Perth Google Business Profile Study: Methodology and Appendix Tables
How the Perth Google Business Profile study was done: the wider list, the random live sample, what each measure counts, the limits, and the full appendix tables.
This note explains how the study was done, what each figure counts, and what it cannot tell you.
Data collected 4 to 5 October 2026. The wider-list values are provider snapshots (median age 15 days). The publication date is the date the article goes live and may differ from the collection dates.
Data source
Business-listing and live Google Business Profile data were collected using DataForSEO’s Business Data API. We validated stored listing fields against live profile pulls before deciding which fields could support the published analysis. DataForSEO supplied the data. The study design, analysis and conclusions are Search Scope’s own. The data come from one provider’s snapshot and live pulls; we have not shown that the provider holds every Perth business.
1. Question and design
We asked what Greater Perth local-service businesses publicly show on their Google Business Profiles, and how visible activity (reviews and posts) varies. The study is descriptive and observational. It does not use ranking data and does not test cause and effect.
There are three groups of profiles, and they are not interchangeable:
- 4,999 address-visible profiles form the broad discoverable frame, the wider list. They are a dated snapshot, and none of the wider-list figures involve a live pull.
- 240 profiles were selected at random for the live deep dive from all the relevant profiles we found, both address-visible and service-area.
- 181 of those 240 have a visible street address. They form the headline live random cohort, and they carry the Wilson confidence intervals. The other 59 are service-area businesses, reported separately.
Only the 240 received the live deep-dive treatment. The 4,999 did not. The two datasets are kept separate throughout:
- The wider list. Every relevant profile we could discover through our searches, as a dated snapshot from a commercial business-listings data provider. We call it the full discoverable list. It is not a census, and it carries no sampling interval.
- The live random sample. 240 profiles drawn at random from all relevant profiles (address-visible and service-area) and read live, with 12 per niche. 181 have a visible street address and 59 are service-area businesses.
2. The 20 niches
Plumbing, electricians, air conditioning, pest control, locksmiths, roofing, solar installation, tree services, landscaping, reticulation, pool servicing, bathroom renovation, house cleaning, carpet cleaning, mobile car detailing, removalists, dentists, physiotherapists, painters and fence contractors. Each niche was defined by one or more Google business categories chosen before retrieval. A business in more than one niche appears in each.
For five niches (plumbing, pest control, landscaping, house cleaning, dentists) we also counted profiles whose primary category is a closely related one, from a list fixed before the final draw: drainage services, gasfitters and hot water system suppliers for plumbing; bee relocation and bird control for pest control; landscape designers, lawn care services and gardeners for landscaping; cleaners and cleaning services for house cleaning; dental clinics, cosmetic dentists, emergency dental services and paediatric dentists for dentists. All other niches count only profiles whose primary category is the niche category.
3. Geography and retrieval
We searched three overlapping circles around Perth (central, south and north), then kept profiles whose coordinates fall inside the Australian Bureau of Statistics Greater Perth boundary. Retrieval is deliberately wider than the boundary, so the boundary decides who is in.
4. Building the list
- We retrieved 8,781 unique profiles after removing duplicates, keeping every niche a business belongs to.
- We removed 93 outside the boundary, 3 excluded as Search Scope-owned near-matches and 874 permanently closed profiles, leaving 7,811 eligible profiles.
- We then kept only profiles relevant to a niche: those whose primary category is the niche category (or, for the five niches above, a listed related category). 1,374 eligible profiles were dropped because the niche category appeared only as an additional category on a business with a different primary category. That leaves 6,437 relevant profiles, of which 4,999 have a visible address and 1,438 are service-area businesses.
- Search Scope-owned, current-client and former-client businesses were excluded to reduce conflicts of interest.
- 58 relevant profiles marked temporarily closed stay in the wider list with a flag and are not eligible for the live sample.
- Profiles with a visible street address (4,999) are the geographic headline population. Profiles without one (1,438) are service-area businesses. We could not independently confirm where in Greater Perth they operate, because the data source gives coordinates that are often shared or default points. They are reported as a separate cohort and are never added to the address-visible percentages. A manual check of 20 of them, using public information only, established Greater Perth coverage for 4 and partial coverage for 3.
