Digital Accessibility Assessment Services

Most agencies commissioning a digital accessibility assessment are not starting from zero. They already have an internal review checklist. They already run automated scans — Lighthouse, axe, WAVE. Staff already care.

What they do not have is an independent answer to the question their board, their funder, or their counsel is actually asking: is what we are doing enough, and if not, what do we fix first?

That is what an assessment is for. Not a scan you could run yourself, but an expert third-party evaluation of your assets and of your process, ending in a prioritized roadmap somebody can execute.

The five things an assessment has to cover

Agencies scope this work in phases, and the sequence matters — testing before you understand the asset inventory wastes the budget on the wrong pages.

1. Your accessibility program, assessed honestly. We review your internal checklist, your review workflow, who is responsible for what, your documentation and QA steps, and your use of automated tools — then say where your current practice is sufficient and where it is not. Critically, this includes drawing the line between what automated testing can safely cover and what genuinely requires expert manual testing. Most internal programs are strong on the former and have no coverage of the latter.

2. An asset inventory review and a sampling plan. Which assets need full manual testing, which can be covered by representative sampling, and which are best addressed once at template, platform or process level. If you already have a proposed sample set, we validate it and tell you where it needs additions to be genuinely representative rather than convenient.

3. Expert testing of the sample. Automated scanning first, because it is fast and thorough at structural defects. Then the manual work that decides whether something is actually usable — described below.

4. A gap analysis and prioritized remediation roadmap. Findings by severity, public impact, usage and remediation complexity. Recurring issues grouped by template, platform or framework, because that is where a single fix pays for itself many times. Suggested ownership, what your team can do internally versus what needs a vendor, and recommendations for tracking and communicating outstanding items — which is what compliance transparency actually requires.

5. Training and knowledge transfer. An assessment that produces a report and no capability leaves you buying another assessment next year. Practical WCAG application rather than theory, how to interpret automated tool output and what it cannot tell you, manual testing techniques your own staff can run, review workflows for content, documents, maps, dashboards and apps, accessible authoring in the tools your team already uses, and the governance and QA steps that keep new content conformant. Live sessions, with materials you keep and can re-run for new starters.

Where automated testing stops

This is the part worth being specific about, because it is the entire justification for the engagement.

Automated tools reliably detect only a minority of WCAG success criteria — the widely cited figure is roughly a third. They are excellent at what they do: missing alt attributes, contrast ratios, missing labels, malformed ARIA. They cannot tell you:

• Whether the reading and focus order make sense, or whether the keyboard jumps from a column heading to a footer.
• Whether alt text is correct rather than merely present. A chart described as “chart” passes automated checks and tells a blind reader nothing.
• Whether a form can actually be completed without a mouse, in order, with errors that announce themselves.
• Whether a data table’s relationships survive being read aloud — header cells associated, scope correct, layout tables artifacted.
• Whether a map, dashboard or data visualization has a usable non-visual alternative, which is the single most common unaddressed gap in public-sector digital content.

Every finding we report is labelled with the method that produced it — automated, manual, screen reader, code review, or user-flow review. That distinction is not pedantry: a finding your own scanner could have caught tells you something about your process, and a finding that required an expert tells you something about your coverage.

What we assess

• Public websites and web applications — templates, navigation, forms, content types, and the recurring components that affect every page.
• Downloadable documents — PDF, Word, Excel, PowerPoint. See document accessibility and Section 508 remediation.
• Images, charts, graphics and data tables.
• Video and multimedia — captions, audio description, transcripts. See accessible video.
• Interactive maps, dashboards and data visualizations — including ArcGIS Hub sites, StoryMaps, dashboards and data portals. Keyboard operability, focus management, screen-reader output, ARIA, and the layers, legends, filters, pop-ups and data controls they expose — plus whether a non-visual alternative exists for the information the visual actually conveys.
• Native mobile applications — iOS and Android.

Maps, dashboards and mobile apps — the assets most assessments skip

These are where public-sector accessibility work usually stops, and they are where the highest-value public information tends to live.

Maps and GIS platforms. The work splits into three layers, and knowing which layer a defect sits in is what stops you spending money on something you cannot reach:

• Content and configuration — what you control, and where most findings are. Alternative text, heading structure, colour and non-text contrast, layer and legend labelling, pop-up markup, and whether a non-visual alternative to the map exists at all.
• Template level — one StoryMap or dashboard template fixed propagates across every asset built from it. This is where the budget goes furthest and it is the first thing we look for.
• Platform level — and honestly, some defects live in the platform vendor’s own code. There the deliverable is a documented workaround plus the specific issue to escalate to the vendor, not an invoice to fix something that is not yours.

The map alternative, concretely. This is the single most common unaddressed gap we see. A map conveys information — which routes serve which areas, where facilities are, how a value varies by geography. A blind or low-vision user needs that information, not a description of the map’s appearance. The usable alternative is the underlying data presented accessibly: a properly marked-up table, a text summary of the pattern, and a non-map route to the same task — an address lookup that returns a result in text. Alt text reading “map of the region” is not an alternative, and it passes automated checks.

