Industry: Senior living

Read senior living demand over months, not weeks, across every community

Families weigh a move for a long time, adult children do much of the reading, and one community's monthly inquiry count is small enough to swing on luck. Here is which window answers which question, community by community.

Continue with Google

Free for 2 sites · No credit card

Or have your agent set it up

Set up DuckDive for my website. Follow https://analytics.duckdive.ai/agents.md

Updated 16 min read

The short answer

Match each question to a fixed window: a week to catch a broken form, 90 days for pace, 12 months for trend. Read memory care, assisted living and independent living pages as separate signals, and judge cost articles by the inquiries they start, not their rate. DuckDive lets your AI agent do this for every community.

  • Before calling a quarter soft, check whether it falls outside the normal monthly spread.
  • Compare assisted living pages with assisted living pages, never whole sites with different care mixes.
  • Through your agent, 12 months is the longest window, so same-month-last-year needs your own records.

Why a 30-day read misleads in senior living

Most senior living marketing analytics advice assumes a 30-day view, yet few families move a parent after one month of searching. The need builds first. In A Place for Mom's 2025 survey of 1,104 family caregivers, 73% said a gradual need prompted their search rather than a single event, and 44% of caregivers who had not started looking thought a care decision was more than a year away.

Then the active search moves fast: more than two-thirds of families who secured care searched for 60 days or less. So your sites serve two crowds, people reading months ahead and people booking within weeks. A 30-day window shows a thin slice of the first and often half of one family's search.

Three more things bend a short read for an operator:

  • Small counts. One community's monthly inquiries are a small number, and small numbers swing on chance.
  • Two people, one decision. The report says adult daughters and daughters-in-law most often lead the search, yet 84% of caregivers in another A Place for Mom study agreed the senior remains the most important voice. The person who reads your pricing page at lunch is often not the one who books the visit, so GA4 can count them as strangers.
  • Triggers you never see. A fall or a hospital stay can move a family from browsing to touring in days. Those bursts and lulls even out over a quarter, not over a month.

Short windows still have a job. Pair each question with a window long enough to answer it.

Choosing the window for each question

DuckDive offers your agent six preset spans, the shortest seven days and the longest twelve months, with 14, 30 and 90 days and six months between. You cannot set your own start date, and nothing reaches past the twelve-month mark. Matched to the question, those presets settle almost every debate a sales and marketing lead has about one community or the whole group.

Which fixed window answers which senior living question
Your questionWindow to ask forWhy that window
Did a visit form or chat widget break?Last 7 days, beside the last 14A broken form falls to zero fast
Is a community's inquiry pace slipping?Last 90 days, beside the 12-month monthly averageNinety days holds enough inquiries to mean something
Is the last quarter really lower?Last 12 months, grouped by monthYou need normal months to judge the latest three
Did memory care search interest move?Last 6 months, beside the last 12Subtracting gives the earlier 6 months, page by page
Which articles start inquiries?Last 6 monthsResearch pages need volume before their rates settle
How does October compare with last October?Not available through the agentA year is the ceiling, so two Octobers never share a window

The second row is the one most weekly meetings need.

Ask your agent
Before Thursday's ops review: for every community site, give me GA4 conversions per month over the last 90 days and per month over the last 12 months. Flag any community whose 90-day pace is more than a quarter below its 12-month pace.

Tools it calls list_websites get_ga4_overview

Explorations forget

Checking long trends in GA4 explorations? Google's data retention help page says the retention setting only affects explorations and funnel reports, and standard properties keep event data for 2 or 14 months. On a 2-month setting, a 12-month exploration comes back short.

Worked example: memory care inquiries over twelve months

Bellmere Gardens is a stand-alone memory care community. Its sales director thinks the last quarter was soft and wants to know before asking for more budget.

Ask your agent
For Bellmere Gardens, show monthly GA4 conversions for the last 12 months and tell me whether the last quarter is really lower or just noisy.

Tools it calls list_websites get_ga4_overview get_ga4_landing_pages

EXAMPLE DATAFictional business: Bellmere Gardens

Use the full calendar months only. Treat December to July as normal and tell me if August to October sits outside that.

  1. list_websitesfinds Bellmere Gardens among the community sites
  2. get_ga4_overview12 months of daily sessions and conversions, summed into calendar months
  3. get_ga4_landing_pagesthe last 90 days, to see which pages the inquiries started on

Not clearly lower yet. December to July averaged about 39 conversions a month. A rough rule for counts says an ordinary month lands within two square roots of its average, so roughly 27 to 52 here. Each of August, September and October stays inside it.

