Evidence behind the claims

Milestone test results

The latest saved mobile optimization results, plus separate cache-reset and benchmark reports. Performance depends on the store’s content, apps, and testing conditions.

Recorded evidence, with an open accessibility target. The latest mobile sessions were captured after the October 25 evening optimizations. The earlier 18:11 PKT demo session is retained as history. Each method and tested revision is identified below; manual accessibility coverage remains incomplete, so WCAG 2.1 AA remains a target.

Records updated · October 26, 2026

Lighthouse results by test context

The latest controlled mobile series has five completed runs per page. The standard cache-reset product check and benchmark use their own samples and are shown separately. Tables show medians and ranges; every individual report is downloadable. Scores are out of 100.

Latest saved controlled mobile session · October 25, 2026

Purely mobile · after optimization

Five completed mobile runs per page after the localization, native account-widget, and gallery-loading optimizations. Captured October 25, 2026, 20:34–20:42 PKT.

Home mobile median: 94 / 100
Collection mobile median: 92 / 100
Product mobile median: 90 / 100

Home and collection exceed 90 in all five controlled runs. Product has a median of 90 and a range of 89–91; 90+ on every product run is not established.

Tested store: purely-bmsyummi.myshopify.com
Tested theme: Milestone (tested as Manifest) · Shopify theme 131750854738
Tools: Lighthouse 13.5.0 · Chrome 151.0.0.0
Sample: 15 completed mobile audits · five runs per page

Scroll the table to view all columns on smaller screens.

Purely mobile · after optimization median scores, October 25, 2026. See this session’s sample and method.
Page	Device	Median performance	Performance range	Automated accessibility
Home	Mobile	94	93–95	100
Collection	Mobile	92	91–94	100
Product	Mobile	90	89–91	100

All 15 individual runs and downloadable reports

Times shown in PKT (UTC+5). Every completed run is included.

All 15 Purely mobile · after optimization runs and reports
Tested page	Run	Performance	Automated accessibility	Reports
Home/	MobileRun 1 · 20:41 PKT	94	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 2 · 20:41 PKT	93	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 3 · 20:42 PKT	94	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 4 · 20:42 PKT	94	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 5 · 20:42 PKT	95	100	View HTML ↗ Download HTML Download JSON
Collection/collections/all	MobileRun 1 · 20:39 PKT	92	100	View HTML ↗ Download HTML Download JSON
Collection/collections/all	MobileRun 2 · 20:39 PKT	92	100	View HTML ↗ Download HTML Download JSON
Collection/collections/all	MobileRun 3 · 20:39 PKT	94	100	View HTML ↗ Download HTML Download JSON
Collection/collections/all	MobileRun 4 · 20:40 PKT	91	100	View HTML ↗ Download HTML Download JSON
Collection/collections/all	MobileRun 5 · 20:40 PKT	93	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 1 · 20:34 PKT	89	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 2 · 20:35 PKT	89	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 3 · 20:35 PKT	91	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 4 · 20:36 PKT	90	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 5 · 20:36 PKT	91	100	View HTML ↗ Download HTML Download JSON

Purely mobile · after optimization settings and provenance

Lighthouse 13.5.0 and Chrome for Testing 151, with standard mobile emulation and simulated throttling. Each run uses a fresh browser, HTTP cache disabled, and authenticated storefront storage retained (disableStorageReset: true). The recorded controlled runs have zero HTTP cache hits. No request blocking, motion changes, or skipped audits.

Deployed commit 902df512949b43b1cbc640f6f5c8ffd94c4941b1. The audit records a post-deployment check of 436 files with no code differences and no protected-file differences.

Mobile simulation: 412 × 823 CSS px; 4× CPU; 150 ms RTT; 1638.4 Kbps throughput.

Mobile results use device emulation. Full reports include the exact URLs, timestamps, configuration, metrics, and findings.

Download Purely mobile · after optimization summary Download report SHA-256 checksums

Separate standard cache-reset check · October 25, 2026

Purely product · standard cache reset

A separate three-run product check captured October 25, 2026, 20:43–20:44 PKT. The completed performance scores are 88, 93, and 95.

Product mobile median: 93 / 100

The median is 93, with a range of 88–95. This is a product-only check; it is not an average across home, product, and collection.

