Resident experience metrics are the countable parts of what a resident goes through in your building. Five of them hold up: retention, maintenance responsiveness, resident effort, active use, and communication reach. URBI records four of those five as a byproduct of running the building, not as a separate reporting exercise.
Why does nobody define resident experience?
Because the phrase sells software better as a mood than a number. It appears as an atmosphere, never as a numerator over a denominator. A building can spend two years improving resident experience and never know if it worked.
The measurement gap is documented. The National Multifamily Housing Council surveyed 40 multifamily companies for its 2023 Customer Experience Technology Report and found every respondent using a resident portal, but only 67% measuring engagement at all. The metrics in use were "simple metrics like views, clicks and time spent in the application" (NMHC, 2023). Views and clicks are page metrics. Resident experience is a person metric. So when you compare options in our resident experience app roundup, ask each vendor which numbers you will be able to produce next quarter.
What is resident experience when you have to put a number on it?
Resident experience is the sum of the interactions a resident has with the building and the people running it. A leak reported at eleven at night. A package that sat behind the desk for three days. A notice half the building says it never received. Each has a start, an end, a channel, and an outcome, which makes it countable.
That gives three measurable families:
- Outcome. Did the thing happen, and did the resident stay?
- Effort. How much work did the resident have to do?
- Adoption. Are residents using the channels you gave them, and are your messages reaching them?
Sentiment is a fourth family and the weakest. A resident who reports satisfaction and files the same complaint three times a year is giving two contradictory readings, and the behavioral one is reliable. Our post on what residents want from their building covers the qualitative side.
Which five resident experience metrics actually hold up?
These five, because each has a defensible denominator, comes out of records you already generate, and fails in a way you can name in advance. A metric that cannot state its own failure mode is a slogan.
| Metric | Definition | How to calculate | Published US benchmark | Failure mode |
|---|---|---|---|---|
| Retention or renewal | Whether residents stayed | Renewals divided by eligible expirations. In a condominium, units transferred divided by total units. | Partial. One large REIT reported 62.1% renewing in Q4 2025. No representative national norm. | Confounded by rent and job moves you do not control |
| Maintenance responsiveness | Submission to first human reply, and submission to resolution | Median and 90th percentile of each, by priority. Never averaged together. | None. The only published US clocks are legal violation deadlines. | Says nothing about whether the fix held |
| Resident effort | Steps and channels needed to complete one request | Touches and channels per completed request, plus one ease question, 1 to 7. | None for housing. The research comes from customer service. | Invisible for abandoned requests |
| Active use | Residents who took a meaningful action in the period | Active residents divided by adult occupants with an account, monthly. | None at resident level. Two thirds of surveyed operators measure engagement. | A content resident looks identical to a disengaged one |
| Communication reach | Share of the intended audience that received and saw a notice | Three per notice: reachable, delivered, seen. Denominator is the intended audience. | None. Nearly always assumed rather than measured. | Seeing a notice is not understanding it |
1. Retention: did they stay?
Retention is the outcome the other four are supposed to predict, which makes it both the most important number and the least diagnostic. By the time it drops, the cause is twelve months old.
Calculate it as renewals divided by leases eligible to renew. Say whether you exclude evictions, portfolio transfers, and month to month holdovers. Those choices move the number several points.
One published reference exists. Equity Residential reported 62.1% of residents renewing in the fourth quarter of 2025 and full year turnover of 40.2% across 73,465 same store units, against 42.6% in 2024 (Equity Residential, Q4 2025). That is one public REIT in specific markets, useful for checking your order of magnitude, not as a target.
How it gets gamed. Buying renewals with concessions and calling it experience. Counting holdovers as renewals. Dropping units where the resident left owing money.
What it fails to capture. Price. A resident can dislike the building and renew anyway because moving is expensive. Turnover costs are real and vary so much by market that any single figure would mislead.
2. Maintenance responsiveness: how fast, and fast at what?
Maintenance responsiveness is two numbers that fail in opposite ways, so collapsing them destroys the information. First response time runs from submission to the first specific human reply. Time to resolution runs from submission to the work being finished.
Fast first response with slow resolution is a capacity problem. Slow first response with fast resolution is a triage problem. Use medians, because one January flood drags a mean for a quarter, and add the 90th percentile. Track reopen rate too, the share of closed tickets that return within thirty days.
Here is the honest part. No published, transparent US benchmark exists for multifamily maintenance first response or resolution time. I checked NMHC and the Foundation for Community Association Research and found none. What exists is regulatory. New York City's Department of Housing Preservation and Development allows 90 days to correct a Class A non hazardous violation, 30 days for a Class B hazardous violation, and 24 hours for most Class C immediately hazardous violations (NYC HPD). Those are legal floors, so set targets from your own baseline. Our guide to condo maintenance request software covers getting those timestamps recorded.
How it gets gamed. Counting the automatic acknowledgement email as a first response. Starting the clock at triage instead of submission. Closing tickets that were never fixed.
What it fails to capture. Whether the resident heard anything during the gap. A leak fixed in six days with four updates beats one fixed in four days in silence, and neither clock shows it.
