive implementation, staff or committee work, and later review. They also show why an Executive Director’s vote as one of seven Directors must be distinguished from the same person’s reports and administrative acts as an officer and employee.
• Discord, marketing and event-insurance records disclose genuine work but not every approval, contract term or outcome. A small, versioned authority map would make those boundaries easier to follow without exposing personnel files, banking details, legal advice, security procedures or confidential conduct cases.

The difference between permission and action

On 8 June 2025, NANOG’s Board authorised the Executive Director and Treasurer to withdraw up to USD 200,000 from reserves between 30 June and 31 July for operating expenses. Six weeks later, the Executive Director reported that USD 75,000 had been withdrawn from savings for June expenses and that the organisation remained ahead of budget for the year to date. The 8 June record and 18 July record do something minutes often fail to do: they let a reader follow authority into an act and then back into a report.

The distinctions are easy to flatten. USD 200,000 was a ceiling, not the sum used. USD 75,000 was the amount reported withdrawn, not a new standing limit. The authority was conferred jointly on two corporate roles, not personally on one office-holder. And the Executive Director’s report was an operational presentation, not a second Board decision, even though the Executive Director also sits on the Board. Each verb identifies a different institutional event.

The chain began with the budget. On 20 December 2024, the Board approved the FY2025 budget by a recorded 7–0–0 vote in the published budget minutes. That decision established the annual plan. The later resolution created bounded reserve authority. The July update recorded a smaller transaction. What remains out of view includes the account, exact transfer date, invoices, allocation of the USD 75,000 and disposition of any unused authority after the window closed. Those omissions prevent a transaction-level audit by the reader. They do not negate the sequence that is visible, and they are not evidence that concealed misconduct occurred.

This short episode supplies a better way to examine executive power than a job description does. Start with the authorising instrument. Identify the person’s capacity. Name the action and its entity. Find the supervision or review. Then separate the public outcome from the facts that remain unknown. Applied across NANOG’s operating surfaces, that method yields neither a celebration of professional management nor an accusation of staff rule. It yields a map of how a corporation gets things done.

A corporate forum, with a corporate limit

NANOG’s current Bylaws define NANOG, Inc. as a Delaware nonprofit corporation providing North American forums for education and the exchange of knowledge about Internet operations. The document says expressly that NANOG is not itself a network operator. Its Board manages and controls the corporation’s property, affairs and business and may delegate within the Bylaws by general resolution. That is meaningful power over NANOG’s own organisation. It does not reach the member networks represented at its conferences, their employers, Autonomous System Numbers, BGP routes, Internet addresses, registries or public law.

The boundary matters because operational visibility can resemble public authority from a distance. Staff can run a conference, administer a communication platform and carry out a conduct process. None of those functions makes NANOG a government, regulator, Regional Internet Registry or network operator. The Executive Director cannot direct how an independent network routes traffic simply because that network’s staff participate in NANOG. Corporate administration and control of Internet infrastructure are different categories of power.

Inside the corporate boundary, the Board holds the governing centre. The live Board Responsibilities page assigns it the hiring, support, evaluation and release of the Executive Director; approval of the annual budget and major commitments, expenditures, loans and leases; protection of assets; and definition of the relationship with the executive, including the degree of delegation. The Bylaws provide the sharper legal grammar: the Board selects corporate officers, while the six elected Directors appoint the Executive Director, determine that person’s salary and may remove the appointee with or without cause.

There are seven voting Directors: six elected members and the Executive Director. The same human being may therefore act in three capacities. Casting a vote is the act of one Director among seven. Presenting an Executive Director Update is ordinarily an officer or employee act. Supervising staff belongs to the operating role. A Board vote cannot be converted into unilateral executive action merely because the Executive Director took part, just as an operational report cannot be treated as a Board resolution. The distinction is not etiquette; it is the basic unit of accountability.

Other offices prevent the executive relationship from collapsing into a two-person diagram. The Chair, with advice from the Executive Director, oversees implementation of Board orders and resolutions. The Treasurer is responsible for finance and fiscal records and supervises the fiscal agent. Staff implement tasks under executive oversight. Volunteer committees recommend, select or review within their remits. Vendors perform contracted work. The public Board-minutes index is where some of these assignments become dated acts, but the presence of a vendor does not prove a compliant procurement, committee advice does not command staff, and staff execution does not author policy.