5. The live sample
- Selection used a fixed hash-ranked rule, with at most one location per brand per niche and each business drawn at most once. The rules were written down and locked before the final draw.
- 12 of the 240 profiles come from a small earlier test run on 4 October that passed the same checks and fell inside the final draw. The rest were read on 5 October.
- Profiles that could not be read were replaced by the next profile in the draw order. Two sampled profiles no longer resolved at read time, and four sampled positions were skipped because the business had already been used as an earlier test profile (each business is used at most once).
- For each profile we read the live record, the visible Google Posts, and the latest 50 reviews, newest first.
6. What each measure counts
- Claimed: the data marks the profile as claimed by its owner. A missing value is recorded as unknown, never as unclaimed. Unknown did not occur in the address-visible profiles.
- Description present: the profile shows any business description. We do not report description length, because the data source cuts descriptions off (we saw a maximum of 750 characters).
- Website is a social or site-builder page: the website shown is a social page or a site-builder address (for example Facebook or Wix) rather than the business’s own domain.
- Services listed: the profile lists at least one service. Service entries are counted separately for the share that carry a description or a price.
- Hours present: at least one day of opening hours is shown.
- Visible post: a post returned by the data source. The source returned at most 10 posts for any profile in our tests, so profiles with exactly 10 may have more. We never report posts per month or posting history.
- Newest visible post within N days: counted from the date we read the profile.
- Owner reply: at least one of the latest 50 reviews carries an owner reply. We do not report how fast owners reply, because the source does not give true reply dates.
- Review-count bands: 0, 1 to 9, 10 to 49, 50 to 199 and 200 or more, fixed in advance.
7. Statistics
- Wilson 95% confidence intervals appear only on the live random address-visible sample.
- The wider list has no sampling interval, because it is the full list of what we found and not a sample.
- The 47% of reviews with an owner reply counts reviews, which cluster within profiles. A standard interval would be too narrow, so none is published.
- We do not test differences between niches, bands or claimed status. Those comparisons are described, not tested.
- The unclaimed group in the live sample is 27 profiles, so its intervals are wide.
- Tables in the appendix show single-niche results only where the live sample has ten or more address-visible profiles.
8. Checking the list against live profiles
For the 240 random-sample profiles we compared the stored list with the live profile. Claimed status, description and address matched for all 240. Listed services matched for 100% of profiles whose snapshot was under 90 days old and 56% of older ones; every older snapshot was a service-area business. Photo counts matched within tolerance for 96% and review counts for 94%. Most counts were unchanged (89% of photo counts and 71% of review counts); where they differed, the live count was mostly higher. We therefore treat all list figures as dated.
A separate group of 60 profiles (40 thinner, 20 fuller) was selected on purpose and read to compare the two ends of the market. It is not part of the random sample, and it is not used for any published figure, including the check above.
9. Limits
One point in time. Posts are a partial visible window. Review analysis uses the latest 50 reviews. Businesses without a discoverable profile are absent, and we have not shown that the data provider holds every Perth business. We cannot see who manages a profile, and products, Insights and other owner-only fields were not available.
10. Data
Aggregate CSV tables accompany the article, as a zip of the eight tables and a README. They contain no business names, identifiers, review text or profile-level rows. The README in the download folder explains the denominators.
Appendix: detailed tables
The tables below come from the study’s analysis files. Numbers are unchanged; some wording has been edited for publication and internal labels removed. Cohorts are never pooled. Wider-list tables are dated snapshots of the full discoverable list, not a census, with no sampling interval. Live-sample tables use the random sample of 240 profiles, with Wilson 95% intervals on the address-visible results. Service-area businesses are reported separately because their exact Greater Perth membership could not be independently confirmed. Associations shown are descriptive only. Percentages are rounded, and the counts in brackets give the exact figure.
