MP Audits
Request an audit

Enterprise e-commerce accessibility engineering

E-commerce accessibility & ADA defense. Engineered at the source.

We audit and remediate the code-level WCAG 2.2 defects that plaintiff bots exploit on active checkout paths — without slow widgets or design compromises.

Engineered forShopifyWooCommerceMagentoCustom headless

Request a compliance audit of your storefront.

One URL is all we need. No app install, no theme access, and nothing added to your store. The brief comes back within 48 hours with the evidence attached.

  • Non-invasive production scan
  • Full checkout & cart verification
  • 100% confidential
2 of 5
storefronts we audited were already paying for an accessibility overlay
45
defects remained on those stores after the overlay had finished running
11
separate WCAG success criteria cited across the storefronts measured

From our own audits, to Sep 04, 2026. Reports are issued privately, one link per storefront, and are never indexed.

§01

What the engagement protects

01

Eliminate Legal Liability

Automated discovery crawlers inspect the markup your server sends, bypassing surface widgets entirely. We fix the vulnerabilities at the code layer, which is where the liability actually sits.

Findings closed at the source, not masked at runtime

02

Protect Checkout Conversion

Inaccessible cart controls, unlabelled buttons and failing contrast cost you completed orders long before they cost you anything else. We test the live checkout path and remove the drop-off.

Cart and checkout verified with a real basket

03

Preserve Store Speed & Aesthetics

No third-party JavaScript loading on every page, nothing slowing your mobile shoppers, and no widget badge sitting in the corner of your brand. The corrections are native to your own code.

Zero added scripts, zero visual footprint

§02

Where overlay coverage ends

An overlay operates on the rendered page. The defects it is meant to cover are in the theme underneath it.

An overlay (UserWay, accessiBe) is a script that loads in a visitor’s browser and rearranges what it can reach, after your page has already rendered. It is a layer on top. It is not a repair.

That distinction determines what a scan finds. Automated discovery crawlers commonly evaluate the document as your server sends it, without executing client-side scripts at all, which leaves the raw markup exposed regardless of what widget is installed. A reviewer working in the browser sees the same defects a second way, because the widget does not correct them there either.

Across the storefronts we have measured, an installed overlay has not yet produced a clean result. The defects are in Liquid and CSS, and that is the layer where they are resolved.

What an overlay cannot reach

  • Your cart’s remove button

    Ships with no accessible name. A screen reader announces “link” and nothing else, on the path to checkout.

  • Your brand colors

    The failing values are compiled into your stylesheet and served to every shopper on every page load.

  • Your product links

    Tap targets under the required size, and links with no readable text behind the icon.

All three are theme-level, all three persist with a widget installed, and all three are among the defect classes most commonly cited in accessibility claims against e-commerce sites.

§03

What a client brief looks like

See how we isolate checkout vulnerabilities in production storefronts.

Every finding is reproducible by the person reading it. Beside each one sits the element as it rendered, the selector that isolates it in your own code, the measured value, and the corrected value — with the arithmetic shown.

Client briefs are confidential. The sample below is a public demonstration storefront, published by Deque Systems as an accessibility teaching site, audited with the same engine on the same settings.

Open the sample brief

57 findings, 11 rules, full risk matrix and evidence exhibits

Exhibit C-1color-contrast
SC 1.4.3
Be Bold
As shipped4.32:1 FAIL

#FF9999 on #344B6E

Be Bold
Corrected4.98:1 PASS

#FFADAD on #344B6E

a[href="mars2.html?a=be_bold"] > h3

Reproduced at the format a client receives. The corrected value is computed against the same background, so layout and design are untouched — the change is one value in the stylesheet.

§04

Verification and documentation standards

Enterprise e-commerce accessibility and theme engineering.

Every audit is a code-level WCAG 2.2 verification of a live storefront, documented to a standard that holds up outside our own office. A finding is only useful if the person reading it can reproduce it — so each one carries the element it was measured on, the criterion it fails, and the arithmetic behind it.

The same brief serves your developer, your counsel, your insurer, and a technical due-diligence review. One document, written once, precise enough that none of them needs to ask us what it means.

Your engineering team
Each finding names the element, the criterion, and the corrected value — enough to action without a follow-up call.
Counsel and insurers
A weighted exposure figure that traces, point by point, back to a specific node in the ledger.
Technical due diligence
A dated, reproducible record of a storefront’s compliance position at a known moment.
Independence
We sell no accessibility widget and take no vendor commission. Nothing in a finding is influenced by what we would be paid to fix.

What ships with every finding

UTC
Timestamped capture
1:1
DOM selector per finding
Ratio
Contrast math shown
MPA
Referenced document
  • Screenshot of the element as it rendered, at the recorded moment
  • The selector that isolates it in your own theme
  • Measured value, required threshold, and the corrected value
  • Toolchain versions, and a written statement of what was not covered
§05

The MP Audits methodology

Three engineering commitments. Each is a decision about where effort is spent, and each is visible in the brief you receive.

A

Code-Level WCAG 2.2 Verification

Findings are measurements taken from your live storefront, not suggestions produced from a checklist.

A general-purpose scanner returns several hundred items in whatever order it found them. That is a list, not an assessment — it says nothing about which defects affect your customers or your exposure. Each entry in our brief records a value read off a specific element, together with the element it came from.

Findings are prioritised rather than enumerated. Seven defect classes are weighted by transaction impact and by how frequently each appears in accessibility claims against e-commerce sites. A keyboard trap on the cart drawer carries the heaviest weight. Colour contrast, the highest-volume finding in any general audit, carries less — it is easy to document, but it rarely prevents a purchase.

The result is a brief your engineering team can work through in priority order, with the reasoning for that order stated on the page.

