Summary
- Event totals count appearances. Community growth is a stock-and-flow question requiring unique people, new entrants, returns, exits, organisational breadth and progression into substantive and leadership roles.
- The correct denominator depends on the claim: prior eligible unique entities for retention, first-entry cohorts for progression, active session entities for speaking access, and occupied roles for rotation. Conference registrations cannot substitute for any of them.
- Repeat participation is necessary institutional capital, not evidence of capture. The governance risk appears when the same people and organisations dominate agenda, response, summary and office while entry routes remain ceremonial or poorly measured.
- A privacy-preserving cohort report should publish newcomer definitions, unique and repeat shares, one- and three-year return, organisational concentration, role progression, attrition evidence and unknowns—then connect findings to access and succession safeguards rather than quotas.
Ten fuller meetings can contain the same hundred people
Imagine a regional institution whose two annual meetings each grow by ten per cent. The annual report celebrates more than a thousand registrations, then compares that total with the previous year. Yet many entities attend both meetings, staff appear at every event, sponsored fellows are counted again when they return, and online users reconnect under several accounts. Across three years, the number of unique community entities may barely change.
Nothing improper has occurred. Returning is often the desired outcome. Policy work takes time, technical trust develops through repeated contact and leadership requires memory. The problem is the noun “growth.” Event appearances have increased; the stock of independent people or organisations may not have.
The illusion becomes consequential when larger totals support claims of expanding representation while authority remains concentrated. A newcomer attends orientation, observes a meeting and disappears. A smaller group authors proposals, chairs sessions, answers objections and fills elected seats. Both groups contribute to the annual attendance figure, but only one shapes decisions.
Community growth therefore requires a flow account. Who entered for the first time? Who returned? Who contributed between meetings? Who progressed into authorship or office? Who left? Which organisations gained or lost presence? How many event appearances were produced by the same people?
No single retention target defines health. A technical community needs regulars and newcomers in different proportions by function. The goal is to distinguish continuity from closure and outreach from durable access. Once the units are visible, institutions can value repeat expertise without using it to simulate a widening mandate.
Attendance is a flow; community capacity is a stock
An event count is produced during a period. It measures registrations, check-ins or another activity. Community capacity is the set of people and organisations able and willing to contribute across periods. Confusing the two is like counting every train journey as a new passenger.
The basic identity is simple. Active community stock at the end of a period equals prior active entities who remain, plus new entrants who become active, minus people who exit or become inactive. Event appearances are generated by this stock at different frequencies. Staff and highly engaged volunteers create many appearances; occasional affected parties create few.
The institution must define “active” for each claim. An active event entity might attend one meeting in two years. An active policy contributor might post, submit evidence or speak on a proposal. An active authority holder occupies a chair, council, board or voting role. These populations overlap but should not be merged.
Publish both unique people and person-events. If 700 event appearances were generated by 430 unique people, say so. Add participation frequency bands: one appearance, two, three to five, and more than five over the selected window. Separate staff and contractors because occupational attendance differs from voluntary return.
Do the same for organisations. Unique declared affiliations can reveal whether more people come from the same employers. Where publicly verifiable, related entities can be grouped at an organisational-control level, with unknowns preserved.
This stock-and-flow view prevents a good retention programme from being misrepresented as recruitment and prevents a one-off outreach spike from being described as sustainable capacity.
The newcomer definition decides whether growth appears
“Newcomer” can mean first registration, first in-person event, fewer than three meetings, first use of an orientation service or self-identification. The definition changes both numerator and retention rate.
ICANN's historical reporting shows the importance of labels. Its 2013 annual material described newcomers as attending fewer than three meetings, while a 2011 Dakar report reported 140 documented newcomers including 12 repeat visits. These programme measures are useful, but neither is automatically a count of first-ever unique entities.
A cohort system should define first verified institutional event participation as entry, with a separate field for first in-person, first remote and first substantive contribution. A person experienced at IETF but new to APNIC is institutionally new but not necessarily new to Internet governance. Both facts can matter without changing identity.