A. Wider list, address-visible, all niches pooled
| Measure | Share |
|---|---|
| claimed (explicit true) | 81% (4071/4999) |
| unclaimed (explicit false) | 19% (928/4999) |
| claim status unknown | 0% (0/4999) |
| description present | 69% (3444/4999) |
| website present | 80% (4014/4999) |
| website is a social page | 2% (104/4999) |
| hours present (1+ day) | 85% (4257/4999) |
| 2+ additional categories | 25% (1251/4999) |
| selected attributes returned | 68% (3388/4999) |
| review topics present | 58% (2900/4999) |
| at least 1 review | 80% (3996/4999) |
| at least 10 photos | 47% (2350/4999) |
| median photos | 8 |
| median review count | 10 |
| mean rating where rated | 4.72 (n=3996) |
Review-count bands (snapshot): 0: 20% (1003/4999); 1 to 9: 30% (1477/4999); 10 to 49: 27% (1330/4999); 50 to 199: 18% (876/4999); 200 and above: 6% (313/4999)
B. Wider list by niche, full table
| Niche | n | Claimed | Unclaimed | Description | Website | Own website | Median photos | 1+ review | Median reviews |
|---|---|---|---|---|---|---|---|---|---|
| plumbing | 546 | 82% (449/546) | 18% (97/546) | 69% (379/546) | 81% (442/546) | 79% (430/546) | 10 | 87% (475/546) | 14 |
| electricians | 687 | 82% (566/687) | 18% (121/687) | 70% (480/687) | 79% (540/687) | 76% (521/687) | 8 | 84% (579/687) | 8 |
| air conditioning | 235 | 78% (183/235) | 22% (52/235) | 64% (150/235) | 77% (181/235) | 74% (175/235) | 5 | 83% (196/235) | 9 |
| locksmiths | 49 | 92% (45/49) | 8% (4/49) | 78% (38/49) | 84% (41/49) | 82% (40/49) | 19 | 94% (46/49) | 49 |
| roofing | 209 | 76% (159/209) | 24% (50/209) | 61% (128/209) | 78% (162/209) | 74% (155/209) | 4 | 74% (154/209) | 3 |
| solar installation | 260 | 75% (196/260) | 25% (64/260) | 61% (158/260) | 74% (192/260) | 74% (192/260) | 3 | 61% (159/260) | 2 |
| tree services | 132 | 83% (110/132) | 17% (22/132) | 64% (84/132) | 78% (103/132) | 71% (94/132) | 8 | 82% (108/132) | 10.5 |
| reticulation | 21 | 100% (21/21) | 0% (0/21) | 90% (19/21) | 86% (18/21) | 86% (18/21) | 12 | 95% (20/21) | 24 |
| bathroom renovation | 77 | 91% (70/77) | 9% (7/77) | 78% (60/77) | 86% (66/77) | 84% (65/77) | 32 | 84% (65/77) | 13 |
| painters | 270 | 83% (223/270) | 17% (47/270) | 73% (196/270) | 74% (201/270) | 71% (191/270) | 13 | 83% (224/270) | 5 |
| fence contractors | 191 | 82% (157/191) | 18% (34/191) | 68% (129/191) | 82% (157/191) | 79% (151/191) | 16 | 82% (156/191) | 9 |
| pest control | 186 | 79% (147/186) | 21% (39/186) | 68% (126/186) | 90% (167/186) | 89% (165/186) | 3 | 70% (131/186) | 3 |
| landscaping | 198 | 86% (171/198) | 14% (27/198) | 71% (140/198) | 78% (155/198) | 73% (145/198) | 12 | 84% (166/198) | 6 |
| house cleaning | 204 | 88% (179/204) | 12% (25/204) | 80% (164/204) | 75% (154/204) | 74% (150/204) | 8.5 | 74% (150/204) | 6 |
| carpet cleaning | 126 | 78% (98/126) | 22% (28/126) | 60% (76/126) | 80% (101/126) | 80% (101/126) | 10 | 68% (86/126) | 7.5 |
| mobile car detailing | 107 | 88% (94/107) | 12% (13/107) | 79% (84/107) | 86% (92/107) | 78% (83/107) | 19 | 86% (92/107) | 8 |
| removalists | 192 | 72% (139/192) | 28% (53/192) | 64% (123/192) | 70% (135/192) | 70% (134/192) | 7 | 79% (152/192) | 13 |
| pool servicing | 64 | 86% (55/64) | 14% (9/64) | 67% (43/64) | 66% (42/64) | 66% (42/64) | 7 | 72% (46/64) | 4 |
| dentists | 469 | 86% (401/469) | 14% (68/469) | 75% (350/469) | 87% (409/469) | 87% (408/469) | 13 | 87% (410/469) | 54 |
