On September 21, 2026, 32 of 113 sampled Phoenix-area Facebook pages had no post newer than 90 days: 28.3%. Twenty-two of those pages had no post within a year. These are observations from the resolved Facebook subsample, not an estimate for all Phoenix businesses or all social platforms.
Updated November 3: we renamed the page to say it measures Facebook pages, and we added where the businesses are, how the random draw works, which Google types each industry kept, a "small business" limit, a disclosure and a note that we emailed 68 of the businesses after the reading. We corrected the exclusion counts, which had counted two runs of the accountants search, and we corrected how we described when the search stopped and what the sitemap dates measure. No finding changed, and no new data was collected.
Updated October 30: we added the headline graphic and its PNG. No numbers changed.
Updated October 7: we clarified the denominators, limits and citation guidance. The observations are still from September 21. No new collection ran for this update.
This is the first original study on boomp.net about the market rather than about our own customers. We built a defined sample of 578 businesses across twelve industries, read each business's own website, read YouTube through its official interface, and read a fixed subsample of Facebook pages the way a reporter would: by hand, in a browser, one at a time. The weaknesses are in the Methodology and Limitations sections below.
Disclosure: we have a stake in this topic, because boomp.net makes social media software and sells done-for-you service plans. That is why the method, the limits and the aggregate data are public.
This graphic is free to use under the Creative Commons Attribution 4.0 license (CC BY 4.0), with credit to boomp.net and a link to this page. Download it as a PNG at 2160 by 2700 pixels.
Checked numbers you can cite
Each percentage uses the denominator in its own row. Do not use the 578-business list as the denominator for the Facebook result.
| Observation on September 21, 2026 | Count | Denominator | Share | What it does not show |
|---|---|---|---|---|
| Facebook: newest post more than 90 days old | 32 | 113 resolved sampled pages | 28.3% | Why posting stopped, or activity on other platforms |
| Facebook: newest post more than 365 days old | 22 | 113 resolved sampled pages | 19.5% | Whether the business was closed |
| Website: no social profile linked | 136 | 501 readable business websites | 27.1% | Whether unlinked social accounts exist |
| YouTube: newest upload more than 180 days old | 51 | 115 resolved linked channels | 44.3% | Video activity elsewhere |
Suggested citation sentence: “In a September 21, 2026 Phoenix-area sample, 32 of 113 manually resolved Facebook pages had no post newer than 90 days.” Link to this report and retain the sample size and date.
Download the four-row citation table, or inspect the complete aggregate dataset. The small table is derived from the same September data. It is not a new study.
How the denominators fit together
The list contained 578 businesses. Website reading produced 501 readable sites, 66 failed or refused reads, and 11 sites excluded from fetching by robots.txt. These three groups sum to 578.
Of the readable sites, 334 linked a Facebook page. We selected 120 linked pages for manual review: 113 resolved, six had been removed and one was unreachable. Only the 113 resolved pages enter the Facebook age percentages. Separately, 119 sites linked a YouTube channel; 115 channels resolved and four did not. A business can appear in several platform measures, so those platform totals must not be added as distinct businesses.
What we measured
- Linked social presence. For each business whose website we could read (501 of 578), how many social profiles the site links to, and which platforms.
- Website freshness. The newest date the site reports in its own sitemap, where it reports one.
- YouTube last upload. For businesses whose site links a YouTube channel (119), the date of the newest upload.
- Facebook last post. For a fixed subsample of 120 businesses whose site links a Facebook page, the date of the most recent post visible on the page. This is the headline measure. 113 pages resolved; 6 pages had been removed and 1 could not be reached.
Facebook: the headline, by industry
Each row is the pages we read for that industry. "No post in over 90 days" means the newest post was more than 90 days old on September 21, 2026.
Per-industry rows contain only 9 or 10 pages each. They describe these observed pages, not a reliable ranking of industries. One page changes a ten-page group by 10 percentage points and a nine-page group by 11. Use the combined 32-of-113 result with its sampling limits when citing the study.