Do not reset newcomer status because an account changes, an employer changes or a person returns after a gap. Use correction and probabilistic matching rather than declaring every unmatched record new. Preserve an unknown band when identity resolution is uncertain.
The denominator for entry share is unique eligible non-staff entities in that event or year under the same verification method. The denominator for programme use is people who met the programme definition and received access. The denominator for progression is the original entry cohort with known follow-up status, while unknowns remain visible.
Stable definitions may produce less dramatic growth. They also produce evidence an institution can act on. A headline newcomer total is publicity; a cohort is a governance measure.
Official reports already show return and newcomer counts can diverge
The RIPE NCC Activity Plan and Budget 2022 presented meeting figures that separated returning entities and newcomers for virtual RIPE meetings. RIPE 81 showed 630 returning and 457 newcomer attendees; RIPE 82 showed 421 returning and 769 newcomers, alongside country distributions.
Those figures demonstrate the value of decomposing totals. Two meetings can have similar overall scale while their return and entry patterns differ sharply. They also raise necessary method questions: how was newcomer status determined, were people deduplicated across accounts, did staff sit in the return category and did “newcomer” refer to the event series or a particular platform record?
Method questions do not discredit the published figures. They define their analytical range. If the method is stable, institutions can compare meetings. If the method changes, the series should show a break. If identity matching is incomplete, a confidence range is better than treating every unmatched account as a new person.
APNIC event reports often publish people, online-only participation, economies and member organisations. Public registration lists can include names, organisations and economies with restrictions on use. These materials can support organisational and recurrence research, but public availability does not remove privacy obligations or guarantee stable identity.
The next step is not merely adding a “returning” slice to a conference graphic. It is following entry cohorts into meaningful activity and measuring whether organisational and role concentration changes. Return is valuable when it builds capacity; it is not by itself proof that the community has widened.
Identity resolution is necessary and dangerous
Counting unique people across years requires matching records. Names change, transliteration varies, accents disappear, employers become account identifiers and email addresses change. Two people may share a name. One person may create several registrations. Exact matching undercounts return; aggressive matching can merge distinct people.
A defensible method uses layered signals under strict purpose limits. A stable entity identifier, where voluntarily maintained, is strongest. Verified email history can support continuity but should be protected. Name, organisation and economy can produce candidate matches for review, not automatic public identity claims.
Publish precision and uncertainty. A “confirmed unique” count uses strong matches. A “probable duplicate” band uses bounded inference. Unresolved records remain unknown. Sensitivity analysis can show how retention changes under conservative and inclusive matching.
Do not use public attendee lists to build expansive personal histories. The objective is aggregate community health. Raw linkage should have restricted access, short retention and a correction route. People should be able to merge duplicate records or separate an erroneous match.
Staff, speakers and contractors need explicit flags because recurring occupational presence can dominate event frequency. Their expertise remains part of the institution, but it should not be reported as independent community growth. Similarly, bots, test accounts and shared viewing rooms must be excluded where identifiable.
The safest system performs pseudonymous cohort matching, then publishes only aggregate bands with minimum cell sizes. Exact individual histories are rarely necessary.
Identity resolution will never be perfect. Governance improves when the uncertainty is reported rather than converted into a confident claim that every unfamiliar row represents new participation.
Return rates need cohort denominators
A common but misleading retention measure divides returning entities at this year's meeting by this year's total. That is a composition share, not a cohort return rate. It says how much of the current room has prior experience, not what happened to people who entered earlier.
For one-year return, the denominator is unique eligible people in a defined entry cohort; the numerator is those who perform a defined return action within the next year. For three-year return, use the same cohort and window. Report deaths, explicit opt-outs and unmatchable records only under transparent rules; do not quietly remove everyone who lowers the rate.
