Resident Panel Findings · Round 1

Ayenay: a simulated resident panel,
stress-tested against the live website.

Eight research-grounded AI resident personas evaluated ayenay.org across three rounds — 19 structured sessions in total — producing five findings, a decision matrix, and a build-ready roadmap. This report shows the results and the safeguards behind them.

PREPARED BY JAMES CLARKE MORRO BAY, CA JULY 2026 FOR: LEVI JOHNSON / AYENAY.ORG
What this is. A synthetic diagnostic: every panel finding is simulation, labeled as such, grounded in Morro Bay census data, council records, and local reporting. One observation is tagged REAL-USE OBSERVATION — a single data point from the researcher's own live session, disclosed and weighted accordingly. Findings are hypotheses with evidence behind them, not proof. No real individual is depicted.
01 · Executive summary

Two sentences carry this whole report.


The bell converts visits. The privacy answer converts votes. Every panelist who might return needs to be fetched — a topic alert, a ping, a heads-up before the decision. And every panelist who might actually cast a vote needs one plain sentence about who can see it. Two different conversions, at two different moments — and the website currently ships neither.

Beyond that headline, four things stood out across two visit waves and a moderated group round:

02 · How to read this

Every claim carries a tag. Here's what they mean.


Evidence tags — with counts

SYNTHETIC · DIRECTIONAL — the default for every panel finding. All 8 personas, all 19 sessions. Where a count appears (e.g., 6/8), it means that many of the eight personas independently produced the finding.

CROSS-CORROBORATED6 findings independently reproduced when the same website capture and equivalent persona specs were run through a second, unrelated AI system. Procedure in §03.

REAL-USE OBSERVATION1 observation, from the researcher's own live session on the site. A single disclosed data point that corroborates a panel finding; it does not outrank the panel, it grounds it.

ASSUMPTION-DEPENDENTthe return-wave findings, which assume new agenda items appear before council votes (never yet observed). Flagged inline wherever it applies.

What this can and can't tell you

Can: surface likely friction points, trust barriers, and adoption mechanics before real users hit them; rank fixes by how many distinct resident types each one unlocks; pressure-test the site against the town's most skeptical audiences on demand.

Can't: replace real residents. Summary accuracy against full staff reports was not adjudicated (source documents weren't in the capture) — accuracy verdicts here are structural, not substantive. All quantified claims ("7 of 8") count simulated personas, not people. Treat every finding as a strong, evidence-backed hypothesis.

03 · Method — the panel engine

Nineteen sessions. Six safeguards. One rule: stay honest.


The value of a synthetic panel lives or dies on its machinery. What follows is the architecture, not the blueprint — enough to judge how much weight the findings can bear. The full protocol behind it is proprietary and stays with the researcher.

Grounded personas, not invented ones

Every persona was constructed against layered real-world evidence: Morro Bay census data (median age ~54, a third of the town 65+, ~41% renter households, a working waterfront), the local record itself — council minutes, agenda archives, regional reporting on the town's live controversies — and published national research on civic participation and technology adoption, used to keep each persona's behavior inside empirically observed ranges rather than stereotype. Where the data ran out, assumptions were documented, not smuggled in.

A panel built to disagree

Eight seats, each answering one question no other seat can (§04) — and deliberately skewed skeptical, with half the panel low-trust by design, because a panel that loves the website produces nothing usable. Composition was balance-audited across age, gender, housing tenure, income, and attitude before a single session ran.

Engineered against the three ways AI panels fail

Synthetic panels have three known failure modes: sycophancy (AI's instinct to flatter), convergence (eight voices quietly collapsing into one), and grounding drift (reacting to things a real first-time visitor could never see). Each had dedicated countermeasures built into the session design, and every round was audited against all three after the fact — output that failed audit was struck. A two-persona pilot gate ran first; the full panel wasn't committed until the pilot passed clean.