| Industry | Pages read | No post in over 30 days | No post in over 90 days | No post in over 180 days |
|---|---|---|---|---|
| Accountants | 10 | 6 (60%) | 6 (60%) | 5 (50%) |
| Dentists | 9 | 4 (44%) | 4 (44%) | 3 (33%) |
| Roofers | 9 | 5 (56%) | 3 (33%) | 3 (33%) |
| Hair salons | 9 | 5 (56%) | 3 (33%) | 3 (33%) |
| Plumbers | 10 | 4 (40%) | 3 (30%) | 1 (10%) |
| Med spas and day spas | 10 | 3 (30%) | 3 (30%) | 3 (30%) |
| Chiropractors | 10 | 5 (50%) | 3 (30%) | 1 (10%) |
| HVAC | 9 | 3 (33%) | 2 (22%) | 2 (22%) |
| Law firms | 9 | 2 (22%) | 2 (22%) | 2 (22%) |
| Landscapers | 9 | 4 (44%) | 1 (11%) | 1 (11%) |
| Auto repair | 9 | 1 (11%) | 1 (11%) | 1 (11%) |
| Electricians | 10 | 2 (20%) | 1 (10%) | 1 (10%) |
| All industries | 113 | 44 (39%) | 32 (28%) | 26 (23%) |
The full distribution of the 113 pages: 69 (61%) had posted within 30 days, 12 (11%) between 31 and 90 days ago, 6 (5%) between 91 and 180 days, 4 (4%) between 181 and 365 days, and 22 (19%) more than a year ago. No page in the subsample had zero posts.
Linked social presence
Of the 501 business websites we could read, 136 (27%) link to no social profile at all. 76 (15%) link to exactly one. 289 (58%) link to two or more.
| Platform linked from the website | Businesses (of 501) | Share |
|---|---|---|
| 334 | 67% | |
| 269 | 54% | |
| YouTube | 119 | 24% |
| 104 | 21% | |
| X | 99 | 20% |
| TikTok | 58 | 12% |
| 27 | 5% | |
| Nextdoor | 9 | 2% |
A business can have a profile it never links from its site, so these are floors, not counts of accounts that exist. What they do measure is whether the business itself points customers at the profile.
Website freshness
Of the 501 readable websites, 362 had a sitemap with a usable last-modified date. Where the date exists, 317 sites (63% of all 501) reported a sitemap change date within the last twelve months and 45 (9%) reported nothing newer than a year old. The remaining 139 (28%) gave us no usable date, and we count them as unknown rather than stale.
| Most recent sitemap date | Websites (of 501) | Share |
|---|---|---|
| Within 12 months | 317 | 63% |
| Older than 12 months | 45 | 9% |
| Unknown (no usable date) | 139 | 28% |
YouTube last upload
Counts only. 119 businesses link a YouTube channel from their website and 115 resolved. 43 had uploaded within 30 days, 18 between 31 and 90 days ago, 3 between 91 and 180 days, 8 between 181 and 365 days, and 43 more than a year ago. 4 could not be resolved to a channel. Put another way, 51 of the 115 linked channels that resolved (44%) had not uploaded in six months.
What this can tell a business owner
This snapshot can help you decide what to check on your own website and profiles. Open the social links your website gives customers. Confirm the destination, current contact details and the date of the most recent post. Give someone responsibility for correcting outdated information.
The study did not measure customer reactions, bookings, sales, owner workload or the effect of posting more often. It cannot establish that a quiet page costs sales, that an industry needs a particular posting schedule, or that software solves the cause of inactivity. A business may be active on a platform we did not read.
Methodology
- Business list. Google Places API (New) Text Search, twelve queries (one per industry), restricted to a rectangle from 33.20, -112.40 (southwest) to 33.75, -111.60 (northeast) that covers Phoenix, Scottsdale, Tempe, Mesa, Chandler, Gilbert, Glendale and Peoria. Each query named Phoenix, for example "plumber in Phoenix AZ". Results were taken in Google's relevance order, page by page, until an industry had 50 included businesses, five pages of results or no further page from Google, whichever came first. We did not choose businesses by hand. By the city in each business's Google address, the list is mostly Phoenix and Tempe: 437 businesses in Phoenix, 96 in Tempe, 19 in Scottsdale, 11 in Mesa, 10 in Chandler, 2 in Glendale, 1 in Gilbert, 1 in Cave Creek, 1 in Guadalupe and none in Peoria. The 113 Facebook pages are 85 in Phoenix, 18 in Tempe, 5 in Chandler, 3 in Scottsdale, 1 in Mesa and 1 in Glendale. We do not publish quiet counts by city, because a city with one page would show one business's status.
- Industries. Plumbers, roofers, HVAC, electricians, dentists, med spas and day spas, hair salons, chiropractors, auto repair, landscapers, law firms, accountants. Each industry kept only businesses whose Google primary type is on its list: plumbers (plumber), roofers (roofing_contractor), electricians (electrician), dentists (dentist), med spas and day spas (spa, massage_spa), hair salons (hair_salon, beauty_salon, barber_shop), chiropractors (chiropractor), auto repair (car_repair), law firms (lawyer) and accountants (accounting, consultant). Google's general consultant type covers 41 of the 50 accountants, so the query, not the type, defines most of that group. Google's place type list has no HVAC or landscaping type (read November 2, 2026), so those two industries kept businesses typed general_contractor or service.