MonthSessionsConversions
December 20252,31041
January 20262,48044
February2,19036
March2,35039
April2,42042
May2,26035
June2,30040
July2,19037
August2,28033
September2,21031
October2,33034
Normal band27 to 52

Taken together, the quarter looks weaker: 98 conversions against about 118 at the old pace, near the bottom of what chance allows for three months. Sessions held at roughly 2,300 a month all year, so families kept arriving; the conversion rate slipped from 1.70% to 1.44%.

My read: not proven, but worth one check this week. Test the visit request form on a phone and look at which pages the last 90 days of inquiries began on. If November and December stay this low, treat it as real.

Google Analytics 4

Two cautions. The band assumes inquiries arrive independently, which fails when an open house or a mailer bunches them into one week. And it works on counts, not rates: test conversions, then use the rate as a hint. Ask the agent to show its arithmetic, and spot-check a month using our guide to checking an AI agent's marketing numbers.

Put your own soft quarter to the test

Link one community's GA4 property, then ask whether its last three months are a real dip or an ordinary wobble.

Continue with Google

Free for 2 sites · No credit card

Care-level pages as separate demand signals

Independent living, assisted living and memory care draw different families at different moments. Memory care inquiries usually come from a caregiver who is already stretched. The Alzheimer's Association's 2026 Alzheimer's Disease Facts and Figures estimates that 7.4 million Americans age 65 and older are living with Alzheimer's, and that nearly 13 million Americans provide unpaid care for people with Alzheimer's or other dementias. A Place for Mom's survey lists dementia symptoms, including memory loss, among the most common specific triggers for a search, at 38%.

So a site-wide number hides the line you care about. If memory care pages lose clicks while independent living pages gain, a campus total looks flat while your hardest wing to fill goes quiet.

Ask your agent
Which of our communities' top Search Console pages are memory care pages, and how did their clicks change over the last 6 months?

Tools it calls list_websites get_search_console_pages

get_search_console_pages returns a site's top pages by clicks, up to 100, for one fixed range. The agent calls it for the last 6 months and the last 12 per community, and subtracts to get the 6 months before. It spots memory care pages by address, so /memory-care/ helps and /our-neighborhoods/ does not. A page in the top 100 for only one window drops out, so ask which pages it skipped.

Totals will not match

Google's Performance report help explains that chart data is aggregated by property while the Pages table is aggregated by page, so chart totals can differ from table totals. Adding up memory care page clicks will never reconcile exactly with a site's total. Compare page groups with page groups.

Give every community site one URL folder per care level, named the same everywhere. Memory care content buried in a general "services" page cannot be split out later.

Research content versus tour requests

Most reading on a senior living site happens well before anyone asks for a visit: what does it cost, what is assisted living, how is memory care different. Cost comes first. In A Place for Mom's survey, high costs were caregivers' top concern about senior living (64%), and 70% said understanding how to finance senior care feels overwhelming. CareScout's Cost of Care Survey put the national median for assisted living at $6,200 a month in 2025 (as of November 2026), so families run the sums before they call.

Those articles need a different scorecard from your tour page.

How to judge research pages and visit pages differently
MeasureResearch pagesTour or visit page
Main jobEarn trust while a family compares optionsTurn a ready family into a booked visit
Usual readerAn adult child, early, often on a work breakA family member who has decided to look
Fair window6 to 12 months30 to 90 days
Judge it bySearch clicks and the count of inquiries startedConversion rate and count
Warning signClicks falling on cost and care-type queriesRate dropping while sessions hold

Next, settle what GA4 counts. Your agent receives one conversion number per day, channel or landing page from DuckDive, with no label saying which event produced it. If brochure downloads, pricing requests and visit requests are all key events on a community's site, an article that hands out brochures looks as strong as your tour page.

  • Make the visit or tour request a key event at every community, set up the same way on each site.
  • Leave brochure downloads and newsletter sign-ups out, or accept that your agent reads one blended number.
  • Keep family portal logins and careers clicks out entirely; they come from residents' families and job seekers.

Worked example: cost-question pages that feed tour requests

Ivystone Commons is a campus with independent living, assisted living and memory care, plus a blog for families. Its marketing manager has been asked to cut content spend.

Ask your agent
Which of Ivystone Commons' top 20 landing pages answer cost or 'what is assisted living' questions, and how do their conversion rates compare with the tour page?

Tools it calls list_websites get_ga4_landing_pages

EXAMPLE DATAFictional business: Ivystone Commons

Use the last 6 months. Put those articles beside /schedule-a-tour and tell me whether they deserve more writing time or less.

