Our site: https://vps-327b7201.vps.ovh.us/DOUGLAS/v1r2/. Round 1, 2026-10-07.
Method. Same rubric for every site; one fresh Codex gpt-5.6-terra (medium) session per site, shown only that site's phone (390×844) and desktop (1440×900) first screens. No names, no history. Scores 1–10, 5 = category average. A single judge per site carries roughly ±0.2 noise. Cookie banners count, as a real visitor sees them. Excluded: ServiceMaster Clean Eugene and Compass (Phoenix) both served bot-check pages, so there was nothing to judge. No national franchise has a Roseburg janitorial office (research/douglas-local.md).
Read this honestly. We rank #1 of 18 at 8.0 because nobody local shows a price. Trust is our weakest score (5.8): we show no reviews, license, insurance or years in business because none are on file. That gap closes only with real proof from the operator.
| # | Site | Overall | Craft | Instant price | Mobile conv. | Trust | Local |
|---|---|---|---|---|---|---|---|
| 1 | umpquaOURS (v1r2) | 8.0 | 8.2 | 9.2 | 8.4 | 5.8 | 8.3 |
| 2 | hellamaidBenchmark, Canada | 7.5 | 7.4 | 8.2 | 7.1 | 7.3 | 8.4 |
| 3 | otteraBenchmark, Chicago | 7.0 | 8.1 | 1.0 | 7.6 | 6.2 | 5.8 |
| 4 | cleanerofficesBenchmark, Montreal | 6.8 | 7.6 | 7.2 | 5.8 | 6.7 | 7.1 |
| 5 | infinityRoseburg local | 6.3 | 7.2 | 1.0 | 6.4 | 4.6 | 7.0 |
| 6 | millenniumBenchmark, Atlanta | 6.3 | 7.6 | 1.0 | 6.8 | 5.9 | 6.7 |
| 7 | nwqualityRoseburg local | 6.1 | 7.8 | 1.0 | 7.1 | 3.5 | 5.8 |
| 8 | xiriBenchmark, New York | 6.1 | 7.8 | 3.5 | 6.2 | 6.8 | 2.3 |
| 9 | system4Benchmark, US franchise | 5.9 | 6.8 | 1.0 | 6.2 | 5.9 | 6.4 |
| 10 | janiking-eugeneFranchise (Eugene) | 4.8 | 6.4 | 1.0 | 3.6 | 4.8 | 5.7 |
| 11 | buildingstarsBenchmark, US franchise | 4.8 | 6.2 | 1.0 | 6.1 | 4.5 | 2.8 |
| 12 | cleaningroseburgRoseburg local | 4.6 | 5.4 | 1.0 | 4.8 | 2.6 | 7.8 |
| 13 | janpro-portlandFranchise (Portland) | 4.3 | 4.8 | 1.2 | 4.1 | 5.3 | 5.6 |
| 14 | mastercareRoseburg local | 4.1 | 4.8 | 1.0 | 4.2 | 4.6 | 3.2 |
| 15 | citywide-portlandFranchise (Portland) | 4.1 | 5.8 | 1.0 | 3.2 | 2.8 | 6.4 |
| 16 | commercialcleaningcorpBenchmark, New Jersey | 3.8 | 5.2 | 1.0 | 2.8 | 3.1 | 4.2 |
| 17 | rosecleaningEugene, serves Roseburg | 3.7 | 3.8 | 1.0 | 4.2 | 2.0 | 5.1 |
| 18 | coverallFranchise (Oregon) | 3.1 | 4.8 | 1.0 | 2.1 | 3.2 | 2.4 |
OURS (v1r2) · https://vps-327b7201.vps.ovh.us/DOUGLAS/v1r2/
Best: The prominent no-contact price estimator makes the value proposition exceptionally immediate and actionable.
Worst: Visible trust proof is thin—there are no reviews, credentials, guarantees, or human evidence to reassure a cautious commercial buyer.
Benchmark, Canada · https://hellamaid.ca/
Best: The prominent above-the-fold estimate form paired with an unusually memorable Canadian cityscape makes the service feel immediately local and actionable.
Worst: The generic residential “house cleaning” positioning and contact-heavy form weaken relevance and speed for a cold B2B janitorial buyer.
Benchmark, Chicago · https://otteracleaning.com/
Best: The striking, highly legible hero and prominent dual CTAs make the value proposition feel premium and immediately actionable.
Worst: There is no visible real pricing or instant-estimate path, while trust relies heavily on a small review count and unsubstantiated claims.
Benchmark, Montreal · https://www.cleaneroffices.ca/
Best: The desktop hero is polished, clearly positions commercial cleaning in Greater Montreal, and foregrounds a cost-calculator CTA.
Worst: On mobile, the oversized copy pushes the primary CTA below the fold and the dense hero makes the next step feel less immediate.
Roseburg local · https://www.infinityprofessionalcleaning.com/
Best: Strong local naming, polished typography, and prominent quote paths create a clear professional first impression.
Worst: There is no visible pricing path and the trust claims are generic, so a cold B2B buyer has little proof before committing to contact.
Benchmark, Atlanta · https://millfac.com/
Best: A distinctive, polished industrial hero and strong headline communicate a professional facilities-services brand quickly.
Worst: There is no visible way to get even a starting estimate, while the trust claims are largely unsupported by concrete proof.
Roseburg local · https://www.nwqualitycleaning.com/
Best: The autonomous-cleaning-robot hero image and clear quote CTA make the service feel modern and immediately legible.
Worst: There is no visible proof or pricing path—only broad claims—so a cautious facilities buyer has little reason to trust or act now.