Tested store: purely-bmsyummi.myshopify.com
Tested theme: Milestone (tested as Manifest) · Shopify theme 131750854738
Tools: Lighthouse 13.5.0 · Chrome 151.0.0.0
Sample: Three product-page mobile audits · standard cache reset

Scroll the table to view all columns on smaller screens.

Purely product · standard cache reset median scores, October 25, 2026. See this session’s sample and method.
Page	Device	Median performance	Performance range	Automated accessibility
Product	Mobile	93	88–95	100

All 3 individual runs and downloadable reports

Times shown in PKT (UTC+5). Every completed run is included.

All 3 Purely product · standard cache reset runs and reports
Tested page	Run	Performance	Automated accessibility	Reports
Product/products/interior-nad-precursor-complex	MobileRun 1 · 20:43 PKT	88	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 2 · 20:44 PKT	93	100	View HTML ↗ Download HTML Download JSON
Product/products/interior-nad-precursor-complex	MobileRun 3 · 20:44 PKT	95	100	View HTML ↗ Download HTML Download JSON

Purely product · standard cache reset settings and provenance

Same Lighthouse version, Chrome binary, emulated device, and simulated throttling as the controlled session, with disableStorageReset: false. These product-only results are kept separate from the five-run controlled series.

Same deployed commit 902df512949b43b1cbc640f6f5c8ffd94c4941b1 and browser/device configuration as the controlled mobile session.

Mobile simulation: 412 × 823 CSS px; 4× CPU; 150 ms RTT; 1638.4 Kbps throughput.

Mobile results use device emulation. Full reports include the exact URLs, timestamps, configuration, metrics, and findings.

Download Purely product · standard cache reset summary Download report SHA-256 checksums

Latest saved benchmark session · October 24, 2026

Shopify benchmark

Separate development store with the official 2,000-product benchmark dataset and a copy of the Purely theme presentation. This is a benchmark configuration, not the Purely demo catalog.

Desktop performance average: 92.3 / 100
Mobile performance average: 93.7 / 100

Separate benchmark configuration; its averages do not combine with the Purely demo mobile sessions.

Tested store: manifest-benchmark.myshopify.com
Tested theme: Milestone (tested as Manifest) · Shopify theme 160219365531
Tools: Lighthouse 13.5.0 · Chrome 151.0.7922.34
Sample: 18 valid runs · three pages × two devices × three runs

Scroll the table to view all columns on smaller screens.

Shopify benchmark median scores, October 24, 2026. See this session’s sample and method.
Page	Device	Median performance	Performance range	Automated accessibility
Home	Desktop	88	85–96	100
Product	Desktop	100	98–100	100
Collection	Desktop	89	84–96	100
Home	Mobile	93	91–99	100
Product	Mobile	93	84–93	100
Collection	Mobile	95	95–95	100

All 18 individual runs and downloadable reports

Times shown in PKT (UTC+5). Every completed run is included.

All 18 Shopify benchmark runs and reports
Tested page	Run	Performance	Automated accessibility	Reports
Home/	DesktopRun 1 · 18:46 PKT	85	100	View HTML ↗ Download HTML Download JSON
Home/	DesktopRun 2 · 18:47 PKT	88	100	View HTML ↗ Download HTML Download JSON
Home/	DesktopRun 3 · 18:47 PKT	96	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	DesktopRun 1 · 18:48 PKT	100	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	DesktopRun 2 · 18:48 PKT	98	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	DesktopRun 3 · 18:48 PKT	100	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	DesktopRun 1 · 18:48 PKT	84	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	DesktopRun 2 · 18:49 PKT	96	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	DesktopRun 3 · 18:50 PKT	89	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 1 · 18:52 PKT	93	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 2 · 18:52 PKT	91	100	View HTML ↗ Download HTML Download JSON
Home/	MobileRun 3 · 18:53 PKT	99	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	MobileRun 1 · 18:53 PKT	84	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	MobileRun 2 · 18:53 PKT	93	100	View HTML ↗ Download HTML Download JSON
Product/products/amazing-ice-sunglasses-collection20	MobileRun 3 · 18:56 PKT	93	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	MobileRun 1 · 18:56 PKT	95	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	MobileRun 2 · 18:56 PKT	95	100	View HTML ↗ Download HTML Download JSON
Collection/collections/collection1	MobileRun 3 · 18:57 PKT	95	100	View HTML ↗ Download HTML Download JSON

Shopify benchmark settings and provenance

The source snapshot was captured on October 24, 2026 at 18:45 PKT. The benchmark used cold network caches and three valid runs per page and device. Its Lighthouse configuration skipped modern-http-insight and bf-cache; see the full report configuration.