| physiotherapists | 776 | 78% (608/776) | 22% (168/776) | 67% (517/776) | 85% (656/776) | 84% (650/776) | 6 | 75% (581/776) | 11 |
C. Wider list, claimed versus unclaimed
| Status | n | Description | Own website | Median photos | 1+ review | Median reviews |
|---|---|---|---|---|---|---|
| claimed | 4071 | 85% (3442/4071) | 85% (3460/4071) | 13 | 88% (3568/4071) | 17 |
| unclaimed | 928 | 0% (2/928) | 48% (450/928) | 1 | 46% (428/928) | 0 |
D. Wider list, review-count bands
| Band | n | Claimed | Description | Own website | Median photos |
|---|---|---|---|---|---|
| 0 | 1003 | 50% (503/1003) | 36% (365/1003) | 58% (584/1003) | 1 |
| 1 to 9 | 1477 | 75% (1105/1477) | 58% (863/1477) | 66% (975/1477) | 3 |
| 10 to 49 | 1330 | 96% (1278/1330) | 81% (1082/1330) | 89% (1185/1330) | 13 |
| 50 to 199 | 876 | 100% (872/876) | 94% (826/876) | 98% (855/876) | 29.5 |
| 200 and above | 313 | 100% (313/313) | 98% (308/313) | 99% (311/313) | 56 |
E. Wider list, chain sensitivity (one location per brand per niche)
| Measure | All locations | One per brand per niche |
|---|---|---|
| claimed | 81% (4071/4999) | 81% (3719/4603) |
| description present | 69% (3444/4999) | 68% (3125/4603) |
| website present | 80% (4014/4999) | 79% (3625/4603) |
| 1+ review | 80% (3996/4999) | 80% (3676/4603) |
Brands collapsed: 4999 locations to 4603 brand-niche units.
F. Wider list, services (recent-snapshot subgroup only)
Address-visible, snapshot under 90 days: n = 4999; services stored 76% (3802/4999); median stored service count among those with services 18. Every address-visible snapshot in the frame is under 90 days old, so this subgroup coincides with the address-visible set; it is still labelled a recent-snapshot subgroup and is not presented as a whole-list prevalence. The current-services prevalence comes from the live random sample (table H). Records 90 days or older (all service-area businesses): services are not used for prevalence.
G. Wider-list service-area cohort by coordinate class
| Coordinate class | n | Claimed | Description | Own website | 1+ review |
|---|---|---|---|---|---|
| unique | 517 | 100% (517/517) | 93% (479/517) | 80% (414/517) | 95% (492/517) |
| shared | 442 | 100% (442/442) | 84% (372/442) | 81% (360/442) | 97% (427/442) |
| apparent_default | 479 | 100% (479/479) | 92% (440/479) | 82% (395/479) | 94% (449/479) |
Sensitivity only (carries the caveat above): addressed plus service-area pooled, claimed 86% (5509/6437), description 74% (4735/6437).
Temporarily closed (kept in the wider list with a flag): 58 relevant profiles.
H. Live sample, address-visible, pooled
| Measure | Share (Wilson 95% CI) |
|---|---|
| claimed (explicit true) | 85% (154/181), 95% CI 79 to 90 |
| unclaimed (explicit false) | 15% (27/181), 95% CI 10 to 21 |
| description present | 70% (126/181), 95% CI 63 to 76 |
| services listed (current, live) | 80% (144/181), 95% CI 73 to 85 |
| selected attributes returned | 62% (113/181), 95% CI 55 to 69 |
| hours present (1+ day) | 90% (162/181), 95% CI 84 to 93 |
| website present | 84% (152/181), 95% CI 78 to 89 |
| website is a social page | 2% (3/181), 95% CI 1 to 5 |
| booking link present | 4% (8/181), 95% CI 2 to 8 |
| 2+ additional categories | 24% (43/181), 95% CI 18 to 30 |
| review topics present | 61% (110/181), 95% CI 54 to 68 |
| at least 1 review (live) | 83% (151/181), 95% CI 77 to 88 |