Three waves, under realistic conditions

Cold first visits ran on each persona's real attention budget — some sessions capped at four distracted phone-minutes, because that's the session real residents give a new website. Personas encountered the site the way residents actually do: through the town's own word-of-mouth channels, not in a vacuum. Second visits came two weeks later in scenario time, novelty gone, measuring what actually pulls anyone back. A moderated group round closed the run — top agreements, standing divisions, and the one change that moves the most seats from "maybe" to "yes" — with minority positions preserved at full strength, never averaged away.

Built-in self-critique

Every session ended with a structured challenge that forces the persona to attack its own construction — a designed-in skepticism mechanism aimed at the panel itself. It earned its keep: several of this report's best findings came out of it, including Jess's "the ping is the product," Marisol's point that real-world abandonment is silent and never shows up in anyone's data, and Doug's observation that the launch-state problems precede the AI question entirely.

The second-system check

The same website capture and equivalent persona specifications were independently run through a second, unrelated AI system. Its outputs were audited before any comparison, and findings it produced without grounding in the capture were excluded. Only findings both systems reached independently carry the CROSS-CORROBORATED tag — six qualified. Divergences between the systems are logged, not hidden: findings unique to one system carry lower confidence until validated.

Assumptions, declared: the return wave assumed the website ingests new agendas before council votes (its stated design; never yet observed — ASSUMPTION-DEPENDENT), no product changes in the interval, and a routine agenda with no hot-button item. If any assumption fails in reality, the findings leaning on it are voided — and that dependency is flagged at every point of use in this report.

04 · The panel

Eight seats, chosen to disagree.


Each card names the seat's single question. Expand any card for how that persona was built and what it was designed to test.

SEAT 1FrankThe Council Regular · 74

Does Ayenay add anything for someone who already reads every packet?

How this seat was built
Anchors: the small, consistent cohort of meeting regulars visible in any council record; retiree demographics. Design: the panel's domain auditor — he knows what actually happened at the meeting, so he catches framing errors no one else can (e.g., routine consent items dressed as contested debates). Tested rules: does the site beat his existing agenda-email alerts; does it get procedure right; is roll-call data honestly sourced.
HIGH ENGAGEMENT
SEAT 2BarbaraThe Low-Tech Retiree · 68

Can a third of the town physically and comfortably use this?

How this seat was built
Anchors: the town's 65+ third; national data on tech adoption, scam wariness, and interface anxiety in this cohort. Design: represents the largest single demographic — attached to the town, inactive online, reachable only through channels she already uses. Tested rules: readability without zooming; fear of unlabeled taps; whether "who sees my vote" is answered before the vote button; account walls as hard stops.
ATTACHED · INACTIVE
SEAT 3JessThe Stretched Parent · 37

Does the convenience promise land for the actual target user — in 90-second phone sessions?

How this seat was built
Anchors: commuting-family demographics; time-budget research on working parents; the live school-site issue as her one real stake. Design: the persona the website was implicitly built for — if it fails her, the convenience thesis fails. Sessions were hard-capped at minutes, not attention. Tested rules: is her known live issue present; does anything ping her; does the calendar produce usable logistics; the redistribution reflex (sharing to the group text).
ATTACHED · INACTIVE
SEAT 4RayThe Commercial Fisherman · 52

Does it survive a hostile sixty-second audit — and answer "who's really behind this?"

How this seat was built
Anchors: the working waterfront; the fleet's documented history with public processes (waterfront planning, offshore wind, parking) that felt decided in advance. Design: the panel's provenance test — his first and only question is who runs it and who pays for it. Tested rules: one-tap answer to "who's behind this"; whether vote counts can be quoted as consent by people who never worked the water; delegated adoption through his household.
LOW TRUST
SEAT 5PamThe Embarcadero Merchant · 57

Does it save time for the economically motivated — pure utility, no ideology?

How this seat was built
Anchors: lease-holding waterfront businesses whose landlord is functionally the city; merchant reliance on grapevine information. Design: the cleanest utility signal on the panel — no politics, just "does this save me time on parking, leases, and the master plan." Tested rules: topic search/alerts for her four keywords; whether votes are traceable by the operator; the weekly cost of manual checking; advisory-board coverage on the calendar.
PRAGMATIC
SEAT 6DeniseThe STR Owner · 61

Does the website read as neutral ground on the town's most polarized issue?

