How to Vet a Website Before Buying a Guest Post: 15 Checks
Most guest post budgets are spent on sites that will never help the buyer rank. They look clean, quote impressive metrics and promise fast placement, but behind the media kit there is often no real audience and no organic traffic. The link goes live, the invoice gets paid and your rankings do not move.
This guide walks through 15 checks in the order we run them, with what good looks like for each and the point where you walk away. It is the same discipline we apply inside our guest post services before a single article goes to any publisher.
Why Vetting Matters More Than It Used To
Guest posting itself is not the problem. Search Engine Journal puts it well: there is a high effort, valuable version and a low effort, spammy version, and the tactic turns toxic when one side gets lazy (Search Engine Journal). Google draws the same line: paid articles with links that pass ranking credit, or optimized anchor text in guest posts, are link spam, and payment is only within the rules when the link carries rel=”nofollow” or rel=”sponsored” (Google Search Central).
Editorial.Link surveyed 518 SEO professionals in 2026 and found that 55.98% believe Google is not effective at identifying and discounting paid links (Editorial.Link). A bad placement may never trigger a penalty. It will simply sit there doing nothing while you paid for it, which is why every check below starts from the reader, not the metric.
If you are still deciding which link format fits your campaign, our comparison of guest posts, link insertions and digital PR explains when each format earns its place.
What You Need Before You Start
Four things: an SEO tool with a link index such as Ahrefs or Semrush; SimilarWeb, even on the free tier, for total traffic estimates and traffic by country (Ahrefs); a free Moz based Domain Authority and Spam Score checker such as Keyword Everywhere's (Keyword Everywhere); and Google itself, where the site: operator and ten minutes of honest reading catch things no dashboard shows.
Traffic and Visibility
Check 1: Real Organic Traffic and Its Trend
What good looks like: steady or rising organic traffic across 2 or more years. Contentellect treats a sharp drop that lines up with a Google core or spam update as an automatic fail, because a single month of data can hide a collapse (Contentellect).
Walk away if: traffic is near zero while the site quotes strong metrics, the trend shows a cliff matching a Google update, or the site claims big traffic that SimilarWeb cannot see at all.
Check 2: Traffic Geography Matches Your Market
What good looks like: most visitors come from the countries you sell in. If your customers are in the US and most of the site's visitors are elsewhere, the link sends few relevant readers (Contentellect). Check the country breakdown in Ahrefs or SimilarWeb's free tier (Ahrefs).
Walk away if: the geography is completely mismatched with your market.
Check 3: The Site Is Indexed in Google
What good looks like: the homepage and recent articles appear in Google. Type site:example.com with the real domain into Google and confirm recent posts show up.
Walk away if: only the homepage is indexed, recent posts are missing, or the site has thousands of indexed pages but no organic traffic. That last pattern points to scaled thin content, which Google's spam policies cover under scaled content abuse (Google Search Central).
Authority and Link Profile
Check 4: Authority Metrics, Read the Right Way
Domain Authority from Moz and Domain Rating from Ahrefs measure the strength of a site's link profile, not Google trust, and Google does not use them. DA runs from 0 to 100 on a logarithmic scale, and a score is only meaningful compared with the sites you compete against (Keyword Everywhere). Contentellect uses DR 20 as a first filter, while noting that DR only shows a site has links, not that anyone visits it (Contentellect).
Walk away if: the site has high DA or DR with near zero organic traffic. Good authority metrics without traffic is one of the strongest red flags in this guide.
Check 5: Link Profile Quality and Anchor Text Patterns
Open the referring domains report, not the raw backlink count. One domain linking ten times is one editorial vote, not ten. Look for links from real sites in related niches, a natural mix of branded anchors and varied phrases. Search Engine Land notes that PBN style link schemes lean on manipulative anchor text and links that lack contextual relevance (Search Engine Land).
Walk away if: the anchor profile is dominated by exact match commercial anchors, or the referring domains are obviously unrelated or spammy.