Different return actions answer different questions. Event return measures continued contact. Mailing-list or online contribution measures substantive engagement. Proposal authorship measures production. Election candidacy and office measure progression into authority. A person may contribute deeply without attending another conference.
Use survival-style reporting where practical: the share of a cohort still active at six, twelve, twenty-four and thirty-six months. Show unknown status separately. Compare cohorts by entry mode, support status, broad region and role only where sample size and consent permit.
Do not treat higher retention as universally better. A programme designed for one-time regulator education may succeed without return. A closed professional circuit may have near-perfect retention and little renewal. State the purpose before selecting the measure.
Cohort denominators make investment review possible. If fellows return to substantive work at a higher rate than unsupported newcomers, the institution can examine which support mattered. If online entrants contribute later despite low physical return, travel should not define engagement.
Organisational growth can lag behind entity growth
More unique people do not necessarily mean more independent organisations or interests. A large registry, vendor or operator may send expanding teams. Fellowship cohorts may join the same established institutions. Consultants may appear under separate firms while serving overlapping clients.
Report unique declared organisations, verified member organisations and related organisational groups separately. Show the distribution of entities per organisation and the share associated with the largest groups. A count of one-person organisations can reveal breadth, while repeat appearances by teams reveal depth.
Affiliation is not mandate. An employee may speak personally. Organisational concentration is therefore a structural measure, not evidence of coordination or invalid speech. Entities should be able to declare active capacity for consequential interventions.
Associations complicate counting. One representative may carry a documented position developed across many members. Counting only the speaker understates the mandate; counting every member as an attendee invents presence. Report the association as one participating organisation and describe the mandate process separately.
Time matters because people change employers. Historical reports should use affiliation declared at the event. A current employer should not be applied retrospectively. Related-company mapping should be based on verifiable control at the relevant date and retain unknowns.
An institution can genuinely expand human participation while organisational diversity contracts. It can also retain a small number of people who connect many new organisations. Both patterns require interpretation beyond the headline total.
The strongest growth claim states the unit: “unique non-staff entities increased,” “first-entry organisations increased,” or “active policy contributors expanded across organisational groups.” Avoid the unqualified “community grew.”
Repeat participation is institutional capital
Regular entities carry procedural memory, technical context and social trust. They know why an earlier proposal failed, which implementation assumption proved wrong and how to frame an objection so it can be resolved. Chairs and authors often emerge from years of unpaid work.
A healthy institution should expect some cohorts to return and advance. Constant replacement would impose onboarding costs, weaken accountability and give staff disproportionate memory. Newcomers need experienced people who can explain both formal rules and practical norms.
The problem is not recurrence but dependence. If only a few people can chair, edit, summarise or negotiate, their absence threatens the process and their preferences may become embedded in routine. The institution then mistakes indispensability for broad support.
Measure knowledge distribution. How many people can perform each critical role? Are minutes, precedents and handovers documented? Do co-chair and shadow roles transfer skill? Do regulars sponsor newcomers into substantive work or merely present to them?
Recognition should not become permanent entitlement. Open calls, term review and transparent selection allow experienced people to continue where they remain effective while testing alternatives. Rotation without succession can be destructive; succession without opportunity is impossible.
Reports should frame high return neutrally. It may show a strong community core. Pair it with entry, progression and concentration. The question is whether continuity creates a wider bench or protects a narrow one.
Treating every repeat entity as suspect would drive away precisely the volunteers required for accountability. Treating repeat participation as proof of growing representation would be equally mistaken. Institutional capital is valuable when it can be shared.
Entry without progression can be ceremonial
Many institutions invest in fellowships, newcomer tracks and travel support. These programmes remove real barriers and can diversify the room. Their governance effect depends on what happens after entry.
Track progression milestones appropriate to the institution: joining a substantive list, making a first intervention, contributing evidence, reviewing text, co-authoring a proposal, serving as rapporteur, joining a selection panel, standing for office or chairing an item. No entity needs to complete all milestones.
