Angry communities rarely call before the annual meeting. AI sentiment tracking reads portal messages, appeals, and maintenance requests to surface a rolling mood score per community — so property managers catch trouble before it becomes a board recall.
A property manager overseeing thirty-five communities does not find out a community is unhappy from a phone call. They find out at the annual meeting, when attendance triples, three homeowners show up with printed complaint lists, and a motion to switch management companies gets seconded before the treasurer's report is even read. By then, the frustration has been building for months — a slow maintenance response here, a fee increase there, an unanswered violation appeal — and nobody on the management side saw the pattern because no single message looked urgent on its own.
This is the core blind spot of portfolio-scale property management. A manager handling twenty to fifty communities reads hundreds of homeowner messages a week: portal questions, violation appeals, maintenance requests, escalations. Individually, each message is routine. In aggregate, they are a leading indicator of whether a community trusts its board and its management company — and that signal is invisible to a human reading one inbox at a time.
LotWize's PMC sentiment tracking feature exists to make that signal visible. It classifies every inbound homeowner message as it arrives, computes a rolling mood score per community, and alerts the property manager when a community's tone is deteriorating — days or weeks before the deterioration shows up as a resignation letter, a board recall petition, or a lost management contract.
Most HOA and PMC software treats homeowner messages as a support queue: something arrives, someone answers it, it gets marked resolved. That workflow optimizes for closing tickets, not for noticing trends. A community that files fifteen individually reasonable maintenance complaints in a month looks, ticket by ticket, like fifteen closed tickets. It does not look like a community that is losing patience with the landscaping vendor — unless someone is tracking sentiment across all fifteen messages together.
The problem compounds with portfolio size. A single self-managed HOA board president reads every message and develops an intuitive sense of community mood over time. A property manager responsible for forty communities cannot develop that same intuition for each one — there simply is not enough attention to go around. The communities that go quiet — not because they're satisfied, but because homeowners have stopped believing complaints get addressed — are the ones nobody notices until it's a crisis: a board voting to put management out for bid, a wave of non-renewals at the annual meeting, or an ombudsman complaint that turns into a public dispute. None of these events are sudden — they're the visible tail end of a decline that was building for weeks.
LotWize classifies sentiment as messages happen, not on a weekly batch. When a homeowner sends a portal message, files a violation appeal, or submits a maintenance request, the message text is passed to Claude Haiku, which classifies it along three dimensions: sentiment (positive, neutral, negative, or hostile), topic (a short label like "parking enforcement" or "maintenance delay"), and urgency (low, medium, high).
Each night, a scheduled job aggregates the last 30 days of classified messages for every community in every active portfolio into a single sentiment score. The score is not a flat average — it uses recency-weighted exponential decay with a 15-day half-life, so a hostile message from yesterday moves the score more than one from three weeks ago. A community that had a rough month in early spring but has since calmed down won't still be flagged red in late spring just because the 30-day window technically includes the bad stretch.
The resulting score runs from -2 (hostile) to +1 (positive), and the portfolio dashboard sorts every community from most concerning to least, so the property manager's attention naturally goes to the community that needs it first — not the community that happens to be top of an alphabetical list.
| Score Range | Label | What It Signals |
|---|---|---|
| +0.5 to +1.0 | Positive | Routine, cooperative tone across messages |
| 0 to +0.5 | Neutral | Ordinary mix of requests and questions |
| -0.5 to 0 | Negative | Frustration building; worth a closer look |
| -2.0 to -0.5 | Hostile | Active anger; needs intervention this week |
Because scoring runs on real message volume, a brand-new community with no message history simply shows "no data yet" rather than a misleading default score. Sentiment tracking is additive to communities that are actively using the portal — it is not a guess.
A dashboard a manager has to remember to check is only half a solution. The value comes from the system proactively flagging the moment something changes. LotWize's nightly sentiment job evaluates two conditions for every community:
A sharp score drop. If a community's sentiment score falls by more than 30% relative to its prior reading, that's flagged as a sentiment_drop alert — a fast decline is more actionable than a slow one.
A hostile cluster. If three or more hostile-classified messages about the same topic arrive within a seven-day window, that's flagged as a sentiment_hostile_cluster warning — a cluster of angry messages on one subject (a fee increase, a delayed pool reopening) usually means a specific, addressable issue rather than generalized noise.
Both alert types generate a short AI-written summary — what happened, the likely impact, and a recommended next step — and land in the property manager's central alert inbox alongside compliance deadlines and invoice anomalies. The manager doesn't have to go looking for trouble across forty separate dashboards; the system brings the communities that need attention to the top.
A sentiment score tells a manager that a community is unhappy. It doesn't tell them what to say about it in front of the board. That's the second half of the feature: an AI-generated pre-meeting briefing that a property manager can pull up before walking into a board meeting.
Selecting a community and generating a briefing produces a structured summary built from the same 30 days of classified messages: a plain-language summary of the overall trend (improving, stable, or declining), the top complaint topics ranked by frequency, a list of specific issues the board should address, suggested talking points for the meeting, and — when warranted — a flagged list of risk areas that need attention before they escalate further.
This is meaningfully different from a manager skimming their inbox the night before a meeting and trying to remember what came up. The briefing is grounded in every classified message from the period, not the handful the manager happens to recall, and it surfaces the message count and score trend alongside the summary so the manager can see the underlying evidence, not just an opaque conclusion — the difference between "I think people are frustrated with maintenance" and "twelve homeowners raised maintenance-delay complaints in the last three weeks, and the common thread is response time on non-emergency work orders."