Check 6: Outbound Link Ratio and Link Farm Signals
Sites that sell links link out far more than they are linked to. Contentellect compares a site's linked domains, the sites it links out to, with its referring domains, and treats a heavy imbalance toward outbound linking as a classic seller signal (Contentellect).
Walk away if: several recent posts each carry multiple commercial outbound links to unrelated industries, or a checker flags unusually high link propensity, which can point to a scraper or low quality site (Keyword Everywhere).
Editorial Quality and Relevance
Check 7: Topical Relevance to Your Niche
Relevance comes first in Contentellect's RAD formula, which stands for Relevance, Authority and Due Diligence. For a travel client in the yachting niche, a post on a technology blog makes far less sense than one on a travel or sailing site, because references from your peers carry more weight (Contentellect). Search Engine Journal makes the same point from the risk side: links from domains that do not make sense for your niche bear investigation (Search Engine Journal).
Walk away if: the site publishes across unrelated verticals in the same week, finance tips next to crypto wallets next to home remodelling, with no editorial thread. That is inventory, not a publication.
Check 8: Editorial Standards and Real Authors
What good looks like: named authors with real bios, visible contributor guidelines, and evidence that someone actually edits. Search Engine Journal's rule for a natural link is that it is an editorial choice, and a site with no editors cannot make editorial choices (Search Engine Journal).
Walk away if: every post is by a generic Admin account or rotating one off names with no bios, or the write for us page promises links to anyone who pays.
Check 9: About and Contact Pages That Exist
What good looks like: an about page that explains who runs the publication, plus a contact page with real names. Contentellect's due diligence step treats real authors and real about and contact pages as proof of legitimacy (Contentellect).
Walk away if: there is no about page at all, or the contact page is a bare form with no names attached.
Check 10: Content Quality and Publishing Frequency
Read the last 10 posts and watch for unrelated topics side by side, sponsored labels on most posts, and commercial anchors pointing to unrelated sites. Then ask: are all the articles guest posts. If they are, run, because the site exists solely to sell placements (Contentellect).
Walk away if: the archive is wall to wall sponsored posts, or publishing happens in bursts of paid content with long silences between.
Trust and Authenticity Signals
Check 11: Design and UX Signals
What good looks like: a coherent design, working navigation, pages that load cleanly and a layout that serves readers rather than ad slots. Be honest about this check's limits: a clean theme costs almost nothing now, and weak sites often look polished. Design is a supporting signal, never a verdict. Search Engine Journal's PBN glossary lists similar design, navigation and colour scheme across suspected sites as a check worth running when you compare several prospects side by side (Search Engine Journal).
Walk away if: the site runs a default template with no customisation, or the page is buried under aggressive ads and popups.
Check 12: Social Presence and a Real Audience
What good looks like: social profiles with genuine activity, comments or shares on articles, and a newsletter with a real archive. Followers can be bought and engagement can be faked, so treat social presence as supporting evidence only. But a site claiming tens of thousands of monthly readers with no comments, no shares and no named authors anywhere is telling you the audience does not exist.
Walk away if: the social icons link nowhere, or there is zero visible reader interaction across the whole site.
Check 13: Spam Score Style Red Flags
Moz's Spam Score runs from 0 to 100 and counts spam signals on a domain, with higher meaning more signals (Keyword Everywhere). Free checkers also show link propensity: how much a domain links out relative to normal.
Walk away if: the spam score is high and other checks are already wobbling. Spam score alone is not a verdict: plenty of legitimate sites trip a few signals. It only becomes a walk away when it agrees with what the other checks already told you.
Risk and Commercial Sense
Check 14: PBN Footprint Detection
A private blog network is a group of websites, owned or controlled by one entity, that exist primarily to pass link equity to a money site and manipulate rankings (Search Engine Land). Search Engine Journal is blunter: a PBN's goal is link building alone to manipulate search engine rankings, which is what makes it a spam tactic (Search Engine Journal).
