Authority metrics help you filter quickly, but they do not answer the most important question: does this site have a real editorial life? Before buying a permanent backlink, evaluate why the site exists, how it earns traffic and what happens to published pages after delivery.
1. Search visibility and audience
Start with visibility, not price. A real publication has pages that rank for topical queries, attract readers and get updated over time. A shell may show link metrics while having weak organic demand and thin architecture.
- Check which sections bring visibility.
- Compare traffic across home, category and article pages.
- Look for sudden drops after waves of commercial posts.
- Check whether the brand itself has search demand.
2. Content quality
Open the latest 20 posts. If topics are random, every article follows the same structure and each page points to a commercial destination, the risk is high. Real publications have categories, recurring editorial patterns, internal links and posts that are useful even without the external link.
| Signal | Real publication | Link-selling shell |
|---|---|---|
| Topic | Clear sections | Random topic mix |
| Posts | News, reviews, guides | Mostly paid articles |
| Links | Relevant and contextual | The link is the point of the article |
| Authors | Named editors or team | No authors or fake profiles |
3. Outbound link pattern
The problem is not simply having outbound links. Real sites cite sources, brands and partners. Risk starts when each new post points to unrelated commercial pages, risky verticals and thin affiliate destinations.
If a donor accepts every topic without editorial rules, its value can decay quickly even when old metrics look strong.
4. The placement page
Evaluate the future page, not only the domain. Ask where the article will live, whether it appears in a relevant section and whether internal links can reach it. A page with no internal support and many unrelated external links rarely delivers what domain-level metrics imply.
Decision scorecard
- Use it: the site is topically close, has visibility, publishes real content and keeps outbound links under control.
- Review manually: metrics look good, but commercial posts are frequent. Inspect recent articles and placement terms.
- Reject it: the site publishes anything, hides placements or sells links into unrelated risky niches at scale.
How to use this in Trastik
Filter donor sites by topic, GEO, traffic and price. Then manually review the shortlist: recent posts, outbound links, indexability and the future placement page.
An example review record
Synthetic example: two anonymous publishers have similar headline metrics. The observations below are invented to show how a reviewer can record evidence; they are not Trastik inventory or client results.
| Observation | Publisher A | Publisher B |
|---|---|---|
| Recent content sample | Several relevant articles with distinct sources and named editorial responsibility | Repeated templates leading to unrelated commercial destinations |
| Article discovery | Sampled pages are linked from relevant sections | Sampled pages cannot be reached from the visible navigation |
| Placement terms | Format and destination requirements are available for review | The proposed placement has no clear location or stated requirements |
| Decision | Continue the manual review; approval is still unresolved | Pause the purchase until the missing evidence is resolved |
No single row proves that a publisher is good or bad. Keep the sampled URLs, review date, evidence source and unresolved questions in an internal worksheet. Metrics without a date should be marked as undated; missing audience data should remain unknown.
Quality evidence and search policy are separate
A maintained publication can still carry a paid link that falls outside a search engine’s rules. Conversely, advertising content alone does not prove a site is a shell. Google distinguishes advertising with appropriately qualified links from buying links for ranking credit. Read its link spam policy and assess the proposed placement, not just the domain’s score.
This guide helps evaluate placement quality. It does not promise ranking gains. Outcomes depend on competition, content, technical health and the broader link profile.