3. Resident effort: how hard was it to get one thing done?
Effort measures how many steps and how many channels a resident needs to complete one task, and it predicts loyalty better than delight does. Matthew Dixon, Karen Freeman, and Nick Toman studied more than 75,000 people using contact centers and self service channels, reported in 2010 that customers mostly want "a simple, quick solution to their problem," and introduced the Customer Effort Score as a better predictor of loyalty than satisfaction or Net Promoter Score (Harvard Business Review, 2010).
That transfers to buildings exactly. Nobody is delighted by a working elevator, and everyone is annoyed by having to call, then email, then walk down to the desk. Measure it two ways:
- Asked. One question at ticket close, scored 1 to 7: the building made it easy to handle my request. Report the share choosing 5 or above, on every closed ticket rather than a sample.
- Observed. Count touches and distinct channels per completed request. One in app tap is one touch, one channel. A call plus two emails plus a desk visit is four touches, three channels.
The observed number changes behavior, because it names the workflow to fix. No published effort benchmark exists for housing.
How it gets gamed. Surveying only residents whose request resolved. Counting the in app steps and ignoring the phone call that started it.
What it fails to capture. Abandonment. The resident who gave up after the second email never answers the survey, so read abandoned request counts alongside it.
4. Active use: is anyone actually using what you bought?
Active use counts residents who took a meaningful action in a period, and the definition of meaningful is where vendor numbers fall apart. An app open is not an action. Neither is a push notification landing on a locked phone.
Define an active resident as someone who submitted a request, booked an amenity, opened a notice, sent a message, made a payment, or registered a visitor. Calculate monthly active residents divided by adult occupants with an account. Track activation separately, as accounts created divided by occupants invited. High activation with low activity means you bought a directory, not a platform.
The published evidence is thin in an instructive way. In NMHC's 2023 survey of 40 companies, all respondents used resident portals and rated engagement 3.9 out of 5, but only 67% measured engagement at all (NMHC, 2023). No credible published resident level benchmark exists, and vendor adoption percentages rarely disclose the window or the denominator. We go deeper in resident app adoption.
How it gets gamed. Dividing by units instead of adult occupants, which roughly doubles the number in a building of couples. Leaving departed residents out of the denominator but their activity in the numerator.
What it fails to capture. Contentment. A resident with no maintenance issues and no interest in the party room is not disengaged. Low use in a small quiet building means something different from low use in a 400 unit tower.
5. Communication reach: did they actually see it?
Communication reach is the share of the intended audience that received and saw a notice, the metric buildings assume instead of measuring. Ask whether residents were told about the water shut off and you will hear that a notice went out. That is a send, not a reach.
Measure three numbers on every notice:
- Reachable. Residents with at least one working channel on file. This is your ceiling, and most buildings never calculate it.
- Delivered. Messages the channel accepted. Bounced email and dead numbers drop out here.
- Seen. Residents who opened the post or the email. The denominator for all three is the intended audience, not the list you happened to send to.
No published US benchmark exists for message reach, and that absence is the argument. An entire category sells communication tools and nobody publishes what good looks like. A building that cannot state its reach cannot claim it communicated. In a dispute over notice of an assessment, "we sent it" is far weaker than "82% of owners opened it." Our comparison of building communication software covers which tools record this.
How it gets gamed. Reporting email open rate alone, which image blocking distorts in both directions. Sending to the whole building so reach looks broad when the notice concerned one stack.
What it fails to capture. Comprehension. A resident can open a notice about an elevator closure and still show up on moving day with a truck.
Which resident experience metrics mislead you?
Two: net promoter score and online review scores. Neither is wrong. Both get asked to carry conclusions their method cannot bear.
What can net promoter score actually support?
It can support a trend line in one building, asked the same way at the same time each year, and little else. One recommendation question with detractors subtracted from promoters produces a number with no diagnostic content. A score of 22 does not say whether the elevators, the concierge, or the assessment increase caused it.
It also samples badly. People who answer an optional survey skew toward the angry and the delighted, so the score gets written disproportionately by the owner in a dispute with the board. A 2023 review in the International Journal of Market Research concluded that "businesses should include varied and different questions and measurements to assess customer loyalty" rather than relying on it alone (Lacohee and colleagues, 2023). Keep it if you run it, hold wording and timing steady year over year, never compare buildings with it, and pair it with questions naming specific services.
What do online review scores really measure?
They measure who was motivated enough to post, which is a different population from your residents. Nan Hu, Paul Pavlou, and Jennifer Zhang documented why ratings pile up at the extremes: only people with strong opinions bother to write, so moderate views go unreported (Hu, Pavlou, and Zhang). In a building the bias is sharper: a review is often the last act of a departing resident.
What reviews can support is text analysis. Read them, do not average them. Four people naming the same broken thing is a real signal. What they cannot support is an estimate of how your residents feel, and they should never be the number a board reports as satisfaction.
One more trap, because it appears in reputable research too. Watch for headline percentages that fold neutral responses into positive ones. The Foundation for Community Association Research reports 86% rating their community association experience positively or neutrally, and the breakdown is 63% good or very good plus 23% neutral (Foundation, 2026). The 86% is accurate. Reporting it as 86% satisfied is not.