Benchmark, New York · https://xiri.ai/
Best: The desktop hero is polished, clear, and makes the verification-led cleaning proposition feel more credible than a typical janitorial homepage.
Worst: It feels like a generic national facilities brand, with no visible local proof and no immediate way to get a real estimate.
Benchmark, US franchise · https://www.system4ips.com/
Best: The prominent real team photo and clear regional service-area headline make the business immediately understandable and more credible than a stock-led competitor.
Worst: The first screen asks for personal contact details before offering any pricing guidance, while intrusive cookie and chat overlays dilute focus and mobile conversion clarity.
Franchise (Eugene) · https://www.janiking.com/commercial-cleaning/eugene-salem
Best: The desktop hero clearly communicates commercial cleaning in Eugene/Salem with a prominent quote CTA and credible service imagery.
Worst: The intrusive cookie modal overwhelms the mobile screen and blocks the primary conversion path before visitors can act.
Benchmark, US franchise · https://buildingstars.com/
Best: The large, clear quote CTA and professional cleaning imagery make the core service immediately understandable.
Worst: It offers no visible pricing path or meaningful local and credibility proof for a cold B2B buyer.
Roseburg local · https://cleaningroseburgoregon.com/
Best: The Roseburg-specific imagery and location-forward headline establish unusually clear local relevance immediately.
Worst: The phone view wastes most of the first screen on an oversized header and generic quote CTA while offering no verifiable trust or price path.
Franchise (Portland) · https://jan-pro.com/portland
Best: Clear regional coverage and a prominent real-person cleaning visual make the service immediately legible.
Worst: The above-the-fold experience is dated and CTA-fragmented, with no clear low-friction path to an actual estimate.
Roseburg local · https://mastercare-cleaning.com/
Best: The page communicates the company name, service category, and phone contact clearly with a cohesive restrained palette.
Worst: It lacks an immediate quote path and compelling proof, making it feel generic and low-converting for a cold B2B buyer.
Franchise (Portland) · https://gocitywide.com/portland/
Best: The Portland-specific hero image and clear service-category grid establish a professional local facilities-services identity on desktop.
Worst: There is no visible quote or estimate path, while the large mobile cookie modal obscures the page and creates immediate conversion friction.
Benchmark, New Jersey · https://commercialcleaningcorp.com/
Best: The desktop hero quickly communicates commercial-cleaning services with clear imagery and a prominent scheduling CTA.
Worst: There is no visible estimate path or trust proof, and the intrusive privacy modal plus inconsistent phone numbers create immediate conversion friction.
Eugene, serves Roseburg · https://cleanrcc.com/
Best: The large phone CTA and Eugene/Salem/Portland mention create a clear service-area signal.
Worst: It feels dated and generic, offers no visible trust proof or pricing path, and makes mobile visitors work through a long hero before the form.
Franchise (Oregon) · https://www.coverall.com/locations/oregon/
Best: The large team image and clear commercial-cleaning headline establish the service category quickly on desktop.
Worst: The mobile screen offers no visible quote path and is heavily obstructed by a privacy modal, creating major friction for a cold buyer.
The service-area map is built from US Census 2024 county and place boundaries, 2020 ZCTAs and OpenStreetMap I-5 and Umpqua rivers, projected with d3-geo (Oregon Lambert parallels). Below: OSM tiles at identical bounds, our layers re-projected to Web Mercator, and the overlay.
Each pin uses the Census internal point (or USGS GNIS for Wolf Creek). Tested: inside its own Census place polygon, and inside the ZIP (ZCTA) listed for that town.
| Town | Zone | Lat, lon | Census place | In place | ZCTA | ZIP matches |
|---|---|---|---|---|---|---|
| Roseburg | core | 43.2236, -123.3520 | Roseburg city | ✔ | 97470 | ✔ |
| Green | core | 43.1514, -123.3874 | Green CDP | ✔ | 97471 | ✔ |
| Winston | core | 43.1201, -123.4245 | Winston city | ✔ | 97496 | ✔ |
| Sutherlin | county | 43.3862, -123.3227 | Sutherlin city | ✔ | 97479 | ✔ |
| Oakland | county | 43.4230, -123.2972 | Oakland city | ✔ | 97462 | ✔ |
| Glide | county | 43.2919, -123.0736 | Glide CDP | ✔ | 97443 | ✔ |
| Myrtle Creek | county | 43.0241, -123.2784 | Myrtle Creek city | ✔ | 97457 | ✔ |
| Riddle | county | 42.9542, -123.3645 | Riddle city | ✔ | 97469 | ✔ |
| Canyonville | county | 42.9273, -123.2789 | Canyonville city | ✔ | 97417 | ✔ |
| Drain | county | 43.6623, -123.3143 | Drain city | ✔ | 97435 | ✔ |
| Yoncalla | county | 43.6002, -123.2924 | Yoncalla city | ✔ | 97499 | ✔ |
| Elkton | county | 43.6374, -123.5691 | Elkton city | ✔ | 97436 | ✔ |
| Reedsport | county | 43.6989, -124.1119 | Reedsport city | ✔ | 97467 | ✔ |
| Winchester Bay | county | 43.6700, -124.1869 | Winchester Bay CDP | ✔ | 97467 | ✔ |
| Cottage Grove | extended | 43.7959, -123.0572 | Cottage Grove city | ✔ | 97424 | ✔ |
| Wolf Creek | extended | 42.6954, -123.3953 | GNIS point (no Census place) | n/a | 97497 | ✔ |