Even the governing description needs maintenance. The responsibilities page says elected Directors serve two-year terms; the current Bylaws say three. The Bylaws supply the current formal rule, while the stale two-year copy is documentary drift. This is not proof of an improperly constituted Board. It is a modest example of the need to identify which source controls and when it was updated.

Nine surfaces, not one executive mandate

The phrase “day-to-day affairs” is broad enough to be useful and too broad to answer difficult questions by itself. NANOG’s documents reveal at least nine distinct operating surfaces: appointment and compensation; staff supervision; finance and payment; procurement and contracts; audit and records; communications and moderation; programmes and events; insurance and safety; and conduct. The allocation changes from one to another.

Finance offers the most explicit contemporary limits. Under the 2025 Internal Financial Controls Policy, all expenses are proposed by the Executive Director. Expenditures below USD 120,000 may be approved by the Executive Director; those at or above USD 120,000 require Board approval. Payment by the bookkeeper still needs approval from the Executive Director and at least one authorised Director, or from two Directors. The authorised Directors are the Treasurer or Chair.

Proposing, approving and paying are separate operations. An executive proposal does not commit NANOG. Authority below USD 120,000 does not remove dual payment approval. The special reserve permission capped at USD 200,000 does not displace the general threshold. Most importantly, a published policy describes the control design; it does not prove that any particular payment complied with it.

The policy also distributes routine checks. Reconciliation should be done by someone who did not receive, record and approve the payments. Where staffing prevents that separation, the Treasurer reviews and signs the reconciliation. Credit-card spending is reconciled monthly, approved by the Executive Director and reported to the Treasurer. Conflict acknowledgements are expected when a Director is appointed or an employee is hired, after a material career change and annually.

Directors, the Executive Director and employees are prohibited from self-dealing, and procurement is expected to use reputable suppliers not controlled by those people or their families. The Board must appoint an Audit Committee each year to oversee the independent auditor.

Compensation illustrates a second division. The Executive Director may adjust staff compensation within the compensation budget the Board approved. The elected Directors, however, determine the Executive Director’s own compensation. Neither “the Board sets every salary” nor “the executive sets compensation” accurately describes the published arrangement. It consists of a Board budget, delegated management of employee pay within that envelope and a reserved decision about the executive’s pay.

These provisions make professionalisation intelligible. Financial administration, contract calendars, registration, committee support, sponsorships, records, moderation and incident response do not wait for a monthly meeting. Paid staff can maintain continuity across staggered volunteer terms and allow Directors to govern rather than operate every workstation. That is the strongest case for an executive office. It is also why the limits should be observable: continuity is a necessary function, not an independent source of authority.

Professionalisation arrived as an overlap

NANOG did not pass in one clean instant from volunteers to employees. The official history, the 2010–2011 transition archive and Merit’s 2011 announcement describe the transfer of the NANOG trademark, resources and operating responsibility from the Merit-era arrangement to NewNOG/NANOG. Together they establish a transition. They do not establish the exact moment when every asset, legal duty, Secretariat function and practical responsibility changed hands.

Contemporaneous traces show multiple actors at once. The August 2011 announcement archive preserves post-transfer communications. The May 2012, September 2012 and October 2012 announcement archives record leadership, Secretariat and event-operating communications. An archived contact page provides a period roster. Across these sources, the Board, committees, an Executive Director, Secretariat functions and vendors coexist. The complete terms of Betty Burke’s appointment and the complete vendor contracts remain private or unretrieved.

The point is not merely historical. Describing 2010–2012 as the moment paid staff replaced volunteer governance would misstate the institutional design still visible today. Directors retained corporate decisions, committees retained defined volunteer work, executives coordinated recurring operations, and staff and vendors carried tasks through. Professional continuity grew within that mixture.

The 2017–2019 period makes the appointment surface easier to see. A 2017 Executive Director search page and a NANOG 71 member presentation documented a Board-led search and transition timetable. NANOG announced Edward McNair’s appointment on 14 March 2018. These records establish a public search process and appointment, not the full candidate pool, interviews, scoring, references, recusals, contract terms or subsequent evaluation.

The 2019 annual report distinguishes the Board’s governance, the Program Committee’s selection of content and the Executive Director and staff’s operation of the organisation. That separation matters whenever a conference programme is used as evidence. The NANOG 96 agenda is an observable programme output; page ownership does not reveal who proposed or selected each item, and staff administration is not unilateral power over technical consensus.