The denominator for each progression rate is the eligible entry cohort, not current role holders. Report time to first milestone and unknown follow-up. Separate supported entities from comparison cohorts carefully; selection into support means differences are not automatically causal.
Quality matters more than titles alone. A newcomer may influence policy through operational evidence without seeking leadership. A fellowship alum may take a formal role but remain dependent on the same sponsor network. Narrative samples and anonymous interviews can explain the numbers.
Institutions should examine who receives invitations to closed preparation, drafting and panel roles. Public meetings may look open while authority accumulates in pre-meeting spaces. Publish selection routes and open calls for agenda roles.
Mentoring can help, but sponsorship should not require alignment with a mentor's views. Pair newcomers with several contacts, document procedures and provide independent complaint routes. The goal is access to work, not absorption into a faction.
An organisation that brings in hundreds of first-timers but advances almost none may still succeed at education. It should claim education, not leadership renewal. If it claims community expansion, the progression evidence should be visible.
Authority concentration is different from attendance concentration
A room may contain many repeat attendees while elected and chair roles rotate broadly. Another may have diverse attendance while the same people control agendas and closure. Attendance recurrence cannot stand in for authority concentration.
Create a role ledger by event and year. Record chairs, co-chairs, proposal authors, agenda setters, panel selectors, rapporteurs, consensus assessors, council and board seats, nominating bodies and formal advisers. Identify selection method, term and declared affiliation at the time.
Measure unique role holders, consecutive service, cumulative role count, organisational concentration and overlap among roles. Distinguish elected, appointed, invited, staff-assigned and volunteer-by-default positions. These routes confer different kinds of authority.
Some overlap is efficient. An author may present a proposal; a chair may serve on several related groups. The risk increases when one person frames, facilitates, summarises and judges the same matter. Role-switch declarations and recusal can manage this without excluding expertise.
The denominator for rotation is occupied role opportunities during the period, with vacancies and uncontested calls reported. Saying that few new people volunteered is not enough unless outreach, qualification and workload are visible. Roles may be designed in ways only incumbents can sustain.
Compare the entity pathway with the role ledger. Did first-entry cohorts eventually appear? Did new organisations reach agenda functions? Did supported entrants progress after funding ended? This reveals whether a growing audience changes governance capacity.
Authority concentration should trigger succession planning, not automatic removal. A capable regular can train a co-chair, document handover and remain available. The objective is resilience and contestability.
Speaking share reveals whether return becomes dominance
Repeat entities often speak more because they know the issues and procedures. That can improve meetings. It can also crowd out first interventions, particularly when regulars hold prepared and response roles.
Measure open-floor turns and time by recurrence band: first event, second or third, established entity and staff. Separate presentations, chair procedure, staff explanation and open discussion. Report answer and summary treatment, not only seconds.
The relevant denominator is allocable deliberative time in the session and valid contributions received, not all conference hours. Compare first-time recognition, queue delay, completion and whether questions receive substantive answers.
Do not impose equal shares. A new attendee may prefer to observe; an established expert may need to explain complex history. Use findings to improve access: first-intervention priority, grouped repeat responses, written evidence channels, newcomer briefings and explicit invitations for missing operational contexts.
Repeated silence among newcomers may indicate satisfaction, learning, language difficulty or low perceived efficacy. Anonymous follow-up can distinguish these possibilities better than inference from a transcript.
Regular contributors should not be penalised for filling an empty queue. Chairs can ask whether others wish to enter before granting repeat turns. When only regulars possess relevant evidence, the institutional response is knowledge transfer and outreach, not arbitrary silencing.
Speaking concentration becomes a governance concern when it aligns with agenda and office concentration, and when newcomers' material contributions receive poorer treatment. The combined pattern is more informative than any one duration figure.
Staff and contractors require their own series
Staff attend because it is their job. They support rooms, explain services, record decisions and often hold the deepest organisational knowledge. Contractors may run platforms, interpretation or engagement programmes. Counting them as recurring community members inflates both total attendance and retention.