Exact tested theme files are identified by the downloadable SHA-256 manifest. A Git commit was not recorded for this snapshot.

Desktop simulation: 1350 × 940 CSS px; 1× CPU; 40 ms RTT; 10240 Kbps throughput.
Mobile simulation: 412 × 823 CSS px; 4× CPU; 150 ms RTT; 1638.4 Kbps throughput.

Mobile results use device emulation. Full reports include the exact URLs, timestamps, configuration, metrics, and findings.

Download Shopify benchmark summary Download report SHA-256 checksums Download tested-theme file hashes

A score of 100 is the automated accessibility result for that page and run. Lighthouse does not score the manual checks needed for a complete accessibility evaluation. These measurements do not establish WCAG conformance or guarantee another store’s results. See Google’s explanations of performance scoring and accessibility scoring.

Earlier retained demo records

The previous three-run Purely session completed at 18:11 PKT preceded the evening mobile stability work. Its 94 desktop / 63.3 mobile averages are historical results, not the latest optimized mobile measurements.

Download 18:11 PKT demo summary and report links Download 18:11 PKT report checksums

The still-earlier session captured at 17:18–17:21 PKT is also retained.

Download 17:18 PKT demo summary and report links Download 17:18 PKT report checksums

Recorded follow-up · October 25, 2026

Accessibility checks

WCAG 2.1 AA target. Manual verification is incomplete. This follow-up covers focused regressions and sampled keyboard, zoom, and reduced-motion behavior. Its scope and recorded revision are separate from the Lighthouse sessions above.

Follow-up demo revision: cb2c4646b595.

Passed
Drawer and heading regression fixtures
18 engine, viewport, and motion combinations.
Passed
Variant focus regression fixtures
54 cases covering focus during asynchronous variant updates.
Passed
Cart reflow regression fixtures
36 viewport combinations, with three focused controls checked in each.
Passed
Filter focus regression fixtures
18 engine, template, and viewport combinations.
Sampled
Native Mac Chrome keyboard, zoom, and reduced-motion samples
Keyboard flows sampled in Chrome 155.0.8059.86 on macOS 15.1; 26 page/zoom measurements at 200% and 400% were recorded.

Findings fixed and retested

Delayed variant responses stealing focus from the next task.
Cart actions becoming unreachable in short viewports and at high zoom.
Routine and password images missing alt attributes.
Filter Apply losing focus after the toolbar refresh.

Still awaiting verification

Spoken journeys with VoiceOver, TalkBack, and NVDA, including announcement retests.
Required native browser, physical phone, and in-app browser coverage.
Real-device virtual keyboard, text size, orientation, and touch checks.
Six valid Lighthouse reports pinned to the final theme revision after all fixes are complete.
The full content-extremes and optional-settings sweep on the final candidate.

Bundled Chromium, Firefox, and WebKit regression fixtures are not native-browser or physical-device certification. DOM and live-region assertions do not verify spoken screen-reader output. W3C’s evaluation guidance explains why human testing is required.

Download accessibility results and limitations

Local workspace check · October 26, 2026

Build checks

These checks describe the local theme workspace. They are separate from the archived demo measurements.

Merchant-addable sections: 42
Page types: 1322 template files including alternate layouts
Minified JavaScript file cap: 16 KB16,384 bytes per authored file
JavaScript check: Passed90 files checked; largest 16,056 bytes (15.7 KB)

The JavaScript cap applies to each minified theme file, not the combined JavaScript downloaded by a page. The checked workspace includes uncommitted changes, so its results are a working snapshot, not a published-release certification.

Download build-check results

Return to Milestone →

Milestone test results

Drawer and heading regression fixtures

Variant focus regression fixtures

Cart reflow regression fixtures

Filter focus regression fixtures

Native Mac Chrome keyboard, zoom, and reduced-motion samples

Findings fixed and retested

Still awaiting verification