Privacy boundaries matter here, and they're worth stating plainly rather than glossing over. The system classifies resident-facing communications: messages sent through the LotWize portal, replies to board broadcasts, violation appeals, and maintenance requests. It does not read private homeowner-to-homeowner messages, and it does not analyze internal staff communications between property managers. The signal comes exclusively from what homeowners are telling the association through official channels — the same channel where dissatisfaction shows up first, long before it reaches a phone call to corporate or a public review.
It's also worth being direct about what the score is and isn't. A -0.4 sentiment score is not a scientific measurement of community happiness — it's a directional signal built from short-message classification, and it works best as a prioritization tool, not a verdict. The system tells a manager where to look first across a portfolio too large to review manually. It doesn't replace the judgment call a manager still has to make once they're looking.
For a single self-managed HOA, sentiment tracking is a nice-to-have — a board president who reads every message already has a feel for community mood. For a property management company running a portfolio, it's closer to a load-bearing feature: attention doesn't scale with portfolio size, but the number of homeowners generating messages does.
A PMC managing 35 communities at an average of 90 units each is responsible for roughly 3,150 households. Even a modest 2% monthly message rate produces 60+ inbound messages a week across the portfolio — more than any manager reads with equal attention every week, and more than they should have to. What they need is a system that reads all of them continuously and points to the three or four communities, out of thirty-five, that deserve a closer look this week — not replacing the manager's judgment, but pointing it at the right community before quiet frustration turns loud.
This pairs naturally with AI-generated board packet narratives, which handle the financial side of meeting preparation, and with personalized homeowner communications, which address the tone-mismatch problem that often drives sentiment down in the first place. A manager who can see a community's mood trending negative, understand why from the briefing, and send calibrated communications back to that community is working with a closed loop — not just a dashboard.
2026 Update: LotWize's homeowner sentiment tracking and pre-meeting briefings are available on the PMC portfolio tier. Explore LotWize for property managers or start a free trial to see how portfolio-wide sentiment intelligence changes what a 40-community caseload looks like week to week.
Sentiment decline is usually gradual and message-by-message invisible — by the time it shows up as a board recall or a lost contract, it's been building for weeks.
LotWize classifies every homeowner message (portal messages, appeals, maintenance requests) as it arrives, then computes a recency-weighted 30-day sentiment score per community using a 15-day half-life so recent tone matters more than old messages.
Two alert types — a 30%+ score drop and a cluster of 3+ hostile messages on the same topic within seven days — surface concerning communities automatically instead of requiring a manager to check every dashboard.
AI-generated pre-meeting briefings turn raw sentiment data into specific talking points, top complaints, and risk areas a manager can bring into a board meeting.
The system reads resident-facing communications only — never private homeowner-to-homeowner messages or internal staff chats — and it's designed as a prioritization tool, not a replacement for manager judgment.
HOA sentiment tracking is an AI feature that analyzes homeowner communications — portal messages, violation appeals, and maintenance requests — to compute a community-level mood score over time. LotWize classifies each message as positive, neutral, negative, or hostile as it arrives, then aggregates 30 days of classifications nightly into a single recency-weighted score per community, ranging from -2 (hostile) to +1 (positive).
The system looks for two patterns: a sharp drop in the rolling sentiment score (more than 30% relative to the prior reading) and clusters of hostile messages about the same topic (three or more within a seven-day window). Both tend to precede visible problems — a board recall, a management contract review — by weeks, giving managers time to intervene before frustration becomes public.
No. It only analyzes resident-facing communications sent through official channels — portal messages, broadcast replies, violation appeals, and maintenance requests. It does not read private homeowner-to-homeowner messages or internal staff communications between property managers.
It doesn't change the ceiling on individual community complexity, but it changes how a manager allocates attention across a portfolio. Instead of reading every message across 30–50 communities with equal attention, a manager can rely on alerts to surface the handful of communities with declining sentiment each week and focus review there.
A satisfaction survey is a point-in-time, opt-in snapshot that most homeowners never fill out. A sentiment score is a continuous, passive signal built from communications homeowners are already sending, so it doesn't depend on response rates and it updates every night rather than once a year.
Trouble in an HOA community rarely announces itself. Start a free LotWize trial to see how portfolio-wide sentiment tracking surfaces a struggling community before it reaches the annual meeting, or read how AI meeting intelligence closes the loop once you're in the room. For the full picture of what AI can automate for property management companies, see our guide to 15 HOA tasks you should never do manually.
LotWize handles violations, resident questions, dues reminders, and meeting packets automatically — so your board gets its time back.
More guides for HOA boards
Filing a lien is the most legally consequential step a self-managed HOA board can take against a delinquent owner. LotWize's lien tracking turns a process most boards run from memory and loose paperwork into an auditable lifecycle — eligibility, intent, recording, satisfaction, and release.
Most HOA boards aren't afraid of automation—they're afraid of losing control. LotWize's AI agent framework solves that with four configurable autonomy levels per agent, from 'suggest only' to 'run and notify me weekly.'
Most HOA boards fear AI automation because they fear losing control. LotWize's autonomy ladder lets each AI agent operate at exactly the independence level your board is comfortable with—from suggest-only to fully autonomous.