Publish staff and contractor participation separately from non-staff community entities. Where staff speak substantively, distinguish factual explanation, impact analysis, facilitation and organisational advocacy. Their contributions remain valuable; the category clarifies the source of continuity.
Staff turnover belongs in institutional-capacity analysis but not in volunteer cohort retention. A growing workforce can make meetings appear to expand even if external participation is static. Conversely, stable staff may support a rapidly changing public.
Former staff and contractors need time-aware status. They may later participate independently. Do not permanently classify them by old employment. Record the role at each event and allow correction.
Sponsored travel providers and programme alumni should not be treated as staff merely because the institution funded access. Funding status can be analysed separately with privacy safeguards. The question is whether attendance was occupationally required and whether the person exercised institutional authority.
Separating series also reveals substitution. If volunteer authorship declines while staff presentation time rises, the institution can investigate workload and support. If external chairs rotate while staff continuity preserves process, the arrangement may be resilient.
The purpose is not to exclude employees from “the community.” It is to prevent payroll continuity from being presented as independent public growth and to show where institutional knowledge actually resides.
Attrition deserves as much attention as recruitment
People leave for many reasons. Their issue concludes, employment changes, travel becomes unaffordable, meetings move to hostile time zones, conduct is poor, workload is excessive or contributions appear to have no effect. Some departure is healthy and voluntary. Systematic departure from particular groups is a warning.
Offer short, optional exit follow-up to cohorts that do not return within a defined period. Ask about relevance, cost, time, language, accessibility, conduct, employer support, procedural clarity and perceived efficacy. Allow “no longer needed” and “prefer not to say.”
The denominator is reachable non-returners invited, with response rate and selection bias stated. Do not generalise a small dissatisfied sample to everyone who left. Combine surveys with programme records, accessibility reports and meeting changes.
Protect privacy. In small sectors or economies, answers can identify people. Aggregate across years or use independent handling. Do not send a repeated-contact campaign to people who simply attended once.
Exit findings should connect to remedies. Scheduling problems call for rotation and asynchronous channels. Role opacity calls for published pathways. Unanswered contributions require disposition tracking. Harassment concerns use conduct safeguards. Funding gaps call for targeted support.
Institutions often know much more about people they recruit than people they lose. That asymmetry favours celebratory entry totals over community health. A credible growth account reports entry and exit with equal care.
No institution should promise to retain everyone. It should know whether avoidable design choices repeatedly remove people whose evidence and affected interests are missing from decisions.
Organisational churn can masquerade as renewal
People change employers while remaining in the same governance circuit. If reports count affiliations, the same entity can appear as a new organisation each time. Corporate acquisitions and rebranding create additional false entry. Conversely, a new person from an established organisation may bring a genuinely different perspective.
Track person and organisation separately with time-aware links. Organisational entry means the first verified participation of that legal entity or group under the chosen definition. Human entry means the person's first verified participation. Neither implies new interests.
Related-company mapping should be conservative and published. Grouping every subsidiary can erase operational independence; treating every brand as independent can inflate breadth. Provide both legal-entity and known-control views where evidence supports them.
Consultants and associations need explicit categories. A consultancy may expose one person to many clients without authority to name them. An association may represent a stable member base while staff change. Count declared participation and mandate provenance separately.
The relevant denominator for organisational retention is prior active organisations eligible for the event or policy area, not all members unless every member had practical opportunity. Entry rate uses organisations newly observed under the stable rule.
Organisational churn can be healthy when new networks enter and old ones no longer need direct participation. It can be concerning when only representatives rotate inside the same dominant companies. Pair counts with speaking, authorship and role measures.
This two-dimensional view stops institutions from celebrating a new logo where the decision network is unchanged, while recognising real human renewal inside long-standing organisations.