How findings are prioritised

  • 25Cart or drawer keyboard trapBlocks the transaction outright for keyboard users
  • 20Unnamed cart and checkout controlsThe shopper cannot tell what the control does
  • 20Unlabelled checkout and account inputsPrevents order completion; usually found alongside
  • 15Color contrast below thresholdHigh volume, but rarely the transaction blocker

Seven factors, 0–100 composite. The full table ships inside every brief.

B

Transactional Path Integrity

We test the route to checkout with an item in the basket, because that is the route that has to work.

Accessibility failures on the transaction path cost twice: once in exposure, and once in orders that never complete. A control a shopper cannot identify is a control they cannot use, whatever the reason they are using assistive technology.

Most audits never reach this surface. An empty basket suppresses the cart drawer on the majority of themes, so we seed one item, exercise the cart, and empty it again. Nothing is ordered, and the storefront is left exactly as we found it.

The drawer is then verified for focus placement, focus containment, Escape handling, and focus restoration. Where no drawer can be confirmed, the brief says so plainly — that is the absence of a drawer, not a drawer that passed, and we never let it read as one.

C

Native Theme Remediation

Corrections are written into Liquid and CSS, then verified on a duplicate theme before anything is published.

These defects live in the markup your server renders. They are present before any client-side script executes and remain present after one has run, which is why a runtime widget cannot resolve them: it cannot change a colour compiled into your stylesheet, the hit-area geometry of a product-card link, or a control that ships without an accessible name.

Our patches are additive by construction. Nothing in your theme is overwritten, the stylesheet loads last so it wins on specificity rather than by force, and the whole layer lifts out cleanly if you ever want it gone.

On a staged engagement we duplicate your published theme and deploy into the copy. Your live storefront serves exactly the code it served this morning, throughout. We re-scan the patched copy, test the checkout path with a screen reader, and hand you a preview link and the delta. You publish when you are satisfied.

§06

The remediation protocol

Four stages, fixed price at each one. You can stop after the audit and hand the report to your own developer — plenty of clients do, and the report is written so that works.

No app is installed on your store. Nothing is added to your storefront. Nothing slows your pages down, and nothing appears in your UI.

Automated testing catches roughly a third of WCAG failures. We say so in every report, and Tier 2 adds the screen-reader testing that covers the rest of the checkout path.

  1. 01

    Audit

    48 hours

    Send one storefront URL. No app install, no theme access, nothing added to your store. You get a referenced report with the evidence attached and a fixed price to close it.

  2. 02

    Patch

    On order

    We write the diffs against the exact theme your store is running, and a stylesheet that only adds — no existing rule of yours is overwritten, and the patch can be removed cleanly.

  3. 03

    Stage

    2 business days

    On Tier 2, our engineers duplicate your theme and deploy into the copy. Your live storefront serves the same code it served this morning, throughout.

  4. 04

    Verify

    Before you publish

    We re-scan the patched copy and test the checkout path with a screen reader, then hand you the delta and a preview link. Your decision, your publish button.

§07

Engagement tiers

Fixed price, scoped to the defects in your report. No retainer, no subscription, and no per-page-view fee that grows with your traffic.

Tier 1Delivered on checkout
$299one-time · USD

Remediation Patch Package

Every defect in your report, closed. Verified CSS and Liquid diffs your developer applies in an afternoon.

  • Verified CSS and Liquid diffs, one per defect in your report
  • assets/a11y-shield.css — additive only, nothing in your theme is overwritten
  • Step-by-step implementation guide written for your developer, not for us
  • Corrected color values computed against your own backgrounds — no redesign
  • Post-implementation re-scan, so you have dated proof the defects are closed

Your developer applies it. Nothing changes on your storefront until you decide it does.

Tier 22 business days from theme access
$599one-time · USD

Turnkey Staging Implementation

Our engineers deploy it for you — to a duplicate of your theme, never the live one. You review, then you publish.

  • Everything in Tier 1
  • Full deployment by our engineers to a duplicate of your live theme
  • Screen-reader testing of the checkout path — the part automation cannot do
  • An official Accessibility Conformance Statement for your records
  • Preview link for your review. You publish. We never press the button.

Your live theme is untouched for the entire engagement.

Both tiers are ordered from inside your report, where the scope is itemised defect by defect. The audit itself is quoted per storefront — send a URL and we will price it.

§08

Institutional backing

MP Audits is the technical compliance division of Mega Promote LLC, a United States limited liability company.

The division exists for one reason: accessibility verification is worth nothing if the party performing it also sells the remedy it recommends. We do not sell an overlay, we do not resell anyone else’s, and we take no commission from any vendor named in a report.

The engineering standard follows from that. A report that will be read by opposing counsel has to survive being read by opposing counsel — which is why every finding carries the element it was measured on, the criterion it fails, and the number that proves it.

Registered office

Mega Promote LLC[Registered address — pending]United States
Legal structure

Mega Promote LLC, a United States limited liability company.

MP Audits operates as its technical compliance division.

Independence

We sell no accessibility widget and take no vendor commission.

No affiliation with Shopify Inc., Deque Systems, or any overlay vendor named in our reports.

Standards referenced

WCAG 2.1 AA · WCAG 2.2 AA

Section 508 · EN 301 549 · ADA Title III (US)

Document discipline

Every report carries a reference number, a UTC timestamp, and the toolchain versions used.

Every report states in writing what automated testing does not cover.

Request an audit

Send one storefront URL. The brief comes back with the evidence attached.

No theme access, no app install, and nothing added to your storefront. If a complaint is already on your desk, note it in the subject line and we will bring the audit forward.

audits@mpaudits.com

  • Non-invasive production scan
  • Full checkout & cart verification
  • 100% confidential