Research · Original measurement · November 2026
How accessible are Fortune 100 homepages?
Automated scanning makes America's largest companies look nearly clean, but the checks a human makes do not. 46 of 100 homepages had zero axe-core violations, yet only 18 passed both the automated scan and every manual check.
Updated November 2026Reviewed by Khushwant Parihar, CPACC
100 homepages, one square each
axe-core 4.10.2 plus manual checks, November 2026
Key findings
Eight numbers from the study
A clean automated scan hid real barriers on 28 of the 46 axe-clean homepages, so automated results alone overstate how accessible these sites are.
54
homepages fail at least one automated WCAG A/AA rule; 28 have a critical violation
21
fail color contrast, the most common automated failure
27
have no skip link, and on 6 more it points to a target that doesn't exist
30
have no main landmark, and 27 have no h1
15
have poor alt text that axe passes: file names, asset codes, placeholders
7
have a verified keyboard or focus problem, mostly from cookie banners
15
scroll horizontally at 320px, failing WCAG 1.4.10 reflow
4
load an overlay widget: UserWay, AudioEye, accessiBe, Recite Me
Automated results
What axe-core found
46 homepages passed every axe-core WCAG A/AA rule, 14 failed exactly one, and 40 failed two or more. The median homepage fails just 1 rule, and 592 failing elements were found across all 100.
Rules failed per homepage
How many axe-core WCAG A/AA rules each homepage failed
- Failing elements in total
- 592
- Homepages with a critical violation
- 28
Most common automated failures
Number of homepages failing each rule at least once
| Rule | WCAG | Homepages |
|---|---|---|
| Color contrast | 1.4.3 | 21 |
| Target size (minimum) | 2.5.8 | 16 |
| Links without an accessible name | 2.4.4, 4.1.2 | 15 |
| Images without alt text | 1.1.1 | 11 |
Target size is new in WCAG 2.2 and already the 2nd most common failure by homepage count, and 1st by element count (138).
The heaviest automated results
Nike alone accounts for 105 of the 592 failing elements.
| Company | What failed | Failing elements |
|---|---|---|
| Nike | 74 focusable elements inside aria-hidden regions, plus 12 contrast failures and broken list and ARIA structure. | 105 |
| Costco | 57 targets smaller than the WCAG 2.2 minimum size. | 57 |
| Tesla | 11 unnamed buttons and 11 elements missing a required ARIA parent role, both critical, plus 10 nested interactive controls. Its 27 critical elements were the most of any homepage. | 38 |
| Home Depot | 27 color-contrast failures. | 37 |
| AbbVie | 25 elements with an invalid lang value (lang="none"). | 26 |
Manual review
What only a person catches
74 homepages failed at least one manual check, and the most common failures are structural basics. Only 26 passed every manual check.
Manual check failures
Number of homepages failing each scripted or manual check
| Check | Homepages |
|---|---|
| No main landmark | 30 |
| No skip link | 27 |
| No h1 | 27 |
| Poor alt text | 15 |
| Fails 320px reflow | 15 |
| Keyboard or focus problem | 7 |
| Skip link target missing | 6 |
| Pinch-zoom disabled | 1 |
33 homepages have no working skip link: 27 have none, and on 6 the link points to a target that doesn't exist. See the skip links guide.
Alt text that passes axe but fails people
axe checks that alt text exists, not that it makes sense. 15 homepages were rated Poor on review. Examples:
- Phillips 66
alt="HumberRefineryP66_00477.webp"- PepsiCo
alt="HOMEPAGE_BANNER_NEW_LAYS"alt="FDS REFORMAT"alt="8"- Coca-Cola
alt="thumb-0"alt="thumb-1"alt="thumb-2"- UnitedHealth Group
alt="image"alt="alt text"- Procter & Gamble
alt="Placeholder"- Nationwide
alt="/staticassets/p-home-qr-code_tcm108…"- Eli Lilly
alt="Pause-icon"alt="arrowIcon-icon"- Morgan Stanley
alt="s"
7 verified keyboard and focus problems
Each one confirmed by following focus with real Tab presses or comparing screenshots. 5 of the 7 come from consent banners or dialogs.
Focus trap
Johnson & Johnson
Cookie banner confines Tab to two buttons and the browser chrome until dismissed
Focus trap
Boeing
Tab cycles cookie-banner controls and a Feedback button; page content unreachable until dismissed
Focus obscured
AbbVie
Focused news links sit behind the cookie banner (focus obscured, WCAG 2.4.11)
Focus trap
American Airlines
Consent panel takes focus on load and Tab cycles inside it until dismissed
Focus trap
UnitedHealth Group
Consent dialog holds focus on load; Tab cycles only through its three buttons
No focus indicator
TJX
Search field shows no focus indicator beyond the caret
No focus indicator
Cardinal Health
Focusable promo card shows no visible focus indicator
The automatic focus-visibility check flagged far more pages (29). Spot checks showed it misfires on shadow-DOM headers (Humana, Prudential) and cookie overlays, so those flags were treated as leads, not failures.