Meeting frequency can manufacture growth
If an institution adds events, person-event totals rise even with a static community. Hybrid formats can create additional registrations for people who previously appeared once. Joint conferences may count the same person in two branded programmes. Longer events create more daily attendance records.
Normalise event activity. Publish unique annual people, unique organisations, event appearances per person and events offered. Compare like formats and mark major changes. A per-event average can supplement, but not replace, annual unique counts.
For longitudinal analysis, choose a stable rolling window, such as three years, that captures recurring policy cycles. Report entry and inactivity under consistent rules. A person who skips one meeting should not necessarily be classified as exited.
Online availability changes opportunity. A rise in low-intensity viewers may represent valuable broader reach even if the core remains stable. Show participation depth bands rather than excluding observation. The claim should be “audience reach expanded,” not automatically “governing community expanded.”
Calendar changes can also alter retention. Two meetings scheduled close together may reduce the apparent return interval. A cancelled year creates artificial attrition. Cohort windows should account for opportunities offered.
Budget reporting benefits from these adjustments. Cost per appearance differs from cost per unique entrant or retained contributor. Institutions can evaluate whether added events widen access, deepen existing participation or both.
The point is not to minimise totals. It is to prevent organisational activity from being mistaken for population growth. More events can be successful even when they serve the same people, if continuity is the stated purpose.
No single concentration index is enough
A largest-group share is easy to understand but misses the rest of the distribution. A Herfindahl-style index captures concentration across organisations but can be sensitive to affiliation cleaning. A Gini coefficient describes inequality in appearances or speaking time but says nothing about role or mandate. Repeat share describes recurrence but not organisational control.
Publish a small dashboard of complementary measures with plain-language interpretation. Unique people and organisations show breadth. Median and upper-range appearances show frequency. Top-one and top-five organisational shares show concentration. Entry and return show flow. Role overlap shows authority. Unknown rates show confidence.
Avoid a composite legitimacy score. Weight choices would hide value judgments and invite gaming. A community with high recurrence and wide organisational reach may be healthy; one with low recurrence and no retained knowledge may be fragile. The measures need context.
Use multi-year trends and comparable subsets. A policy meeting should not be compared directly with a training event. An all-virtual year should be marked. Small sessions require ranges and privacy suppression.
Triggers can be defined after observing a baseline. Sustained rise in top-group share may require outreach. Low cohort progression may require pathway review. High chair overlap may require succession. None automatically invalidates substantive decisions.
Independent review of identity and affiliation methods improves trust. Entities should be able to challenge classification without being asked to disclose more than necessary.
The dashboard succeeds when it produces better questions. It fails when a green arrow becomes a substitute for examining who actually shaped a consequential outcome.
A compact annual cohort report is possible
An annual report can provide a rigorous account in several pages. Begin with definitions and method changes. Report unique non-staff people, staff, organisations and person-events. Break unique people into first-entry, recent-return and established cohorts. Show one- and three-year return for prior entry cohorts.
Add substantive progression: list contribution, speaking, authorship and formal role, with cohort denominators. Show organisational and role concentration, attendance-mode transitions, broad regional distribution and unknown affiliation. Report exits through a limited voluntary survey with response rates.
Include a participation funnel for high-impact decisions rather than claiming the annual population authorised them. Link event reach, session presence, contribution and the actual decision method.
Publish privacy safeguards, identity-resolution confidence and correction procedures. Suppress small cells. Keep raw linkage separate from public attendee lists and delete it under a stated schedule.
End with actions and later verification. If remote entrants seldom progress, test mentoring or asynchronous authorship. If established entities dominate chairs, open co-chair pathways and measure the next cohort. If organisational breadth rises but speaking remains concentrated, adjust queue and agenda roles.
The report should retain positive findings. Strong return can show successful community-building. Regular contributors can be recognised for service. New entry can be celebrated where definition is stable. Honest metrics do not require a narrative of failure.
What they require is that growth refer to a specified stock, flow or capacity. The largest number on the page need not carry every claim.