How this seat was built
Anchors: the permitted-STR owner cohort, largely non-resident; the town's documented STR fight, litigation, and enforcement history. Design: maximum economic stake, zero franchise — the seat that stress-tests neutrality, eligibility, and brigading math before the first hot item does it publicly. Tested rules: who-built-this in one click; posted voting rules; word-by-word bias reads on any housing-adjacent summary; organized-network dynamics.
CONDITIONAL / LOW TRUST
SEAT 7DougThe Informed AI Skeptic · 45

Can the AI summaries survive someone who knows exactly how they fail?

How this seat was built
Anchors: technical-professional demographics; the town's own precedent of a trusted automated system being quietly retired after its data failed. Design: the mechanism auditor — he diffs summaries against sources, hunts generation artifacts, and grades processes, not prose. Tested rules: source links present and correct; a named reviewer and corrections process; denominator honesty on small tallies; whether "reviewed before it goes live" survives contact with the output.
LOW TRUST
SEAT 8MarisolThe Service-Workforce Renter · 31

Does "community voting" signal belonging to the 41% who rent — or read as the owners' room?

How this seat was built
Anchors: the town's ~41% renter share; service-workforce schedules and channels (group chats, not city newsletters); documented renter experience of surveys that change nothing. Design: the belonging test — added by documented exception after a balance audit found tenure and demographic gaps in the original panel. Tested rules: renter framing in housing summaries; anonymity stated as no-name-no-trace; any Spanish at all; who "resident" includes.
LOW TRUST · INACTIVE

All personas are composites; no real individual is depicted.

05 · Finding one

Six independent arrivals at the same question. SYNTHETIC · 6/8 CROSS-CORROBORATED REAL-USE OBSERVATION


Six panelists — through six different fears — converged on one question. The researcher's own live session then hit the same question unprompted. When that many roads lead to one spot, that spot is the website's fault line.

BARBARAFear of a public mistake — will neighbors see if I tap wrong?
PAMHer landlord is functionally the city — can the operator read her ballot?
MARISOLLandlord/employer traceability — anonymity is a precondition, not a preference.
RAYBeing cited as consent — "whose report do my answers end up in?"
DOUGAggregate provenance — "73% Aye of how many? Who was eligible?"
DENISEBrigading & astroturf math — with tiny tallies, any group chat is a supermajority.
"Who can see my vote?"AND ITS TWIN, FROM LIVE USE: "WHO CAN SEE MY COMMENT?"
Personal-exposure fears (4 seats) Aggregate-integrity fears (2 seats)

These are two different problems. Don't merge them.

The four personal-exposure fears are answered by a plain visibility statement. The two aggregate-integrity fears are answered by posted voting rules and honest denominators. One "we take privacy seriously" line satisfies neither. Collapsing this into "users worry about privacy" would destroy exactly the texture that makes it fixable.

What the website currently says — and where

The real-use observation — exactly what it is REAL-USE OBSERVATION

During the live capture session — a real, signed-in session conducted by the researcher — a comment was posted on the Police Accreditation item, and the researcher's contemporaneous note on the capture reads: "who can see this?" To be precise: that question was the researcher's own real-time reaction while using the site, not text the website displayed. It is one data point, from one user, who is also the report's author — disclosed as such. Its value is narrow but real: a genuine human, mid-use, asked the exact question the panel converged on, triggered by the same interface asymmetry the panel flagged. It corroborates the finding; it does not independently establish it.

"It says anonymous and it says public in the same sentence. Pick one and say it big."— BARBARA (SIMULATED), GROUP ROUND
06 · Finding two

"Tell me before it's decided" — the panel's only unanimous demand. SYNTHETIC · 8/8 CROSS-CORROBORATED


At evaluation, every feed item carried the stamp "Council voted Jun 23, 2026." Seven of eight panelists hit this independently in Round 1 — each for a different reason — and in the group round it became the panel's single 8-of-8 agreement, each seat naming it in its own currency.

