An accessibility specialist or digital compliance officer at a mid-size company runs a quarterly audit and discovers that hundreds of images across their CMS have alt text that was bulk-generated months ago — some of it wrong, some of it missing the subject entirely, some of it stuffed with keywords in a way that degrades screen reader experience. They have no systematic way to identify which images have bad alt text without reading every one manually.

This problem exists because alt text generation and alt text quality assurance are treated as the same step, but they serve different stakeholders. The person who ran the bulk generation cared about speed and coverage. The accessibility team cares about accuracy and compliance, and they come in after the fact with no tooling designed for their job.

Complaints like 'alt texts were created for some images that I didn't want alt text for' and 'some descriptions could be more in depth and it sometimes misses the focal point' describe exactly the kind of errors that create real WCAG compliance failures — decorative images that should have empty alt attributes but have descriptive text instead, or complex informational images that have shallow one-line descriptions. These are not aesthetic problems; they are auditable defects.

Compliance audits recur on a schedule, triggered by policy, legal review, or contract requirements. Each audit cycle requires someone to re-examine the same corpus of images as new content is added. Without a tool that flags likely compliance failures specifically — not just 'this text might be generic' but 'this image appears informational and has a description under 10 words' — auditors are doing manual pattern-matching at scale, which is slow and inconsistent.

What to build

Build a crawl-and-audit tool that scans a domain or CMS export, classifies each image by type (decorative, informational, complex), compares existing alt text against WCAG sufficiency criteria for that type, and produces a prioritized defect report with suggested rewrites — exportable as a CSV the compliance team can work through or hand to a developer.

Where to start

Start with higher education institutions, which face consistent WCAG 2.1 AA audit pressure from accreditors and legal counsel, often have large image-heavy CMS sites built years ago, and have accessibility office staff who are actively looking for audit tooling rather than needing to be convinced the problem exists.

The hard part

Classifying whether an image is 'decorative' versus 'informational' well enough that compliance teams trust your defect flags requires high precision — a false positive (flagging good alt text as defective) erodes trust in the audit output and makes the tool feel like more work than manual review.

How it makes money

Annual contract priced per domain audited — $1,200/year for a single domain with quarterly re-scans, $3,500/year for up to ten domains — sold to the accessibility office or IT compliance team, not to the marketing team that originally bought the alt text generator.

See the evidence. The complaints behind this idea, the products they came from, and similar ideas in SEO Tools.

More ideas in SEO Tools