Native mobile apps are a different discipline from web testing, not the same checklist re-run. Assessment covers key user flows, navigation and controls, forms, map and location features, content presentation and touch target sizing, screen-reader compatibility with VoiceOver and TalkBack, keyboard and switch access where applicable, and behaviour under the platform settings people actually use — Dynamic Type and larger text, reduce motion, increased contrast.

And one thing we report that most assessments do not: whether each mobile issue is specific to the app, shared with your web platform, or inherited from a third-party SDK. Those three have entirely different owners, fix costs and timelines, and conflating them is how a remediation budget gets spent in the wrong place.

WCAG 2.1 or 2.2 — and why the answer matters in a solicitation

The DOJ Title II web rule adopts WCAG 2.1 Level AA as its technical standard. Many agencies now specify WCAG 2.2 Level AA in procurement, and that is the better choice: 2.2 adds nine success criteria and obsoletes one (4.1.1 Parsing), so it is effectively a superset — conform to 2.2 AA and you satisfy 2.1 AA.

We assess against 2.2 AA and map findings to both, plus Section 508 where federal funding or federal procurement is in play. Naming the standard and the specific criterion against each finding is what makes a report usable as evidence rather than as an opinion.

The deadline dates that apply to you are in the banner above and in our ADA Title II deadline guide — maintained in one place so they are right.

Assessment and remediation should be separate tracks

We do both, and we will still tell you to keep them separate on paper.

An assessor who is also bidding to fix everything they find has an obvious incentive to find a lot. And a roadmap written to be executable by anyone is more valuable than one only its author can act on — which is why our findings carry stable identifiers you can hand to your own team or a third-party vendor.

If you want the assessment alone, that is a complete engagement and the roadmap is yours to take anywhere. If you want help afterwards, we provide document and PDF remediation, remediation guidance for web content, accessible template development, and follow-up validation testing after your team or another vendor completes the fixes.

For a single document rather than a programme, the PDF Accessibility Audit is a fixed-price version of the same discipline, with a full sample report you can read before you buy.

Buying it: contract vehicles and federal award rules

For public agencies this is often the part that decides the timeline rather than the technical scope.

• GSA Schedule holder and NASPO ValuePoint contract holder, so many state, local and regional bodies can purchase through a vehicle they already hold rather than running a new solicitation.
• SAM.gov registered with an active Unique Entity Identifier — routinely required on the title page of a proposal, and a disqualifier when missing.
• Veteran-owned small business.
• Set up for procurement governed by the Uniform Guidance (2 CFR Part 200), including the competitive-selection requirements at 2 CFR 200.318–200.320, which apply when the work is funded by a federal award.
• Able to supply what these solicitations actually ask for: itemized pricing by task with labor rates broken out, a work plan mapped to your stated scope, a timeline, references, a certificate of insurance, and a separately redacted copy where public-records law requires one.

See GSA translation services and NASPO ValuePoint for the contract detail.

What an assessment cannot do

It is not a certification. There is no such thing as being “ADA certified” — no such certification exists, and a vendor offering one is selling something that is not real. An assessment documents conformance against named technical standards at a point in time.

It is not a legal opinion. Whether your organization carries a particular obligation, and what your exposure is, is a determination for your counsel. We give you an accurate technical picture and we will not tell you that you are safe.

It is not permanent. Digital assets change weekly. That is exactly why the training and governance phase matters more than the findings list.

And some findings are not yours to fix. Where a defect sits in a platform vendor’s code — a mapping platform, a CMS, a third-party mobile SDK — no amount of remediation budget on your side will close it. We identify those explicitly, give you the documented workaround, and tell you exactly what to escalate to the vendor. A report that quietly lists them alongside things you can actually fix is a report that will waste your money.

The half most vendors cannot cover

Public agencies carry two access obligations on the same content, and almost always buy them from two vendors.

Title VI of the Civil Rights Act requires meaningful access for people with limited English proficiency. The ADA and Section 508 require access for people with disabilities. The same public notice, the same benefit application, the same board packet, the same video is subject to both — and a document translated into six languages and left inaccessible in all six has solved half the problem six times over.

Because this is a language company that added accessibility rather than an accessibility firm that subcontracts translation, one vendor covers both, with consistent terminology across the spoken, written and translated versions. See education language access, healthcare language access, and certified translation.

(A note on citations, since stale ones appear in a lot of procurement boilerplate: Executive Order 13166 was revoked in 2025. Title VI itself is untouched and remains the correct authority — we explain the change here.)

Start where it costs you nothing

If you are scoping a solicitation or building a budget case, the free accessibility assessment gives you a scoped plan and a price without a procurement. If you already have a scope and a deadline, request a scoped proposal and tell us the standard, the asset inventory and the submission date.