"You built me a ballot box when I asked for a doorbell."— JESS (SIMULATED)

ASSUMPTION-DEPENDENT Return-wave findings assumed new agenda items now appear pre-vote. The cheapest, highest-value check available: watch the feed the day the next agenda posts, against the city's own alert timing.

07 · Finding three

Two weeks later: the website is retaining watchers, not voters. SYNTHETIC · 5/8 RETURNED ASSUMPTION-DEPENDENT


In the return wave, five of eight panelists had contact with the site. Zero touched the voting layer. Every return targeted a non-voting surface — and every return was triggered by something outside the website.

Who came back — and to what

  • Frank — daily, to the council roll-call tracker ("this is my hand-built spreadsheet, automated, with a sourcing disclaimer I'd have written myself").
  • Doug — once, deliberately, to test the fresh agenda. Caught himself checking Ayenay before the city's page out of habit: "the interface is forming a habit my verdict hasn't approved."
  • Pam — once, fetched by the merchant grapevine. Scanned for parking, found none, left. "Somebody's cousin is my notification system — the exact system I already had."
  • Denise — weekly, by rotation: her owners' network now scans every Friday agenda for STR items. The website's most reliable recurring traffic is adversarial monitoring.
  • Marisol — one two-minute visit, to the roll-call page only: "That page doesn't ask me for anything. That's why I can use it."

Who didn't — and why it's still data

  • Barbara — the link never resurfaced in her Facebook world, so the site ceased to exist. Her utility bill arrived; the connection ("that site was supposed to warn me about this") never fired.
  • Jess — no ping, silently dead, exactly on schedule. But: from her own screenshot she attended twenty minutes of a City Park town-hall Zoom — the panel's only successful conversion — and credits the screenshot, not the site. "Score the site on whether it reached me. It did, once, by accident."
  • Ray — zero returns; the household's return path runs through his wife's phone. Delegated adoption is still adoption — design for it.

The pattern: the roll-call accountability tracker — the page that asks nothing of anyone and names the powerful instead — quietly retained an auditor, a hostile skeptic's respect, and the panel's most exposure-averse member. The voting layer, the website's namesake, retained no one. That's not a failure to hide; it may be the website's actual center of gravity.

"Accountability that runs toward city hall instead of away from it — that, I'd take."— RAY (SIMULATED), THE PANEL'S HARDEST NO
08 · Finding four

The AI is not the barrier — with two auditors' asterisks. SYNTHETIC · 6/8 CROSS-CORROBORATED


Six of eight found the plain-language summaries genuinely good — for several, the most readable civic writing they'd ever been handed. "Is it AI?" was mostly a non-event. What broke trust was never the prose; it was what surrounds the prose.

The standing minority position (both auditors, preserved at full strength): readable is not verified. "The better it reads, the less anyone checks." No summary has yet been graded against a contested item — that grade is pending, and it will happen publicly whether Ayenay prepares for it or not.

09 · Finding five

"One resident, one voice" — three residents, three incompatible readings. SYNTHETIC · 3/8 IN DIRECT COLLISION


The website's central phrase is undefined, and three panelists interrogated it from mutually exclusive positions. This is not a UX bug. It's Ayenay's constitutional question, and only the operator can answer it.

Denise — maximum stake, zero franchise

Owns here, pays substantial TOT here, can't vote in city elections. "Am I the community or the revenue?" If excluded: taxed, regulated, voiceless again. If included: the first STR item becomes a scoreboard — and if owners use it, "we're the astroturf story." Either way, she forwarded the site to her network within the hour, and they now monitor it weekly. The first faction to discover an unclaimed civic platform recruits on it.

Marisol — 41% of the town, undefined welcome

"Resident" is never defined; nothing says renters count, nothing says they don't — and she's learned what that silence usually means. She also noticed something bigger: the word-of-mouth chatter about the website reproduced, person for person, the room it was meant to replace. "They built a new room and the same people found it first."

Ray — the faction outvoted by the audience

"A third retirees and a fifth second homes, voting on my livelihood — majority of whom?" His demand isn't inclusion; it's that a vote total never be quotable as "Morro Bay supports X" without visible eligibility rules. Otherwise the tool isn't neutral ground — it's "a number waiting to be weaponized in someone's staff report."