The 2023–2024 leadership change repeats the broad pattern without opening the personnel file. October 2023 Board minutes record executive-search and transition work. The 2024 annual report says the Board approved extending an employment offer to Jonathan Black in February 2024 and describes his onboarding. The record does not establish private reasons for McNair’s departure, Black’s comparison with other candidates, the length of the contract, severance, performance metrics or a private evaluation; it also does not establish the exact decision sequence for any Executive Director departure. Public silence on those subjects is not evidence of an improper process.

Across the four periods—2010–2012, 2017–2019, 2023–2024 and 2025–2026—the same caution applies. Institutional announcements can show a decision and operating documents can show activity. Neither gives access to every private term or proves that staff displaced the volunteer bodies whose authority remains in the governing instruments.

Compensation is a disclosure, not a verdict

Tax filings make the professional layer measurable, but not self-explanatory. NANOG’s 2016 Form 990 identifies the filer as NEWNOG Inc., using the NANOG name, and gives EIN 27-2534183. It reports seven voting governing members, six of them independent under the form’s definition, and no delegation of substantial management duties to a management company or other person. It also says the Executive Director reviewed the return and the Board approved Executive Director compensation.

Betty Burke appears as Executive Director and Secretary with USD 121,448 in compensation and USD 8,517 in other compensation. The listed Directors received zero compensation in the disclosed summary. These figures describe what the organisation reported for the year. They do not establish the value of Burke’s work, the motivation for the amount or the propriety of the decision. The filing does not reproduce the full deliberation, comparable salaries or recusals.

The 2022 Form 990 reports the same seven-member and six-independent counts, no management-company delegation, Executive Director review of the return and Board approval of Executive Director compensation. It lists Edward McNair at USD 223,513 in compensation and USD 13,621 in other compensation. It does not disclose a peer group, entity recusals or a contemporaneous written rationale.

The apparent tension between executive review of the return and Board approval of executive pay should be described, not dramatised. The acts can coexist. An officer may review an institutional filing while elected Directors reserve the compensation decision. The public record does not show every stage of circulation or deliberation, so it cannot prove either rigorous transaction-specific independence or improper self-review.

“Independent” itself has a narrow meaning in this context. The IRS explanation of Form 990 Part VI defines an independent voting member for federal reporting. A six-of-seven count is not automatically a finding of Delaware-law independence, proof of compliance with NANOG’s conflict policy or evidence that a particular compensation or contract decision was conflict-free. Form 990 is organisational self-reporting; availability of the filing does not amount to IRS approval of the disclosed governance practice.

ProPublica’s Nonprofit Explorer provides a separate access layer to later filings. Its summary reports 2023 compensation of USD 230,064 for Edward McNair, USD 150,000 for Gregory Newman, USD 119,803 for Darrieux Williams Harvey and USD 118,413 for Valerie Wittkop, with other compensation disclosed separately. For 2024 it reports USD 124,615 for Jonathan Black, USD 155,625 for Gregory Newman, USD 137,918 for Darrieux Williams Harvey, USD 136,317 for Valerie Wittkop and USD 117,521 for Shawn Winstead, again plus separately disclosed other compensation.

Those numbers show that staffing is a material corporate expense and that the annual disclosure changed during a leadership transition. They do not answer whether any person was well paid for the value delivered, whether professionalisation improved performance, or who took part in each decision. The unknowns include the complete salary comparables, peer set, compensation-consultant work and identities or recusals of the entities. A defensible public improvement would disclose process metadata—authority, comparator class and recusal method—while keeping evaluations and private employment information private.

What an audit does not say

The 2023/2022 audited financial statements, in an auditor’s report dated 9 June 2024, give an opinion on whether the statements fairly present NANOG’s finances under US generally accepted accounting principles. Management is responsible for the statements and for designing, implementing and maintaining the related internal controls. The auditor expressly did not express an opinion on the effectiveness of those controls.

That ceiling is central. The opinion cannot certify that a delegation was valid, a procurement was competitive, or every invoice followed the approval process. It does reveal scale. For 2023, the statements report USD 972,531 in salaries and wages and USD 191,837 in contract services. They say sponsorship agreements, including in-kind sponsorship contracts, were signed by sponsors and the Executive Director. This is a specific signature surface, not a general licence to sign every contract.

