1. Start with the Site, Not the Score
Open the publisher's homepage and its ten most recent posts before you look at a single metric. Ask three things. Who writes here: named staff, named contributors, or nobody at all? What is the site about: one subject, a few related ones, or everything from crypto to recipes? And would your article look at home among those ten posts?
The last question is the one that decides most purchases. A placement that a regular reader would recognise as part of the site keeps its value. A placement that looks dropped in from elsewhere invites the search engine problems described in our summary of the paid placement rules. The written section at the top of each profile here answers the first two questions for you, with sources.
2. Read the Two Authority Scores as a Pair
Moz Domain Authority and Ahrefs Domain Rating both try to describe how strong a domain's backlink profile is, but they are built on different link indexes. When they agree, you have two independent readings pointing the same way. When they sit far apart, one index is counting links the other discounts or never found.
A wide gap is not a verdict, it is a reason to look closer. Check which score moved recently and whether the referring domains behind it are real publications. Never buy on a single score, and be wary of any seller who quotes only the higher of the two.
3. Test the Traffic Figure Before You Trust It
Semrush's organic traffic is an estimate of visits from search results, built from rankings. Divide it by the number of ranking keywords and you get visits per keyword, a quick test of whether the traffic is broad or rests on a handful of terms. Every profile here shows that ratio in its measured figures table.
Then ask what the estimate leaves out. Visits from people who type the address, use an app or arrive from social media are not in it. The extreme case is a search engine or a log-in tool: people open it directly, so its organic estimate can be a tiny fraction of its real audience. For an ordinary blog the gap is smaller, but the lesson holds. A low organic estimate on a site with a loyal direct audience can still be a good placement, and a high estimate on a site nobody returns to can be a poor one.
Watch for traffic that has collapsed. A publisher with plenty of keywords but almost no visits, or whose numbers fell sharply after a core update, may be selling yesterday's reach.
4. Look at Where the Links Actually Point
Three counts on every profile tell you how a site's links are spread. Backlinks to the homepage URL, to the hostname, and to the whole domain. On large sites the domain figure dwarfs the others because links spread across products and subdomains. On a small blog, check that the authority sits on the hostname where your article will appear, not on a forum or a shop on another subdomain.
Then compare total links with referring domains. A modest number of referring domains sending a huge number of links usually means sitewide links, such as footer or blogroll links, which inflate counts without adding much. Moz's Spam Score adds a third angle: it measures how much a site resembles sites Moz has seen penalised. A high score is not proof of anything, but combined with a lopsided link profile it is a reason to walk away.
5. Check the Domain's History
The registry record on each profile gives the registration date, the last change and the expiry. Read them against what the site claims about itself. A large brand that launched on an older domain it bought is harmless. The same pattern on a small site can mean something else.
Google's spam policy describes expired domain abuse as an expired name "purchased and repurposed primarily to manipulate search rankings by hosting content that provides little to no value to users". Its examples include affiliate content on a site once run by a government agency, and casino content on a former elementary school site. If a ten-year-old domain has only a year of posts, look up what used to be there before buying.
A registration that expires within weeks is worth a question too. It usually means nothing, since most owners renew, but a publisher about to let the name lapse is not one to pay for a permanent placement.
6. Make Sure It Is Still Publishing
A placement on a site that stopped publishing two years ago will be seen by very few people. The sitemaps listed on each profile show how much a site has published; the dates on its latest posts show whether it still does. Steady, recent publishing is a better sign than a burst of posts that all appeared in the same week, which often marks a site built to sell links.
7. Read Its robots.txt
Every profile lists how the publisher's robots.txt treats the main AI crawlers. This rarely changes whether you should buy, but it tells you how the publisher thinks about where its content travels. One point is often misunderstood. Blocking Google-Extended stops Google using a site's content to train Gemini, and Google says it does not affect the site's inclusion or ranking in Search. A publisher that blocks it has not hurt its own search visibility.
8. Ask the Questions Only the Publisher Can Answer
Some of what matters is not visible from outside. Before paying, ask:
- Will the post be labelled as sponsored, and how? A visible label protects readers and you.
- Which link attribute will the link carry? For a paid placement the honest answer is
sponsoredornofollow. - Who edits the article? A publisher that edits guest posts is protecting the readership you are paying for.
- Where will it appear? On the main blog or in a separate section that is hidden from the homepage, excluded from the sitemap or set to noindex?
- How long will it stay live, and what happens if the site is sold?
- Can you see two or three earlier sponsored posts? How they read and how they rank tells you more than any promise.
9. Red Flags That End the Conversation
- A "write for us" page that accepts every niche, from casinos to health to software.
- A price list that is the same for every topic and every link type, with "dofollow" as the selling point.
- Recent posts that are all outbound-link carriers with no visible readership: no comments, no shares, no authors.
- Metrics quoted from a screenshot instead of a source you can check yourself.
- Pressure to pay before you have seen the site's earlier sponsored work.
A Worked Example
Take a hypothetical home-improvement blog offered to you for a placement, with Moz DA 34, Ahrefs DR 12 and an estimated 900 organic visits a month from 2,400 keywords. The numbers are invented for illustration; the reasoning is the point.
Step 1 comes first: its last ten posts are all about renovation, written under two named authors, so your article on kitchen lighting would fit. Step 2 raises a question, because a 22-point gap between DA and DR means the two link indexes see very different profiles. Step 3 adds another: under half a visit per keyword suggests many rankings on page two or beyond rather than a real search audience. Step 4 explains both. Its referring domains are few but send thousands of links, the pattern of sitewide footer links, which one index counts and the other discounts. Step 5 settles it: the domain was registered twelve years ago, but its first posts are from last year.
None of that proves the blog is bad. It gives you three precise questions to put to the seller, and a seller who answers them well is worth more than one with a higher score and no answers.
10. Only Then, Compare Price with Evidence
Price comes last because it only means something against what you have found. Two sites at the same price can be very different buys once you know their audience, their link profile and how they treat sponsored work. Shortlist two or three publishers from the directory or a category hub, run this routine on each, and then contact us for current pricing. Each number used above is defined on Metrics explained.