  1. list_websitesfinds the Ivystone Commons website
  2. get_ga4_landing_pagestop 20 landing pages for the last 6 months, with sessions, conversions and rate

Five of the top 20 landing pages answer cost or care-type questions. Together they drew 10,690 sessions and 172 conversions. The tour page drew 182 conversions by itself, at a rate roughly ten times higher.

Landing pageSessionsConversionsRate
/schedule-a-tour1,14018216.0%
/pricing1,920713.7%
/blog/al-or-memory-care1,310221.7%
/blog/va-benefits-for-care860141.6%
/blog/assisted-living-cost3,860461.2%
/blog/what-is-assisted-living2,740190.7%

By rate, the articles look weak. By count, they started almost as many inquiries as the tour page and reached thousands of families not yet ready to book. A landing page is the first page of a session, so a daughter who reads the cost article on Tuesday and returns on Saturday to book gets credited to whatever page she lands on second.

My read: keep writing, and fix /blog/what-is-assisted-living first. It has the second-largest audience and the lowest rate. A clear link to pricing and to a visit request near the top of the page is the usual first change.

Google Analytics 4

If Ivystone counts brochure downloads as key events, part of that 172 is brochures. When a research page ranks for months and still starts almost nothing, work through the checks for landing pages that rank without converting.

Comparing communities with different care mixes

Larkhaven Assisted Living offers one care level. Ivystone Commons offers three. Bellmere Gardens offers only memory care. Rank their whole websites against each other and you are ranking their care mix, not their marketing.

NIC MAP's third-quarter 2026 figures put independent living occupancy at 91.7% and assisted living at 89.1%, against a national senior housing rate of 90.4% (as of November 2026). Care levels fill at different speeds.

Building age matters too. NIC's lease-up research describes new properties starting low, accelerating in the second year and stabilizing by the third and fourth years after opening. Judge a second-year community against its own climb.

Ask your agent
Larkhaven Assisted Living has no memory care or independent living. Set its organic clicks over the latest half-year beside the half-year before that, and benchmark it only against the assisted living pages on our other community sites, not their whole sites.

Tools it calls list_websites get_search_console_summary get_search_console_pages

EXAMPLE DATAChange in organic search clicks, latest half-year over the prior half-year: whole site versus assisted living pages only
CommunityCare levelsWhole siteAL pages
Larkhaven Assisted LivingAL-8%-8%
Ivystone CommonsIL, AL, MC+11%-3%
Bellmere GardensMC+4%None

The whole-site column makes Larkhaven the laggard. The assisted living column narrows the gap: Ivystone's growth came from independent living, and its assisted living pages slipped too. Larkhaven's question shifts from "what are we doing wrong" to "is assisted living search softening in our markets".

Before you rank communities with different care mixes

  • Compare like pages. Assisted living pages against assisted living pages, never one campus total against another.
  • Tell the agent the mix. DuckDive does not know which care levels a community offers, so put them in the prompt.
  • Group new buildings apart. Communities still on their lease-up climb get their own peer set.
  • Divide by units when sizes differ. Paste licensed units by care level and ask for inquiries per 10 units.
  • Stay at 6 or 12 months. Care-level page groups are small, and short windows exaggerate their swings.

Owned by an investor? See how private equity firms read each portfolio company's website. Also run apartments? Read apartment community website analytics.

Three decisions to make before connecting a single community

A VP with twenty or thirty campuses rarely has time to open GA4 for each one, so a slow quarter at a smaller community can go unnoticed until occupancy reports it. DuckDive gives every community website one connection point for GA4 and Search Console, with Google Ads, Local Services Ads, Yext and Local Falcon added for campuses that use them. It stores the history it imports, so a portfolio-wide question is answered from data already on hand rather than a fresh request to Google for each campus.

  1. Agree one care-level folder pattern for every campus

    Settle on /independent-living/, /assisted-living/ and /memory-care/ across the portfolio, and redirect the old addresses. Your agent sorts pages into care levels by reading their paths, so a pattern that differs from campus to campus turns every care-level comparison into guesswork. In the same pass, confirm each site records the visit request as a key event in the same way, as described under research content above. If you change key events partway through a year, note the date: months either side of it may not be counting the same thing.

  2. Map each multi-care campus to the right number of websites

  3. Chase access for acquired communities first

    Senior living communities are bought and sold, and a campus acquired last year may still report into the seller's Google account or an old web vendor's. Get your team added as a GA4 user and a Search Console user on those properties before setup day; everything else can be connected in one session. To skip the clicking, paste the setup prompt into an agent that supports remote MCP servers. You click Continue with Google (or Allow) once, and the person running setup needs a DuckDive admin account with actions allowed.