The statements also report USD 1,501,943 in Board-designated future meeting commitments under hotel contracts running into 2025. Such commitments show why contract administration and institutional memory matter. They do not publish every authorising motion or vendor comparison. The 2024 audited statements list USD 889,735 in salaries and wages, USD 193,934 in contract services and hotel commitments extending into 2026. Again, the opinion does not cover internal-control effectiveness.

NANOG’s financial-reports index improves discoverability, but a collection of audits and returns is not a transaction ledger. Historical procurement thresholds before the 2025 policy remain unknown. So do transaction proposals, invoices, payment approvals, bank mandates and reconciliation sign-offs. The records cannot show whether every expense complied with the later policy, and the 2025 policy cannot retroactively prove historical compliance.

One audit sequence is nevertheless unusually complete. On 11 December 2024, following an Audit Committee recommendation, the Board approved entering an agreement with Smith and Schaefer for the coming fiscal year. The December minutes also assign staff a separate business-address action, placing Board approval and staff implementation side by side. On 8 June 2025, the Executive Director presented draft audited statements and a draft governance letter. The Board and Audit Committee reviewed risks identified by the auditor, the minutes reported that mitigation steps had been taken, and the Board accepted the 2024 financial statements.

This track has an authority source, named vendor, executive presentation, committee and Board review, and final acceptance. It still does not disclose the engagement’s amount and full scope, the vendor’s independence statement or procurement comparison. Nor are the governance letter, workpapers, complete risks, remediation owners, deadlines and closure evidence public. “Mitigation steps had been taken” is not evidence that every issue was closed. Board acceptance cannot transform a financial-statement opinion into an opinion on controls.

From a charter to a Workshop seat

Programmes provide a different kind of complete chain. On 17 January 2025, the Board approved a Workshop Committee charter and assigned staff to add the committee to the nominations list, according to the January minutes. At the continued February meeting, the Board accepted appointees and designated a liaison in the February minutes. On 18 July, the Board approved the start of a Workshop pilot. Committee and staff effort then became visible to prospective entities.

At NANOG 95, the programme offered the hands-on “BGP For Data Centers” Workshop, limited to 50 people and requiring preregistration. The current Call for Content describes Workshops as 120–180 minutes, separately registered and free to attendees. A staff announcement in the May 2026 attendee archive offered two NANOG 97 Workshops, each capped at 50.

Here the Board created the committee and approved the pilot; staff administered nominations and public communication; appointees and a liaison supplied the volunteer structure; public sessions supplied an observable output. That does not mean staff authored all content, or that a capacity of 50 became 50 attendees. The record does not give the nominations, candidate assessments, attendance, completion, learning evaluation or later decision about the pilot. An offering establishes activity and capacity, not participation or educational effect.

The Workshop chain is strong precisely because its uncertainty is bounded. It is possible to say what was authorised, who implemented parts of it, how the committee connected to the Board and what the public could register for. It is equally possible to say what has not been measured. A short pilot report with offered capacity, actual attendance, completion, an identified learning measure and the follow-on decision would add accountability without identifying entities.

Discord and the difference between operation and publication

Communications policy divides rule-making from administration. The Bylaws and the mailing-list Usage Guidelines place platform administration and minimal moderation with staff under Executive Director oversight, while the Board approves changes to acceptable-use rules. In February 2025, the Board approved Discord Usage Guidelines, said they would be posted publicly and made Discord NANOG’s official chat platform.

The following records show operation and review. The March minutes say the Moderation Committee was reviewing guidelines. The 20 June minutes record a Board-directed check for uniform moderation holds, as well as Executive Director support and insurance work. An August 2025 newsletter archive publicly promotes Discord. Board policy, committee review, staff operation, executive oversight and public use can therefore be connected.

Publication is the missing link. The material reviewed for this article did not independently locate a standalone page containing the Discord Usage Guidelines. Minutes saying that a policy “will be posted” are not evidence that a retrievable version appeared. The record also does not supply a version history, effective date, moderation log, individual warnings or sanctions, appeal records or outcome measures. Some case records should remain confidential, both to protect entities and to avoid turning enforcement into public spectacle.

A stable policy text and effective date are different: publishing them would verify the rule without disclosing a case.