Does any of this transfer from rentals to a condominium?
Four of the five transfer directly and one does not, and most pages on this topic pretend otherwise. Nearly all published resident experience benchmarking comes from multifamily rental, where there is a lease, a renewal date, and an operator accountable to a profit seeking owner. A condominium has none of those.
Maintenance responsiveness, effort, active use, and reach carry over unchanged, because interactions exist in both. Retention does not, because there is no renewal event. An owner does not renew. They sell, or they do not. Substitute ownership side outcomes:
- Unit turnover. Deeds transferred in a year divided by total units. Slow by nature, so read it across two or three years.
- Arrears rate. Share of units behind on common expenses. Partly financial, partly a trust signal.
- Meeting participation. Attendance and proxy return against quorum, and whether board seats were contested.
- Owner satisfaction. Specific questions about specific services, worded the same way each year.
For a national reference, the Foundation for Community Association Research commissioned Zogby Analytics to survey 3,000 residents of community associations between February 26 and March 14, 2026, with a margin of error of plus or minus 1.8 percentage points. It found 82% saying their elected board strives to serve the community's best interests and 75% saying the manager provides value (Foundation, 2026). Read it as a national floor, not a benchmark for your building: the sample mixes condominium and homeowner association members and is not broken out by type.
Where should a building that measures nothing start?
Start with maintenance, because the data already exists and the number moves fastest. Then reach, then use, then effort, and retention last. Retention goes last because it is the slowest and most confounded, not because it matters least.
| Order | What to start measuring | Time before the number means anything |
|---|---|---|
| 1 | Median first response and resolution time, by priority | About 30 closed tickets. A month in a large building, a quarter in a small one. |
| 2 | Reach on every notice: reachable, delivered, seen | Useful on the first notice. A trend after roughly six. |
| 3 | Monthly active residents against adult occupants with accounts | Three months. Month one is inflated by onboarding. |
| 4 | Effort: one ease question at close, plus touch and channel counts | Around 50 responses for the asked number. The observed number reads sooner. |
| 5 | Retention, or unit turnover and arrears in a condominium | One full lease cycle, or two to three years of transfer records. |
Two rules save wasted work. Define each metric in writing before you calculate it once, because the first argument at the board table will be about the denominator. And show owners a trend, never a single value. Our post on owner and board reporting covers presenting these.
How does URBI produce these numbers?
URBI records four of the five as a side effect of normal operations, because the interactions happen inside one system instead of four. That is the real precondition. You cannot compute effort or reach when a request starts as a phone call, becomes an email, lands in a spreadsheet, and gets answered on a bulletin board.
- Maintenance clocks are timestamps, not estimates. Service tickets run seven stages, TRIAGE, BACKLOG, TO_DO, IN_PROGRESS, IN_REVIEW, COMPLETE, and ARCHIVED, with four priority levels and a full activity log on every ticket.
- Reach is recorded per resident. News posts track who saw the post, per resident. The manager picks the audience, whole building, board only, specific floors, or a custom group, so the denominator is the intended audience by construction.
- Sentiment is continuous rather than annual. Arthur handles resident contact across voice, SMS, email, and in app chat, and every call is transcribed, summarized, sentiment scored, and tied to the resident record. That is a running read, not one survey a year.
- Effort becomes visible because the channels connect. Message threads are persistent and searchable, and inbound email routes back into the same operator thread through a per thread reply key, so replying to a building email does not open a second ticket.
- Reporting does not require an export. HERO, the manager facing AI, runs questions against the live database and produces reports on desktop web. Every action, including every HERO tool call, writes to a permanent log.
Active use is the honest exception. URBI records the actions, but the definition of active is yours to set and defend. No platform can decide whether opening a notice counts. If you are new to it, what URBI is covers the platform, and URBI for residential buildings covers how boards and managers use it.
What else do operators ask about resident experience metrics?
What are resident experience metrics, in one sentence?
They are the countable properties of the interactions between a resident and the building, grouped into outcome, effort, and adoption. In practice: retention or owner turnover, maintenance first response and resolution times, effort per completed request, monthly active residents, and communication reach per notice. Sentiment is context, not the measurement.
Is there a published benchmark for maintenance response time?
Not one that stands up. Neither NMHC nor the Foundation for Community Association Research publishes a first response or resolution benchmark, and vendor figures rarely disclose sample or method. The only published US clocks are regulatory violation deadlines, which are legal minimums, not experience targets. Set your baseline from your first quarter of data.
Should a condo board use net promoter score?
Only as a trend line, and only alongside specific service questions. A single recommendation question gives a board no idea what to fix, and it oversamples owners with strong feelings, which in a condominium usually means owners in a dispute. A 2023 review in the International Journal of Market Research recommended varied loyalty questions instead of the score alone.
Pick two of the five, define them precisely, and start next Monday. Most buildings already generate the records and never read them, far easier to fix than a real data gap. To talk through which two make sense for your building, or to see how URBI produces the maintenance clocks and reach numbers out of normal operations, write to us at hello@myurbi.co.