    Paste this into your agent
    Set up DuckDive for my website. Follow https://analytics.duckdive.ai/agents.md

After that, wait. Most imports finish within minutes, but a community with years of history may need longer. A community whose GA4 history is still arriving shows thin early months that look like a collapse, which is exactly the false alarm this page is meant to prevent.

Ask your agent
Check all of our community sites and name any where the GA4 or Search Console import is failing or still running, so I know which twelve-month reads to hold off on.

Tools it calls list_websites get_sync_status

What DuckDive reads from each property: connecting Google Analytics 4 to your AI agent. For how DuckDive handles dozens of location websites in general, see the multi-location overview.

Questions to take to your CRM instead

DuckDive reads what Google, Yext and Local Falcon see. Most of a senior living sale happens after the website.

  • Move-ins, deposits and occupancy. Those live in your sales CRM and resident management software, neither of which DuckDive reads.
  • Calls from cost articles. Families often ring the community straight from a cost article. Unless the call-button tap is a key event on that site, your agent sees the article visit and not the call, and DuckDive has no call tracking source to fill the gap.
  • How ready each family is. Whether an inquiry came from someone moving a parent next month or a daughter planning for next year is something your sales counselors record after the first conversation. Web analytics cannot see it.

To trace each move-in to its first touch, a senior living CRM with lead-source tracking is the better tool. DuckDive answers the earlier question: which communities and care levels draw families online, and whether a dip is real.

Common doubts about long-cycle data

How do we read a memory care wing that opened mid-year?

Pick the shortest fixed window that still includes the opening, usually the last 6 months, and look at the /memory-care/ pages on their own. Because get_ga4_overview returns daily points, your agent can also add up sessions and conversions from opening day onward inside the 12-month range, though that sum covers the whole site rather than the wing. Judge the wing against a lease-up climb, not against wings that have been full for years.

Which months should we leave out of the twelve-month baseline?

Any month your team knows was not ordinary for that community: a visit form that broke, a site rebuild, a large open house or a mailer that bunched inquiries into one week. Name those months in the prompt and ask the agent to build the normal band from the rest. Remove more than three and the band rests on so few months that almost anything passes as normal, so wait for more data instead.

Do placement-agency referrals show up anywhere?

Not in DuckDive. A placement adviser passes a family's details to your sales team directly, so the family may never visit your website before the tour, and nothing reaches GA4. If a referred family browses your site afterwards, the visit looks like ordinary direct or organic traffic with nothing marking it as a referral. Count referrals in your CRM, and remember them when a community's website inquiries fall while its tours hold steady: the gap may be referral volume, not marketing.

Does resident or health information pass through DuckDive?

DuckDive reads marketing numbers such as sessions, conversions, clicks and page paths, and has no source for health records or your CRM. The risk sits on your own sites: a memory care inquiry form that writes a diagnosis or a resident's name into its thank-you page address sends that text to GA4 as a page path. Keep those details out of URLs, and never paste resident details into a chat; your AI provider's own policies govern what happens to a conversation.

Can a regional director who uses ChatGPT ask these questions?

Only on a ChatGPT Business or Enterprise workspace; Free and Plus plans cannot use the connector. Claude, Cursor, Claude Code and other clients that support remote MCP servers with OAuth sign-in work too. A director who would rather not set up an agent can ask Drake inside DuckDive.

Sources

  1. Senior Care Search Trends: Navigating Options in the U.S. A Place for Mom (published November 20, 2025). Read on November 16, 2026.
  2. The cost of waiting to talk about senior care. A Place for Mom (published January 21, 2026). Read on November 16, 2026.
  3. Data retention. Analytics Help, Google. Read on November 16, 2026.
  4. 2026 Alzheimer's Disease Facts and Figures. Alzheimer's Association. Read on November 16, 2026.
  5. Performance report (Search results): Overview and basic setup. Search Console Help, Google. Read on November 16, 2026.
  6. Cost of Care Report. CareScout (Cost of Care Survey, July through November 2025). Read on November 16, 2026.
  7. Conversions vs. key events in Google Analytics. Analytics Help, Google. Read on November 16, 2026.
  8. Senior Housing Occupancy Nears Record High as Majority of Markets Surpass 90%. NIC (October 1, 2026). Read on November 16, 2026.
  9. Lease-Up Trends Show First Year Critical: NIC SHARK Second Segment Report. NIC (October 24, 2024). Read on November 16, 2026.

Last reviewed . Numbers in EXAMPLE DATA blocks are fictional.

Give every community a fair twelve-month read

Connect your community websites once, then ask whether a dip is real, which care level is softening and which articles start visits.

Free for 2 sites · No credit card · Working in an agent? Let it set DuckDive up