| at least 10 photos | 50% (91/181), 95% CI 43 to 57 |
| any visible post | 43% (78/181), 95% CI 36 to 50 |
| exactly 10 visible posts (potentially endpoint-censored) | 17% (30/181), 95% CI 12 to 23 |
| newest visible post within 90 days of pull | 12% (22/181), 95% CI 8 to 18 |
| newest visible post within 180 days | 16% (29/181), 95% CI 11 to 22 |
| posts with a button or link (among profiles with visible posts) | 68% (53/78), 95% CI 57 to 77 |
| at least one owner reply in fetched reviews (among profiles with reviews) | 62% (94/151), 95% CI 54 to 70 |
Descriptive (no CI): median photos 10; median live review count 10; median services among those listing services 16.5; median days populated for hours 6; visible posts, median among profiles with any 7. Owner replies on 47% (1677/3571) of fetched reviews (reviews, not profiles, are the unit, and they cluster within profiles, so no interval is shown). Description length is not reported: the endpoint truncates descriptions (observed maximum 750 characters, about half of descriptions at 495 or more), so length is endpoint-censored. Profiles listing at least one priced service: 6% (8/144), 95% CI 3 to 11 of profiles listing services. Services with a description among all listed services: 7% (266/3555). Services carrying a price: 1% (42/3555); priced at or below $1.00 (placeholder candidate): 0. Review recency and censoring (profiles with reviews, n = 151): newest review within 90 days 56% (85/151); 50-review cap reached with more on the listing 29% (44/151). Window flags: 90-day window censored for 1% (2/151), 180-day for 3% (4/151). Review counts in a window are reported only where uncensored; ‘latest 50’ is never turned into a historical rate.
I. Live sample by group (our own grouping of the 20 niches)
| Group | n | Claimed | Description | Services listed | Any visible post | 1+ review | Median photos |
|---|---|---|---|---|---|---|---|
| trades | 94 | 84% (79/94), 95% CI 75 to 90 | 65% (61/94), 95% CI 55 to 74 | 81% (76/94), 95% CI 72 to 88 | 41% (39/94), 95% CI 32 to 52 | 81% (76/94), 95% CI 72 to 88 | 9 |
| recurring services | 64 | 88% (56/64), 95% CI 77 to 94 | 75% (48/64), 95% CI 63 to 84 | 88% (56/64), 95% CI 77 to 94 | 41% (26/64), 95% CI 29 to 53 | 84% (54/64), 95% CI 74 to 91 | 12.5 |
| professional services | 23 | 83% (19/23), 95% CI 63 to 93 | 74% (17/23), 95% CI 54 to 87 | 52% (12/23), 95% CI 33 to 71 | 57% (13/23), 95% CI 37 to 74 | 91% (21/23), 95% CI 73 to 98 | 10 |
J. Live sample by niche, where ten or more address-visible profiles (indicative, wide intervals)
| Niche | n | Claimed | Services listed | Any visible post | 1+ review |
|---|---|---|---|---|---|
| electricians | 11 | 64% (7/11), 95% CI 35 to 85 | 91% (10/11), 95% CI 62 to 98 | 36% (4/11), 95% CI 15 to 65 | 73% (8/11), 95% CI 43 to 90 |
| air conditioning | 10 | 70% (7/10), 95% CI 40 to 89 | 80% (8/10), 95% CI 49 to 94 | 40% (4/10), 95% CI 17 to 69 | 90% (9/10), 95% CI 60 to 98 |
| solar installation | 11 | 91% (10/11), 95% CI 62 to 98 | 18% (2/11), 95% CI 5 to 48 | 45% (5/11), 95% CI 21 to 72 | 73% (8/11), 95% CI 43 to 90 |
| tree services | 10 | 100% (10/10), 95% CI 72 to 100 | 90% (9/10), 95% CI 60 to 98 | 50% (5/10), 95% CI 24 to 76 | 90% (9/10), 95% CI 60 to 98 |
| painters | 10 | 100% (10/10), 95% CI 72 to 100 | 100% (10/10), 95% CI 72 to 100 | 40% (4/10), 95% CI 17 to 69 | 100% (10/10), 95% CI 72 to 100 |