- Exclusions. Counting each business once per query, we examined 673 results and excluded 95: 43 with no website, 39 for the wrong primary type, 8 for a website shared by more than two businesses (chains), 2 with a website on a social or directory host, 2 repeats of a business then listed under another industry, and 1 with no ratings. 578 businesses were included: between 39 and 50 per industry. The type filters for HVAC, landscapers and accountants were changed after a first run of the search, and the accountants query was run again. The 2 repeats are shops typed car_repair that the first run had listed under HVAC; the final HVAC filter removed them, so neither HVAC nor auto repair kept them. Our collection log counts both accountants runs, so it shows 727 results and 149 exclusions.
- Website read. One read per business of its own website, on September 21, 2026: robots.txt first (we honor it; 11 sites disallowed us and were not read), then the homepage and sitemap, with a named research user agent, a 15 second timeout and a one-second pause between sites. We read /sitemap.xml, following redirects, and up to five child sitemaps on the same site. A sitemap kept only at another address counts as unknown. 501 of 578 responded with a page; 48 refused our user agent with a 403, which we recorded and did not work around.
- Social links. Profile links found on the homepage. Share buttons, embedded plugins and tracking pixels were not counted.
- YouTube. The YouTube Data API, through its official channel and playlist interfaces, for the newest upload on each linked channel.
- Facebook. From the 334 businesses whose website links a Facebook page, a fixed random subsample of 10 per industry (120 pages) was drawn before any page was read. The draw is a mulberry32 shuffle with seed 20260921 plus the industry's position in the Industries list above (0 for plumbers to 11 for accountants), over candidates sorted by Google place ID. The business list stays private, so readers cannot repeat the draw, but the seed and the draw rule were fixed in our code before the draw. A person opened each page in a browser on September 21, 2026 and recorded the date of the most recent post, or that the page was removed or unreachable. No script fetched or read any Facebook page. A helper script only opened each address in that person's own browser. After the reading ended on September 21, we emailed 68 of the 113 businesses whose Facebook pages we read, about this study and our software.
- Buckets. Days are counted from September 21, 2026. "0-30" means 0 to 30 days inclusive, "31-90" means 31 to 90, and so on; "365+" means more than 365 days.
- Go/no-go. Before any of this page was written, we fixed a rule for whether it would be published at all: at least 100 Facebook pages read, at least 8 for every industry reported, and either 25% or more of pages with no post in over 90 days or a 20-point spread between industries. All four conditions passed. Had they failed, the collectors and counts would have stayed in our records and nothing would have been published.
Limitations
- The Facebook number is a subsample of 113, not all 578. We read this subsample by hand. The combined rate describes the sampled pages. The small industry groups do not establish a population ranking or precise industry rates.
- Instagram, TikTok and LinkedIn were not read at all. A business could be quiet on Facebook and active on Instagram. We make no claim either way.
- Sitemap dates are self-reported and often missing. 28% of sites gave us no usable date. A site can also change without updating its sitemap, or ship a sitemap that stamps every page with today's date. Treat website freshness as the weakest measure here.
- Google's relevance order is not a random sample of all businesses. The inclusion rules require a website and ratings, and the results follow Google's relevance order within industry quotas. We did not measure the direction or size of the resulting selection bias.
- Mostly Phoenix and Tempe, one month. 533 of the 578 businesses are in Phoenix or Tempe. The results say little about other Valley cities, other regions or other seasons.
- Small industries. Nine or ten pages per industry. A single page moves an industry's rate by 10 or 11 points.
- "Small business" is a label, not a measurement. We did not check employee counts or revenue. Chains were screened only by the website rule in Exclusions, which left out the third and later business on one website, so a website can appear at most twice in the list.
- Missing website observations. The 66 failed or refused reads and 11 robots-blocked sites are excluded from website presence and freshness percentages. They remain part of the 578-business list. Missing data is not evidence of inactivity.
Get the data
The aggregate dataset is a CSV of industry, measure, bucket and count: phoenix-social-staleness-2026.csv. It contains no business names, websites or page addresses, and none will be released. You may reuse this data under the Creative Commons Attribution 4.0 license (CC BY 4.0). Please credit boomp.net and link to this page. The citation block at the end of this page has the format.
If you want help creating posts, boomp.net Packs is $9 a month for ten finished posts you publish yourself. Core at $39 a month adds connected publishing, and you approve each post before it goes live unless you turn on auto approval. The done-for-you service at $275 a month adds founder review. None of these options changes what this study can prove.