Reflow and zoom
Page loaded in a true 320 CSS px viewport, the same width as 1280px at 400% zoom
- homepages scroll sideways at 320px
- 15
- widest typical overflow (Elevance Health)
- 94px
- Amazon's fixed desktop layout
- 1000px
Most failing homepages overflow by less than 100px, and on most of them the first screen looks fine. The overflow comes from wide elements further down. Amazon's desktop homepage has no viewport setting at all.
TJX is the only homepage that disables pinch-zoom (user-scalable=no).
Only 4 use an overlay
Homepages loading an accessibility overlay widget script
| Company | Overlay vendor |
|---|---|
| Cardinal Health | UserWay |
| Energy Transfer | AudioEye |
| Enterprise Products Partners | accessiBe |
| Coca-Cola | Recite Me |
Enterprise Products and Coca-Cola both still fail reflow, and Coca-Cola also has Poor alt text.
By industry
Insurers lead, automotive and retail trail
All 12 insurance homepages had zero axe violations, and 8 of the 18 cleanest homepages overall are insurers. Automotive & transport and retail homepages carry the most barriers.
| Industry | Homepages | Zero axe violations | Clean on both | Median barrier points |
|---|---|---|---|---|
| Insurance | 12 | 12 | 8 | 0 |
| Industrial | 1 | 0 | 0 | 3 |
| Health care | 16 | 8 | 2 | 7.5 |
| Technology | 15 | 7 | 3 | 9 |
| Energy | 8 | 3 | 1 | 10.5 |
| Financial services | 13 | 5 | 1 | 12 |
| Consumer goods & food | 8 | 4 | 1 | 13.5 |
| Telecom & media | 5 | 2 | 1 | 21 |
| Retail & e-commerce | 10 | 3 | 1 | 24 |
| Aerospace & defense | 4 | 0 | 0 | 24 |
| Automotive & transport | 8 | 2 | 0 | 27 |
Industry groups are our own, broader than Fortune's sector labels; Berkshire Hathaway, for example, is counted under financial services rather than insurance. Groups are small (1 to 16 homepages), so read differences between neighbouring rows as directional. Even among insurers, MetLife, New York Life, Nationwide and Prudential fail manual checks that axe cannot see: missing skip links and main landmarks, poor alt text, or no h1.
Leaders and laggards
Who passed, and who carries the most barriers
The 10 homepages with the most barrier points
Barrier points rank pages only: impact weight times failing elements per axe rule (at most 10 elements counted), plus 3 per failed manual check
| Company | Fortune rank | axe rules / elements | Critical elements | Manual check fails | Barrier points |
|---|---|---|---|---|---|
| Tesla | 43 | 7 / 38 | 27 | Repeated vague link text | 136 |
| Nike | 99 | 7 / 105 | 8 | No h1 | 128 |
| Home Depot | 25 | 4 / 37 | 1 | No skip link, no h1, no main landmark, poor alt text | 73 |
| Albertsons | 57 | 3 / 17 | 12 | No skip link, no h1, repeated vague link text | 72 |
| Coca-Cola | 98 | 4 / 18 | 1 | No skip link, no main landmark, poor alt text, vague links, reflow | 70 |
| FedEx | 50 | 3 / 20 | 1 | Skip link target missing, no main landmark | 58 |
| Galaxy Digital | 76 | 3 / 24 | 1 | No skip link, repeated vague link text | 55 |
| Verizon | 28 | 3 / 14 | 4 | No h1, reflow | 52 |
| Bank of America | 20 | 3 / 12 | 9 | No h1, no main landmark | 51 |
| Apple | 4 | 3 / 15 | 2 | No skip link | 47 |
Tesla and Amazon were tested in a real Chrome from India because they blocked headless browsers; Tesla's result includes a country-selector banner shown to that location.
The 18 clean homepages
Passed the axe-core scan and every manual check. A floor, not proof of conformance.
- Walmart
- Alphabet
- CVS Health
- McKesson
- Microsoft
- JPMorgan Chase
- State Farm
- Valero Energy
- Progressive
- Archer Daniels Midland
- Allstate
- Charter Communications
- Uber
- USAA
- TIAA
- Liberty Mutual
- Travelers
- MassMutual
8 of the 18 are insurers. Walmart is the only retailer on the list.
For practitioners
What this means for your own testing
Cookie banners are the new keyboard trap.