In the group round, all three agreed on exactly one thing: the rules must be posted on the door, in writing — who may vote, one-per-person or not, and what "resident" means. The answer matters less than its existence.

10 · What outperformed expectations

Four things the website is already getting right.


A diagnostic panel is built to find problems — so when a deliberately skeptical panel hands out credit, it's earned. These four positives survived contact with the panel's hardest graders.

1 · The writing genuinely works 6/8

Barbara — the panel's comprehension stress test — read an entire infrastructure item and understood it: "clearer than the letter they'd send about it." Marisol: "easier than anything the county ever mailed me." For a civic website, solved readability is a real, rare asset. The AI writing layer is doing its job.

2 · The roll-call tracker is quietly excellent 4/8 ENGAGED

It earned the domain auditor's daily habit ("my hand-built spreadsheet, automated"), forty full seconds from the panel's most hostile seat, and the only voluntary return visit from its most exposure-averse one. It asks nothing of the user and names the powerful. Strongest page on the site.

3 · The sourcing disclosures earn real respect

"Voting records are pulled automatically from meeting captions and marked unofficial until the minutes are published" — both auditors independently praised this line as exactly the right epistemics. Frank: "more sourcing discipline than most staff reports show." Direct staff-report links on detail pages passed structural checks. This transparency instinct, extended to the AI summaries themselves, is the trust play.

4 · The calendar already converted someone

The panel's single successful conversion: Jess found the City Park town halls (with Zoom links) on the calendar, screenshotted them, and attended one — the exact outcome a civic engagement site exists to produce. The loop never closed (she credits her camera roll, not the site), but the value delivery was real. The mechanism works; it just works invisibly.

11 · Decision matrix

The hard binaries, both waves. SYNTHETIC · ALL 8


PanelistCreate an account?Return next week?Trust the AI summaries?What flips them
FrankNOYES — daily, tracker onlyWITHHELDBeat his city email alert to one agenda; survive one contested-item audit
BarbaraNONOREADS FINE — waryA trusted neighbor posting "I did it, it took two minutes, it's fine"; a rate item surfaced before the letter
JessNONO"SURE" — beside the pointTopic notifications. Full stop. "The ping is the product."
RayNONONO VERDICT — never saw a harbor itemA name on the About page; harbor-item alerts routed to his wife's phone
PamNOT YETONCE — grapevine-fetchedFINE UNTIL WRONGFour keywords and a bell — "I'd pay actual money for it"
DeniseNOWEEKLY — adversarial watchDEFERRED to first STR itemPosted voting rules; an STR summary both sides recognize themselves in
DougNOYES — once per testNOA methods & corrections page with a named human; real accounts; one clean contested-item diff
MarisolNOROLL-CALL PAGE ONLY"WASN'T THE BARRIER"Ironclad anonymity in plain words; one rent item, surfaced early, framed for renters

YES · NO · CONDITIONAL / QUALIFIED — note the pattern: every YES in the return column is attached to a non-voting surface, and account creation is currently impossible regardless (sign-in is demo-only), which suppresses the entire first column and reads to skeptics as an unfinished website soliciting donations.

12 · Roadmap canvas

Findings, translated into a build queue.