Cross-institution circulation can mimic independent expansion
The same professional network often appears across IETF, ICANN, RIR, network-operator and Internet-governance meetings. A person can be a newcomer to each event series while remaining a long-established entity in the wider field. Counting institution-specific entry is still useful, but it should not be described as broad social renewal without context.
Ask entrants voluntarily about prior participation in related institutions using broad bands rather than named histories: none, occasional, regular or leadership experience. Non-response remains valid. This distinguishes institutional onboarding from first access to the wider governance environment.
Cross-institution entities provide valuable translation. They carry technical developments into policy, explain regional effects and prevent institutional isolation. Their overlap becomes concerning only when the same network occupies multiple formal channels that are later counted as independent support.
Role-aware analysis can show whether a proposal was authored, reviewed and endorsed by genuinely distinct bodies or by overlapping people acting under different mandates. Shared personnel are not proof of coordination. The record should preserve capacities and avoid counting institutions as independent constituencies where the underlying authorship is substantially common.
The denominator depends on the claim. For onboarding success, all first entities in that institution are relevant. For widening the global contributor base, entrants with no prior related experience are the narrower numerator. For cross-system capacity, experienced transfers are a positive category.
Privacy limits matter because a detailed map of affiliations can become a dossier. Publish aggregate overlap and declared formal roles only. The objective is to interpret growth, not police mobility.
Institutions should welcome experienced newcomers while making their description precise. A community can expand its institutional reach by attracting people from adjacent forums even if the wider population does not grow. That is integration, a valuable result distinct from renewal.
Success requires capacity to leave
A mature community should not require permanent attendance to preserve a person's contribution. Documents, recorded reasons, maintained code, clear precedents and trained successors allow entities to step away without institutional loss.
Measure off-ramping as well as return. Do chairs complete handovers? Can authors transfer maintenance? Are former role holders consulted without retaining informal vetoes? Can volunteers take a year away and re-enter without rebuilding status from zero?
This capacity reduces unhealthy retention pressure. People may continue because no successor exists, because reputation depends on visibility or because a process relies on undocumented knowledge. High recurrence then reflects institutional fragility rather than preference.
Succession support includes co-roles with real responsibility, concise handover records, term planning, accessible archives and recognition for maintenance. Former leaders can remain advisers under bounded roles. New leaders need room to differ rather than merely execute inherited choices.
The denominator for successful transition is roles ending during the period, with planned and emergency endings separated. Report successor readiness, vacancy duration and whether the outgoing holder continued to exercise decisive functions.
A community that allows graceful exit can welcome return without dependence. Its unique-person count may fall while capacity improves. This is another reason raw growth cannot serve as the sole objective.
Repeat participation is healthiest when it remains a choice supported by meaningful work, not an obligation created by institutional memory failure.
Healthy communities combine memory, entry and contestability
A governance community is not healthy because everyone is new. Nor is it healthy because the same experts reliably return. It needs memory to avoid repeating mistakes, entry to receive new evidence, progression to distribute capability, and contestability so authority can change hands.
Repeat entities become a problem only when recurrence is converted into presumed representation or permanent control. Newcomers become decorative when their badges enlarge totals but their contributions, return and pathways remain unknown. Organisers become misleading when they count appearances while speaking of people.
The repair is methodologically modest: stable identities under privacy limits, clear newcomer definitions, cohort denominators, organisational layers, role ledgers and honest unknowns. The harder repair is institutional: give entrants real work, give regulars a way to transfer knowledge and give affected outsiders a route into consequential decisions without requiring a professional conference career.
Growth should be stated precisely. The audience grew. The number of unique entities grew. More organisations entered. A cohort returned. The contributor bench widened. Leadership rotated. Each is valuable and testable.
When institutions can say which one occurred, retention stops looking like a deception and becomes evidence of capacity. When they cannot, a larger attendance total may merely show that a familiar community has learned to count itself more often.