Check the domain's history first: an expired domain rebuilt as a blog or a sudden switch of topic is a red flag (Contentellect), and Google's spam policies cover repurposed expired domains under expired domain abuse (Google Search Central). Then check hosting and ownership: shared IP addresses, reused analytics and tracking codes, and repeated registration data are the key technical tells (Search Engine Land). Finally compare design and content across sites, including duplicate content found by searching a distinctive sentence in quotes.
Walk away if: two or more of these signals point the same way. One signal alone proves nothing: legitimate networks exist, and the difference is transparency and intent (Search Engine Land).
Check 15: Price Versus Quality Sanity Check
Editorial.Link's 2026 survey of 518 SEO professionals found the average acceptable price for one high quality backlink was $508.95, and BuzzStream's 2026 analysis of 257,267 sites found a guest post costs $295 on average bought directly from the site and $461 through a vendor (Editorial.Link, Contentellect). For a deeper breakdown, see our guides to what a guest post typically costs and how much link building costs overall.
Walk away if: the price is far below market for the claimed metrics. Cheap plus high DR is the classic signature of inflated metrics on a site with no audience.
Red Flags That Mean Walk Away Immediately
| Red flag | Why it kills the deal |
|---|---|
| Zero organic traffic paired with strong authority metrics | The authority is manufactured, not earned |
| Traffic collapsed after a Google update and never recovered | The site was caught doing something Google disliked |
| Every article is a guest post or sponsored post | The site exists to sell links, not to serve readers |
| Commercial anchors to unrelated industries in every post | Your link will sit in a link farm neighbourhood |
| No about page, no named authors, no real contact details | There is no publication behind the placement |
| Expired domain repurposed with unrelated content | Google treats this as expired domain abuse |
| Two or more PBN footprint signals | You are buying into a link scheme |
| Price far below market for the claimed metrics | Inflated metrics on a site with no audience |
Why We Vet Every Site Before We Place a Post
Every site in our outreach list passes these 15 checks before a pitch goes out, and clients see the vetted list before anything is bought. That is slower than blasting a vendor spreadsheet, and it is why our placements hold their value through updates.
If you want placements on real publications with real audiences, without running this checklist yourself, talk to us about our guest post services. We will show you the sites first, and you approve the list before we spend a dollar. If you would rather compare providers first, our roundup of the best guest posting services for 2026 explains what separates serious providers from resellers.
Frequently Asked Questions
What is the fastest way to check if a guest post site has real traffic?
Paste the domain into Ahrefs or Semrush and look at the organic traffic trend over 2 or more years, then cross check total traffic and top countries in SimilarWeb. If both tools show little or nothing while the site claims a large audience, the audience does not exist.
What DA or DR should a guest post site have?
There is no universal number. Compare the score with real sites in your niche, and never read DA or DR without traffic. Contentellect uses DR 20 as a first filter, then checks traffic: a metric without visitors is the classic sign of an inflated profile.
How can you tell if a site is a PBN?
Look for footprints, not single clues: shared hosting or IPs, reused tracking codes, hidden or repeated ownership records, similar designs across sites, duplicate content, and high metrics with no organic traffic. Two or more matching signals is the point where suspicion becomes a decision.
Are paid guest posts against Google's guidelines?
Paying for a placement is not automatically a violation. Google's spam policies treat it as link spam when the article's links pass ranking credit or use optimized anchor text without a rel=”nofollow” or rel=”sponsored” qualifier.
What is a good price for a guest post in 2026?
Editorial.Link's 2026 survey put the average acceptable price for one high quality backlink at $508.95, while BuzzStream's 2026 analysis found guest posts average $295 direct from publishers and $461 through vendors. Treat these as market context, not quotes.
Should I buy guest posts on sites outside my niche?
Only if the audience overlap is genuine. A reference from a peer site in your niche carries more weight, and links from irrelevant domains invite closer scrutiny. If the site covers many unrelated verticals with no editorial thread, it is selling inventory, not audience.
How many failed checks should make you walk away?
Two clear failures. One uncertain result is a question to investigate, two failed checks is a pattern. A single bad placement costs money and adds nothing, so walking away early is the cheaper option.