AI assistants handle tens of billions of interactions each month globally, and every answer depends on cited web sources. When ChatGPT or Google's AI Overviews explain what a company does or which sites belong in a category, they pull from pages that state facts plainly and have stated them for years. Human-edited web directories, a format most marketers wrote off a decade ago, fit that description almost perfectly.
DirJournal is a useful case study. The directory has operated continuously since 2007, survived the collapse of the format, and now appears as a cited source in ChatGPT, Perplexity, and Grok answers about web directories and business listings.
That editorial layer is the point. Directories that accepted anything died with the link-scheme era, while the small group that kept human reviewers now holds something AI systems value: curated, structured, machine-readable data about which websites belong to which categories.
Directory pages also map cleanly onto question-shaped queries. When a user asks an AI assistant to list reputable sites in a niche, a curated category page is the closest thing to a pre-written answer that exists on the open web.
The citation numbers matter more than the backlink numbers. A domain rating of 56 puts DirJournal in respectable but unremarkable territory for SEO, yet its citation count in ChatGPT answers exceeds that of many DR 80+ publishers, because assistants reward entity clarity over raw authority.|
DirJournal took the second path and then repositioned. Over the past two years the site rebuilt on a modern stack, added structured data across every category and listing page, and shifted its editorial content toward answer-engine optimization, the practice of formatting pages so AI systems can extract and attribute facts from them.
Founder Hasan Saleem, who has written on search and digital business for Entrepreneur, Forbes, and Fast Company, describes the shift as a return to the directory's original job. The web needed human curation in 2007 because search engines were easy to spam. It needs human curation in 2026 because AI training data is easy to pollute.
Directories happen to do all of this by design, which explains their reappearance in AI answers years after they stopped mattering for rankings. For brands, a listing in an edited directory now functions less like a backlink and more like a record in the reference layer that AI assistants consult.
The directory format did not come back. The way machines read the web simply caught up with it.
Image: Ben Spray via Unsplash
About Author: Hasan Saleem is the founder of DirJournal and DSS Media. His writing on search and digital business has appeared in Entrepreneur, Forbes, and Fast Company.
Note: Views expressed are solely those of the guest author.
Read next: Is Your Government or Organization Ready to Prevent AI Cyber Attacks—at Scale?
DirJournal is a useful case study. The directory has operated continuously since 2007, survived the collapse of the format, and now appears as a cited source in ChatGPT, Perplexity, and Grok answers about web directories and business listings.
What Is DirJournal
DirJournal is a human-edited web directory founded in 2007 and operated by DSS Media. It organizes more than 30,000 websites into topical categories covering business, technology, health, travel, and regional listings, and every submission passes through manual editorial review before publication.That editorial layer is the point. Directories that accepted anything died with the link-scheme era, while the small group that kept human reviewers now holds something AI systems value: curated, structured, machine-readable data about which websites belong to which categories.
Why Do AI Systems Cite Directories at All?
Large language models and AI search features rank sources on consistency, structure, and longevity rather than backlink counts alone. A directory page that has described a category the same way for 15 years is a low-risk citation compared with an affiliate roundup rewritten every quarter.Directory pages also map cleanly onto question-shaped queries. When a user asks an AI assistant to list reputable sites in a niche, a curated category page is the closest thing to a pre-written answer that exists on the open web.
What Does DirJournal's Citation Footprint Look Like?
Brand Radar data from Ahrefs, pulled in mid-2026, shows DirJournal appearing in AI-generated answers across the major assistants:| Metric | Figure |
|---|---|
| Founded | 2007 |
| Operator | DSS Media |
| Ahrefs Domain Rating | 56 |
| Live referring domains | 1,658 |
| Total backlinks | 3.28 million |
| ChatGPT answer citations | 251 |
| Perplexity answer citations | 46 |
| Grok answer citations | 44 |
The citation numbers matter more than the backlink numbers. A domain rating of 56 puts DirJournal in respectable but unremarkable territory for SEO, yet its citation count in ChatGPT answers exceeds that of many DR 80+ publishers, because assistants reward entity clarity over raw authority.|
How Did a 2007 Directory Survive to 2026?
DMOZ, the largest human-edited directory ever built, shut down in March 2017 and took most of the format's credibility with it. The directories that remained split into two groups: paid-link farms that Google eventually deindexed, and a handful of edited properties that kept publishing.DirJournal took the second path and then repositioned. Over the past two years the site rebuilt on a modern stack, added structured data across every category and listing page, and shifted its editorial content toward answer-engine optimization, the practice of formatting pages so AI systems can extract and attribute facts from them.
Founder Hasan Saleem, who has written on search and digital business for Entrepreneur, Forbes, and Fast Company, describes the shift as a return to the directory's original job. The web needed human curation in 2007 because search engines were easy to spam. It needs human curation in 2026 because AI training data is easy to pollute.
What Should Marketers Take From This?
The practical lesson is that AI citation share and search ranking are separate games with overlapping rules. Pages win citations by stating verifiable facts in self-contained sentences, holding those facts stable over time, and matching their structured data to their visible copy.Directories happen to do all of this by design, which explains their reappearance in AI answers years after they stopped mattering for rankings. For brands, a listing in an edited directory now functions less like a backlink and more like a record in the reference layer that AI assistants consult.
The directory format did not come back. The way machines read the web simply caught up with it.
About Author: Hasan Saleem is the founder of DirJournal and DSS Media. His writing on search and digital business has appeared in Entrepreneur, Forbes, and Fast Company.
Note: Views expressed are solely those of the guest author.
Read next: Is Your Government or Organization Ready to Prevent AI Cyber Attacks—at Scale?