5 of the 7 verified keyboard issues come from consent banners or dialogs. Test the default first-visit state, not a session where consent is already given.
Alt-text quality needs a human.
The content team, not the code, produced most Poor ratings: file names and CMS asset codes flowing straight into alt attributes. See how to write alt text.
WCAG 2.2 is already biting.
Target size is the 2nd most common axe failure, and focus obscured by sticky banners (2.4.11) showed up on AbbVie.
Structure basics are still missing at scale.
Between 27 and 30 homepages each lack a skip link, an h1 or a main landmark: cheap fixes with large payoffs for screen-reader and keyboard users.
Overlays are rare and don't fix the basics.
Two of the four overlay sites still fail reflow, and one has Poor alt text. Read why overlays don't fix accessibility.
- Open the URL auditor
Run the automated half on your own homepage.
The URL auditor runs axe-core against any page. Then do the manual half: Tab through it, zoom to 400%, and read the alt text.
Method
How the homepages were tested
Each company's homepage was tested in one study on 14 and 15 November 2026, using ranks 1 to 100 of the 2026 Fortune 500 as published by DemandSage. The full automated and manual run took place on 14 November, and every company-specific finding on this page was confirmed on the live site the next day.
92 homepages ran in headless Chrome with a US locale. Eight were tested in a real desktop Chrome from India instead: seven served bot challenges to headless traffic (Amazon, Marathon Petroleum, StoneX, Tesla, Eli Lilly, Oracle, American Airlines), and Nationwide rendered blank. Some of those eight showed country prompts.
Limits. This covers one page per company during Black Friday promotions, and homepages change daily: by the second day a few axe counts had shifted, most of all on Albertsons and FedEx, so treat the figures as a snapshot. No screen reader testing was done, and no forms, checkout flows, mobile apps or PDFs were tested. Zero axe violations is not WCAG conformance: automated rules catch only a minority of issues, which is why the manual checks matter.
Barrier points, used only to rank pages, add up each failed axe rule's impact weight (critical 4, serious 3, moderate 2, minor 1) times its failing elements, counting at most 10 elements per rule. Each failed manual check adds 3.
| Check | How it was done |
|---|---|
| Automated scan | axe-core 4.10.2, WCAG 2.0, 2.1 and 2.2 A and AA rules only (best-practice rules excluded), desktop viewport 1366×900 |
| Keyboard | 30 real Tab presses from the top; focus order, traps and hidden focus stops logged; focus visibility measured by pixel difference, then the flagged cases verified by screenshot |
| Reflow | Page loaded in a true 320 CSS px viewport (1280px at 400% zoom); fail = horizontal scrolling |
| Structure | Skip link and whether its target exists, h1 count, main landmark, lang attribute |
| Alt text | Up to 25 alt values per page read and rated Good, Fair or Poor |
| Links | Repeated vague link text (“Learn more” ×2 or more) with no accessible-name override |
| Other | Viewport zoom lock, autoplaying video without a pause control, overlay widget scripts |
Sources
- DemandSage: Fortune 500 Companies List (2026), ranks 1 to 100 used for scope
- Deque: axe-core, version 4.10.2, automated rules engine
Frequently Asked Questions
How was the Fortune 100 accessibility study done?
Each of the 100 homepages, taken from ranks 1 to 100 of the 2026 Fortune 500, was tested in one study on 14 and 15 November 2026. axe-core 4.10.2 ran the WCAG 2.0, 2.1 and 2.2 A and AA rules at a 1366×900 desktop viewport. Scripted and manual checks covered keyboard focus (30 real Tab presses), reflow in a true 320 CSS px viewport, skip links, headings, landmarks, alt-text quality, zoom locks and overlay scripts. Every company-specific finding on the page was confirmed on the live site.
Does a homepage with zero axe violations meet WCAG?
No. Automated rules catch only a minority of WCAG failures. In this study 46 homepages had zero axe-core violations, but 28 of those still failed at least one manual check, such as a missing skip link, no main landmark, poor alt text or a keyboard trap. Passing both is a floor, not proof of conformance.
Why are cookie banners a keyboard accessibility problem?
5 of the 7 verified keyboard issues came from consent banners or dialogs: some confine Tab focus to the banner until it is dismissed, and on AbbVie the banner covers focused links (WCAG 2.4.11 Focus Not Obscured). The problem only appears on a first visit, so test the default state before consent is given.
How many Fortune 100 homepages use an accessibility overlay?
4: Cardinal Health (UserWay), Energy Transfer (AudioEye), Enterprise Products Partners (accessiBe) and Coca-Cola (Recite Me). Two of the four still fail the 320px reflow test, and Coca-Cola also has poor alt text.
Essential Accessibility Resources
Comprehensive tools, checklists, and guides to help you create inclusive digital experiences