Nor does control of a NANOG platform expand NANOG’s corporate remit. Staff moderation can affect participation in that platform under an approved policy. It does not control a person’s employment, outside speech, network or routes. Even within NANOG, ordinary moderation and membership expulsion concern different entities and different decision-makers.

Carlton, sponsorship and an unearned causal arrow

The marketing sequence begins with broad Board decisions. The FY2025 budget was already in place, and in February the Board approved the 2025 Strategic Plan. In March, the Executive Director reported that NANOG would enter an agreement with Carlton Group and that Denver sponsorship stood at 55 per cent of target. In April, the Executive Director reported Carlton meetings, a postcard intended for community distribution and updates to sponsorship material; sponsorship then stood at 81 per cent. The April minutes preserve that snapshot. In May, the Executive Director reported that postcards had been created and distributed and that strategic-plan work was under way, as the May minutes record.

This is evidence of executive reports, vendor contact and tangible products. It is not a Carlton-specific Board approval. The Board’s recorded approval of Smith and Schaefer cannot be borrowed to complete a different vendor’s chain. The contemporary policy may have allowed executive approval below USD 120,000, but the public record here does not give the Carlton price, scope, threshold class, procurement path, full deliverables, authorship or return on investment. It also cannot establish that no confidential Board authorisation existed.

The movement from 55 to 81 per cent happened alongside the reported work. It does not prove that Carlton caused the increase. Sponsor timing, staff effort, existing relationships, event content or external conditions could also have contributed. The public NANOG 94 Conference Kit shows messaging, logos and social assets; it does not attribute authorship to Carlton. A postcard is an output. A sponsorship percentage is a contemporaneous indicator. Neither is a causal evaluation.

Useful disclosure need not compromise negotiation. Nonconfidential metadata could name the authorising instrument, actor and capacity, date, generic scope, threshold class, conflict treatment and review owner. That would show whether the agreement rested on Board approval, delegated authority within a threshold or some other instrument, without revealing commercial terms. It would also keep reporting honest by separating delivery from outcome and outcome from causation.

Insurance is not a safety finding

Event insurance appears in repeated Executive Director updates rather than one fully documented approval chain. In January, insurance was being explored or planned. March brought another status report. By April and May, it was described as in place or secured. NANOG 94 was then held in Denver, which the venue page confirms. The June minutes record badge reprints, badge theft, the removal of unbadged people and an action to monitor the issue in future.

This is evidence of planning, reported implementation, a held event, incident response and prospective review. It does not disclose the insurer, premium, coverage, limits, exclusions, claims or final injury protection. It also does not say whether the future monitoring produced a change. Insurance may be prudent, but the existence of a policy does not prove that it made the conference safer. The fact that the conference happened does not prove there were no claims. The presence of badge incidents does not prove negligence, and recording them does not prove that remediation worked.

Safety illustrates why both transparency maximalism and institutional reassurance are inadequate. Publishing badge identifiers, names of incident subjects, vulnerabilities or detailed response procedures could expose people and weaken security. Saying only that insurance was secured and an event took place would obscure the missing evidence. A proportionate summary could name the category of incident, responsible reviewing role, nonconfidential corrective-action status and measures not taken, while withholding exploitable detail.

The marketing and insurance tracks are partial for different reasons. Marketing lacks a public contract approval and causal evaluation; insurance lacks contract detail and outcome measurement. Both show work. Neither permits a conclusion about procurement compliance, vendor performance, safety or claims.

Conduct and the edge of ordinary discipline

NANOG’s Code of Conduct makes a consequential distinction. The Executive Director and Ombuds investigate reports and recommend responses. The Executive Director may implement ordinary discipline. Suspension or expulsion from membership, however, requires Board action. The public Ombuds page describes a reporting and support channel, not a dossier of cases or outcomes.

This policy-level chain is complete enough to allocate responsibility. A report reaches designated investigators; they assess and recommend; the Executive Director can apply an ordinary remedy; the Board alone can impose the membership consequence. Staff are not powerless to administer a forum between Board meetings, but their operational role does not grow into unilateral expulsion authority.

The case-level evidence is intentionally absent. Public sources do not reveal complaints, witnesses, investigative material, ordinary remedies, Board membership decisions, appeals or results. Those records may contain private facts and legal advice. Their absence is not proof that a review failed to occur. The public interest lies in reliable process information—recipient, investigator, recommending role, remedy class, decision-maker and appeal route—not in exposure of complainants or respondents.

