> Markdown version of [/videos/802-data-mining-accessibility](https://www.wearedevelopers.com/videos/802-data-mining-accessibility). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Data Mining Accessibility Carl Groves proves most accessibility failures aren't complex edge cases, but repeating basic structural mistakes. Break the endless audit-fix cycle by shifting your testing left. - **Speakers:** [Karl Groves](https://www.wearedevelopers.com/@karl-groves) - **Event:** WeAreDevelopers LIVE - **Published:** November 3, 2023 - **Duration:** 56:00 - **URL:** https://www.wearedevelopers.com/videos/802-data-mining-accessibility ## Summary Drawing upon a dataset of nearly 15 million automated and manual accessibility errors, Carl Groves demonstrates that web accessibility struggles stem less from ignorance of esoteric edge cases and more from repeating basic structural mistakes. Analysis reveals that 90% of automatically discovered issues aggregate into just five core content types, predominantly centering on missing image alt text, poor navigation labeling, and document structure failures. Strikingly, manual accessibility audits uncover these same overarching meta-issues, debunking the common myth that automated tools flag superficial problems while manual testing discovers distinct, complex bugs. Whether caught by a machine or a human, the vast majority of web accessibility failures compromise basic content perception and operate at a severe WCAG level a or aa standard.\n\nThe single greatest differentiator between accessible and inaccessible websites is not the testing methodology used, but when that testing occurs. Sites subjected to pre-deployment evaluations boast less than half the defect density of platforms reliant on post-release audits. Organizations trapped in continuous audit-fix cycles show no historical improvement over first-time testers because they address symptoms rather than the root causes of poor software engineering.\n\nBy embracing an extreme accessibility framework, teams can break this cycle and build sustainable practices by shifting testing to the earliest phases of the software development lifecycle. Treating accessibility failures as fundamental software quality bugs rather than specialized design flaws allows developers to prevent defects using immediate feedback loops like pair programming, linting integrations, and accessibility-specific acceptance tests. Manual testing should ultimately be reserved for nuanced user experiences and dynamic interactions, rather than wasting human effort on predictable labeling and contrast defects that automation effortlessly handles. **Keywords:** automated accessibility testing, manual accessibility audits, wcag level a, wcag level aa, defect density, pre-deployment testing, audit-fix cycle, extreme accessibility, software quality metrics, test-driven development, accessibility acceptance tests, pair programming, accessible development lifecycle, shift-left testing, aria properties ## Chapters 1. **Analyzing web accessibility data at scale** (00:03) — An automated review of millions of web pages reveals patterns in how organizations code accessible interfaces. 1. **Primary content types causing automated accessibility failures** (04:28) — The vast majority of automatically discovered compliance errors stem from just five fundamental content categories like missing structural labels. 1. **Breaking down automated issues by compliance principles** (06:38) — A large percentage of discovered web defects fail baseline compliance criteria tied to basic perceivable sensory characteristics. 1. **Understanding the density of accessibility code bugs** (09:19) — Analyzing page-level error density reveals how granular unit tests identify hundreds of specific compliance failures per source file. 1. **Assessing limits of automated accessibility test coverage** (12:05) — A breakdown of software accessibility criteria highlights why only a specific subset of requirements can be verified using binary tool assertions. 1. **Common patterns discovered in manual accessibility audits** (16:10) — Representative sample testing reveals a high prevalence of severe obstacles blocking keyboard navigation and complex structural commands. 1. **Correlating manual testing findings with automated results** (23:09) — Comparing human audits with machine logs demonstrates a near-perfect overlap in discovering the most frequent interface failure patterns. 1. **Timing impact on deployed website accessibility issues** (28:23) — Integrating test suites before code deployment cuts the average number of live accessibility bugs by more than half. 1. **Escaping the continuous accessibility audit and fix cycle** (30:43) — Engineering teams can maintain permanent accessibility standards by integrating functional checks into continuous integration workflows and pair programming habits. 1. **Dealing with false positives and developer linting resistance** (39:27) — Resolving early compliance alerts inside development environments requires balancing opinionated testing logic against contextual scenarios. 1. **Navigating third-party tracking scripts and community perfectionism** (48:00) — Achieving incremental accessibility wins requires protecting validated code from injected marketing analytics tags and avoiding counterproductive gatekeeping dynamics. ## Related Moments - [Verifying accessibility compliance utilizing automated auditing and user testing](https://www.wearedevelopers.com/videos/460-future-proof-css) (from "Future-Proof CSS") - [Addressing web accessibility challenges in development teams](https://www.wearedevelopers.com/videos/649-making-interactions-accessible-to-all-users) (from "Making Interactions Accessible to All Users") - [Integrating robust accessibility practices into agile development workflows](https://www.wearedevelopers.com/videos/1340-wearedevelopers-live-performance-and-ai-social-media-decline-developer-events-and-more) (from "WeAreDevelopers LIVE - Performance and AI?, Social Media decline, Developer Events and more") - [Reviewing key learnings for continuous accessibility integration](https://www.wearedevelopers.com/videos/100170-continuous-accessibility) (from "Continuous Accessibility") - [Defining the goals of continuous accessibility testing](https://www.wearedevelopers.com/videos/100170-continuous-accessibility) (from "Continuous Accessibility") - [Building actionable and positive accessibility tooling for developers](https://www.wearedevelopers.com/videos/1792-developer-time-is-valuable-use-the-right-tools-kilian-valkhof) (from "Developer Time Is Valuable - Use the Right Tools - Kilian Valkhof") ## Related Articles - [How Developers Can Focus on Maintaining Satisfaction With Accessibility](https://www.wearedevelopers.com/magazine/109-how-developers-can-focus-on-maintaining-satisfaction-with-accessibility) - [The State of Accessibility on the Web in 2025: WebAIM Million Report Findings](https://www.wearedevelopers.com/magazine/570-the-state-of-accessibility-on-the-web-in-2025-webaim-million-report-findings) - [12 ways to fix an inaccessible web site - the CODE100 Accessibility Scavenger Hunt results](https://www.wearedevelopers.com/magazine/511-12-ways-to-fix-an-inaccessible-web-site-the-code100-accessibility-scavenger-hunt-results) - [Find all accessibility problems in a page to take part in the CODE100 final!](https://www.wearedevelopers.com/magazine/507-find-all-accessibility-problems-in-a-page-to-take-part-in-the-code100-final) ## Related Jobs - [Staff Developer Advocate, GitHub Security Lab](https://www.wearedevelopers.com/jobs/ext/1921051-staff-developer-advocate-github-security-lab) at **GitHub** - [Engineer, Offensive Security Organization](https://www.wearedevelopers.com/jobs/ext/1992296-engineer-offensive-security-organization) at **Twilio** - [QA Automation Engineer](https://www.wearedevelopers.com/jobs/ext/1414297-qa-automation-engineer) at **Bitpanda** - [Senior Web Designer, Growth](https://www.wearedevelopers.com/jobs/ext/102722-senior-web-designer-growth) at **Intercom, Inc.** - [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat** - [Staff, Analytics Engineer](https://www.wearedevelopers.com/jobs/ext/569504-staff-analytics-engineer) at **Twilio**