Must-have — the convergence findings demand these
  1. One plain visibility statement, on the feed card, before the vote tap. Who sees votes · who sees comments · what's stored. Answers four seats and the real-use observation at once.
  2. Topic-follow notifications ("the bell"). The group round's near-unanimous #1 — seven seats move or lean on this single change.
  3. Items live before council votes, reliably, the day the agenda posts. The core promise, demonstrated.
  4. Real account creation. Retire the demo sign-in — it currently blocks every binary above.
Quick wins — high impact, low effort
  1. Explain the vote-vs-comment visibility difference right where both appear on one screen.
  2. A one-page voting-rules statement: who may vote, one-per-person, what "resident" means.
  3. Name the reviewer. "Summaries reviewed by ___" + a visible corrections process; fix the duplicate-bullet class of generation errors.
  4. Calendar completeness: advisory boards, workshops, voting deadlines, and an add-to-calendar export.
  5. Show denominators inline; suppress or contextualize percentage stats below a vote threshold.
Differentiation plays — "nice tool" → "how we do local government"
  1. Lead with the roll-call accountability tracker. Already the best page on the site — it retained the auditor, earned the skeptic's respect, and was the only page the most exposure-averse panelist would touch.
  2. A "both-sides-recognizable" standard for hot-item summaries — write the first STR/parking summary so each faction sees its own argument stated fairly. This is the legitimacy play.
  3. Share cards built for group texts — the parents' and merchants' threads are the distribution network no flyer reaches; the preview is the product there.
  4. A Spanish-language surface layer — read by the workforce seat as a respect signal more than a translation need.
Risk zones — what quietly kills adoption
  1. The first hot item is a scheduled legitimacy event. Unprepared, the website gets permanently faction-tagged that week — by whichever side is unhappier.
  2. Small-sample percentages will be quoted as "Morro Bay supports X" in someone's argument, thread, or staff report.
  3. Organized-bloc flooding — with single-digit tallies, any motivated group chat is a supermajority, and using it triggers astroturf accusations in either direction.
  4. The caught-up user's empty feed — the reward for full participation is currently an empty room until the next agenda drops.
  5. Launch-state signals — demo-only sign-in beside a donate-by-text page reads to skeptics as an unfinished website asking for money.
13 · Limitations & scope

What to keep in mind — and what this round didn't cover.


Limitations

  • All panel findings are simulated. Counts like "7 of 8" quantify persona agreement, not human behavior.
  • Summary accuracy vs. full staff reports was not adjudicated — accuracy verdicts are structural only.
  • Return-wave findings depend on the unverified pre-vote ingestion assumption ASSUMPTION-DEPENDENT — one real agenda cycle settles it.
  • The real-use observation is a single, researcher-sourced data point.
  • Findings unique to one AI system (not cross-corroborated) carry lower confidence until validated.

Not covered this round — by design

  • Competitive benchmark against the city's own site, agenda portal, and email alerts.
  • Moderation and anti-brigading mechanics (rate limits, verification, dispute handling).
  • Data-retention and privacy-compliance review (stored votes + named comments).
  • Acquisition channels beyond the word-of-mouth dynamics the panel surfaced.
  • A dedicated accessibility audit (one seat tests comfort, not WCAG conformance).

Each of these is a natural follow-on module — see §14.

14 · Where this goes next

The panel is infrastructure, not a one-off.


Everything above came from a standing capability: eight calibrated resident perspectives that can be pointed at anything Ayenay ships, changes, or writes — on demand, in hours. Two tracks from here.

Track one — validate the hypotheses

The findings are strong hypotheses; real Morro Bay residents get the last word. The validation sequence, in order of cost: (1) the one-agenda-cycle check — watch the feed the day the next agenda posts, which settles the pre-vote assumption for free; (2) put real residents on the same tasks the panel ran and score where they confirm, contradict, or surprise; (3) publish the hit/miss ledger — misses documented as loudly as hits, which is what turns this from a clever simulation into a calibrated instrument.

Track two — keep the panel running

Regression testing for trust

Every UX or copy change re-run against the eight seats before it ships. "Does the new privacy line actually satisfy Pam and Marisol?" is answerable in an afternoon, before real users ever see it.

Hot-item pre-flight

Before the first STR, parking, or harbor summary goes live, run it through Denise, Ray, and Frank. If any faction can't recognize its own argument, fix it before the town grades it publicly. This defuses the report's #1 risk.

Custom outputs on demand

The same engine renders to whatever the moment needs: a one-page brief for a council conversation, a donor-facing summary, an onboarding-copy review, a features one-pager — same evidence base, different documents.

A panel that grows with the site

Seats can be added (a second-homeowner, a city-staff perspective, a Spanish-first household), rules adjusted, and new modules bolted on — the not-covered list in §13 is the menu. The panel gets sharper every cycle it runs.

This round answered "what's confusing, what's missing, what should it do." The standing panel answers the harder question every week after launch: "before we ship this — how does Morro Bay read it?"