Two small executive roles reinforce the rule against reading a narrow power broadly. The conference-rates policy gives the Executive Director a limited registration-fee waiver role with Board support; it is not general unilateral pricing authority. In an election tie, the Bylaws make the Executive Director a witness to a random selection. Witnessing that procedure does not confer control over nominations, tabulation rules or election outcomes.

Programmes and events follow the same logic. The Board determines the number and location of conferences and approves corporate commitments. The Program Committee selects programme content. The Executive Director and staff administer operations. Volunteers can advise or select without commanding employees; employees can implement without originating policy; a contractor can deliver a service without proving the procurement was compliant. The institution works through these differentiated capacities, not through one executive title.

The case against indiscriminate disclosure

Any proposal for greater visibility should begin by accepting the strongest objection. A volunteer Board cannot itself reconcile accounts, administer registrations, negotiate every hotel detail, maintain platforms, support committees, process sponsorships, coordinate insurance and respond to incidents each day. Paid professionals preserve records, contractual memory and service continuity across Board elections. Asking Directors to operate every process would consume their time, blur governance with execution and make volunteer service harder.

Privacy, bargaining, security and legal privilege are equally real. Candidate references and comparative assessments depend on confidentiality. Performance reviews and employment agreements may contain personal data. Vendor files contain prices, negotiating positions and trade secrets. Bank mandates and payment credentials create fraud risks. Audit workpapers and governance letters may include sensitive control or legal information. Insurance limits can reveal operational exposure. Conduct files can endanger complainants, respondents and witnesses.

The existing public record is not barren. It contains Board votes, Treasurer participation, committee liaisons, an annually appointed Audit Committee, tax and financial disclosures, staff outputs and repeated Executive Director reports. Silence in public minutes does not prove that a private approval, executive-session discussion, recusal or internal delegation did not occur. A missing public delegation matrix is not itself proof of defective delegation. Earlier procurement thresholds may have existed before the 2025 policy.

The evidentiary ceiling must therefore remain low enough to be credible. No case-specific conclusion follows here about legality, fiduciary duty, employment rights, negligence, procurement compliance or personal motive. NANOG’s present Delaware good standing is also unknown because no current state certificate was obtained. The sources cannot answer whether professional staff improved continuity relative to a comparable volunteer-only organisation; there is no counterfactual. They cannot tell whether a public update omitted a private approval.

The fact that such questions remain open is a reason for careful language, not a warrant for suspicion.

The most important unknowns cluster by surface. Personnel files would be needed for employment terms, benefits, confidentiality clauses, severance, performance metrics, departure reasons, candidate pools, interviews, scores, references and recusals. Compensation review would require peer sets, consultant analysis and decision-entity records. Financial verification would require transaction proposals, invoices, payment approvals, bank mandates and reconciliation sign-offs. The audit chain stops before the complete governance letter, workpapers, control deficiencies, remediation owners and proof of closure.

Vendor assessment stops before the Carlton scope, procurement route, authorship and return on investment. Platform and programme evaluation stops before Discord case logs and appeals, or Workshop attendance, completion and learning. Insurance and conduct remain bounded by security and privacy.

That is a powerful argument for selective confidentiality and professional continuity. It deserves more than a ritual acknowledgment. The answer is not to abolish the executive office, expose private records or make a volunteer Board narrate every minor task. It is to publish a small amount of structured information at the boundary between decision and execution.

A public authority map that fits the risk

NANOG’s disclosure problem is one of assembly. A reader can often find the pieces, but must cross governing documents, policies, minutes, filings and public outputs to determine what an act meant. A compact, versioned authority map could reduce that burden. Its rows would cover the nine operating surfaces. Its columns would identify the authority source, acting capacity, threshold where applicable, implementation role, review owner, observable output and deliberately withheld field.

The finance row could state that the Board approves the budget and spending at or above USD 120,000, the Executive Director proposes all expenses and may approve those below the threshold, and payment remains dual. It would preserve the Treasurer’s separate role and link transaction-specific resolutions rather than treating a USD 200,000 reserve ceiling as the general rule. An annual control-attestation summary could state which policies were reviewed, how many exceptions were identified and whether nonconfidential remediation closed, without publishing accounts, invoices or credentials.