| fence contractors | 10 | 80% (8/10), 95% CI 49 to 94 | 80% (8/10), 95% CI 49 to 94 | 60% (6/10), 95% CI 31 to 83 | 80% (8/10), 95% CI 49 to 94 |
| house cleaning | 10 | 100% (10/10), 95% CI 72 to 100 | 100% (10/10), 95% CI 72 to 100 | 60% (6/10), 95% CI 31 to 83 | 90% (9/10), 95% CI 60 to 98 |
| mobile car detailing | 10 | 90% (9/10), 95% CI 60 to 98 | 100% (10/10), 95% CI 72 to 100 | 40% (4/10), 95% CI 17 to 69 | 70% (7/10), 95% CI 40 to 89 |
| removalists | 12 | 75% (9/12), 95% CI 47 to 91 | 83% (10/12), 95% CI 55 to 95 | 50% (6/12), 95% CI 25 to 75 | 83% (10/12), 95% CI 55 to 95 |
| dentists | 12 | 83% (10/12), 95% CI 55 to 95 | 75% (9/12), 95% CI 47 to 91 | 58% (7/12), 95% CI 32 to 81 | 100% (12/12), 95% CI 76 to 100 |
| physiotherapists | 11 | 82% (9/11), 95% CI 52 to 95 | 27% (3/11), 95% CI 10 to 57 | 55% (6/11), 95% CI 28 to 79 | 82% (9/11), 95% CI 52 to 95 |
Not shown stand-alone (fewer than 10 address-visible observations): plumbing (n=8), locksmiths (n=1), roofing (n=9), reticulation (n=8), bathroom renovation (n=6), pest control (n=8), landscaping (n=8), carpet cleaning (n=8), pool servicing (n=8).
K. Live sample, claimed versus unclaimed
| Status | n | Description | Services listed | Any visible post | 1+ review | Median photos |
|---|---|---|---|---|---|---|
| claimed | 154 | 82% (126/154), 95% CI 75 to 87 | 87% (134/154), 95% CI 81 to 91 | 51% (78/154), 95% CI 43 to 58 | 90% (138/154), 95% CI 84 to 94 | 13 |
| unclaimed | 27 | 0% (0/27), 95% CI 0 to 12 | 37% (10/27), 95% CI 22 to 56 | 0% (0/27), 95% CI 0 to 12 | 48% (13/27), 95% CI 31 to 66 | 1 |
L. Live sample, review-count bands
| Band | n | Claimed | Description | Services listed | Any visible post | Owner reply in fetched reviews |
|---|---|---|---|---|---|---|
| 0 | 30 | 53% (16/30) | 37% (11/30) | 57% (17/30) | 7% (2/30) | n/a |
| 1 to 9 | 58 | 78% (45/58) | 59% (34/58) | 69% (40/58) | 28% (16/58) | 40% (23/58) |
| 10 to 49 | 49 | 100% (49/49) | 82% (40/49) | 94% (46/49) | 49% (24/49) | 73% (36/49) |
| 50 to 199 | 31 | 100% (31/31) | 90% (28/31) | 94% (29/31) | 77% (24/31) | 74% (23/31) |
| 200 and above | 13 | 100% (13/13) | 100% (13/13) | 92% (12/13) | 92% (12/13) | 92% (12/13) |
M. Service-area businesses in the live sample
n = 59. Claimed 100% (59/59), 95% CI 94 to 100; description 97% (57/59), 95% CI 88 to 99; services listed 98% (58/59), 95% CI 91 to 100; any visible post 61% (36/59), 95% CI 48 to 72; 1+ review 90% (53/59), 95% CI 80 to 95. Sensitivity only (addressed plus service-area, carries the caveat): claimed 89% (213/240), 95% CI 84 to 92; services listed 84% (202/240), 95% CI 79 to 88; any visible post 48% (114/240), 95% CI 41 to 54.
N. Stored list versus live profile (240 random-cohort profiles)
Profiles compared: 240
| Field | Agreement |
|---|---|
| Claimed status matches | 240/240 (100%) |
| Description presence matches | 240/240 (100%) |
| Address presence matches | 240/240 (100%) |
| Services presence matches | 229/240 (95%) |
| Service count within tolerance | 228/240 (95%) |
| Photo count within tolerance | 230/240 (96%) |
| Review count within tolerance | 225/240 (94%) |
Services present same, snapshot under 90 days: 215/215; 90 days or older: 14/25 (older snapshots that are service-area: 25 of 25).
Photo counts: unchanged 213, live higher 25, live lower 2. Review counts: unchanged 170, live higher 62, live lower 8.
Back to the study
The findings are in the Perth Google Business Profile study.