Contract metadata could identify the authorising instrument, approval date, actor and capacity, generic scope, threshold class, conflict or recusal process and review owner. It would make the documented Smith and Schaefer approval visibly different from the reported Carlton agreement. Publication of that distinction would not require pricing, bids, legal advice or trade secrets.

Personnel metadata could separate the elected Directors’ power to appoint, remove and compensate the Executive Director from the Executive Director’s supervision of staff and pay adjustments within the approved compensation budget. Search and compensation reports could name the process, comparator class and recusal method without naming unsuccessful candidates or publishing evaluations. Minutes could label the Executive Director as voting Director, presenting officer, employee supervisor or contract manager when the capacity affects interpretation.

For Workshops, a public pilot record could connect charter, appointees, liaison, staff administration and sessions, then add attendance, completion, whatever learning measure was used and the follow-on decision. For Discord, NANOG could provide a stable approved policy, effective date and version history while withholding case material. For event safety, it could publish incident categories, responsible review and nonconfidential closure status without revealing personal or exploitable information. Conduct materials should preserve the distinction between investigation, ordinary remedy and Board-only membership action.

The map must also state what NANOG cannot delegate because it never possessed the authority: control of member networks, employers, ASNs, BGP routes, number resources, registries or public law. That negative space would prevent a corporate role from being mistaken for community sovereignty.

Two narrower repairs follow the same principle. The Board Responsibilities page should be reconciled with the Bylaws’ three-year term. When an electronic vote or written consent is later ratified, nonconfidential metadata about the original decision can be linked so the later minutes do not carry the entire explanatory burden. Public versions of Board-approved platform policies should carry dates. Pilot outputs should identify what was not measured. None of this requires passwords, account identifiers, medical details, private evaluations, legal advice or trade secrets.

An executive office measured by its links

NANOG’s Executive Director has a substantial operating surface: day-to-day administration, employee supervision, proposals and bounded financial approvals, reports, specifically evidenced sponsorship signatures, committee support, communication oversight, event work and ordinary conduct remedies. But the surface is divided by Board direction, reserved decisions, Chair and Treasurer responsibilities, dual approvals, volunteer committees, audits and the corporation’s limited mandate.

The three strongest action chains make the architecture visible. A Board-approved budget led to joint reserve authority and a reported USD 75,000 withdrawal. An Audit Committee recommendation led to approval of a named auditor, executive presentation, risk review and Board acceptance. A Workshop charter led to staff nomination administration, appointees, a liaison, a pilot and public sessions. Their omissions are specific rather than limitless.

The four partial chains define the edge. Discord links policy approval to operation but not to a verified standalone guidelines page or case record. Marketing links strategy to a reported Carlton relationship and public materials without a contract motion, attribution or causal return. Insurance links executive reports, a held event and incidents without coverage detail, claims or proof of safer outcomes. Conduct supplies a clear policy path while appropriately withholding private cases.

Professional staff are not an intrusion into volunteer governance simply because they are paid, and privacy is not evidence of concealment simply because it limits public detail. Equally, competence does not answer an authority question, and an output does not prove an outcome. Trust is strongest when readers can see which capacity acted, under which instrument, within which limit, with whose review—and where the public record properly stops.

SEO and social metadata

• SEO title: The Verbs Behind NANOG’s Executive Office
• SEO description: An evidence-led map of how NANOG’s Board, Executive Director, staff, committees and vendors move decisions through finance, audits, Workshops, communications, events and conduct—and where public authority remains incomplete.
• Social title: How NANOG Turns Board Decisions Into Operations
• Social description: NANOG’s records reveal bounded executive action across nine operating surfaces. The strongest chains show oversight; the gaps call for proportionate authority metadata, not private-file disclosure.

Image metadata

• Alt text: Synthetic editorial cross-section of a nonprofit operations desk, with an abstract Board decision passing through a clearly separate executive station into finance, Workshop, communications and event workstations, then returning as audit and reporting records; no people, logos or readable text are shown.
• Caption: An executive role becomes legible when authority can be followed into implementation and back through review.
• Accessibility description: A wide editorial illustration uses distinct desk zones, pathways, shapes and colour to separate Board governance, executive coordination, finance, programmes, communications, events and reporting. The composition communicates through spatial relationships rather than readable labels and contains no brand marks or identifiable people.
• Synthetic image provenance: Original synthetic editorial illustration concept. It must not imitate a historical document, depict a real person, reproduce a logo or include readable policy text.