Every proposal you will ever receive leads with domain rating, because it is one number, it is free to look up, and it sounds like quality. It is none of those things in the way the proposal implies. It is a normalised estimate of one site's whole backlink profile, invented by a tool vendor, and it was never designed to answer the question people use it to answer.
The question people are actually asking is: will this placement do anything for me? Domain rating cannot answer that, and knowing why it cannot is the difference between buying inventory and buying readership.
What domain rating actually measures
It measures the strength of a domain's inbound link profile relative to every other domain the tool has crawled, on a logarithmic scale. Three consequences follow, and all three matter commercially.
It is site-wide. A domain rating of 72 tells you the site collectively has a strong profile. It tells you nothing about the specific page your link would sit on, which may have three inbound links and forty monthly visitors.
It is logarithmic. The distance from 30 to 40 is not the distance from 60 to 70. Moving a site from 60 to 70 takes roughly an order of magnitude more acquisition. This is why "we will get you DR 60+ links" is a claim about inventory the seller already controls rather than a claim about quality.
It is a vendor's model, not a search engine's. Google does not publish domain rating and does not use it. It is a useful proxy invented to make an invisible thing legible. Proxies drift from what they proxy, and this one has drifted a long way, because the moment a number became a purchasing criterion an industry grew up manufacturing it.
What manufactured authority looks like
The pattern is consistent enough to name. A site with a high domain rating and almost no page-level traffic, publishing eleven outbound links per article under an anonymous byline, with a publishing history that begins the same year it started accepting contributions.
The authority is real in the sense that the links exist. It is not real in the sense that anyone reads the site. It was acquired to be sold, and the audience never existed.
The tell is not the domain rating itself. It is the divergence: high site authority alongside near-zero readership on the page you would appear on. A genuine publication with DR 72 has pages that people read. If the numbers separate that sharply, the number has been built rather than earned.
What topical relevance means, measured
Relevance gets asserted more than it gets measured, which makes it sound soft. It is not soft. Three things make it checkable.
Does the publication already cover your category? Not "technology" — your category. Search the site for the three terms that define what you sell. If nothing comes back, the publication does not cover you, whatever its section headings say.
Does the specific page sit inside that coverage? A trade publication that covers your category can still place you in an unrelated round-up written for search traffic. The page has to belong to the topic, not the site.
Do the surrounding links point at your neighbourhood? What a page links out to describes what it is about more honestly than what it says it is about. If a page links to four vendors in your category, it is a category page. If it links to a payday loan site and a casino, the domain rating is the least of your problems.
When DR 45 beats DR 72
The trade publication wins whenever its readers are your buyers and the general-business site's readers are not. In B2B SaaS this is most of the time, and the reason is arithmetic rather than preference.
A vertical SaaS company selling to clinical trial coordinators has a total addressable audience of perhaps forty thousand people worldwide. The general-business site with DR 72 reaches half a million readers, of whom maybe two hundred are coordinators. The DR 45 trade title reaches nine thousand readers, of whom seven thousand are.
The placement in the trade title is worth more on every dimension that matters: it reaches more of the right people, it sits in a page a buyer might actually arrive on from search, and it signals topical association to a search engine that has spent a decade getting better at understanding topics.
Sales teams notice this before analytics does. The reliable early sign that a programme is working is not a ranking change. It is that somebody in sales starts sending a placement to prospects because it makes the argument better than the deck does.
When domain rating is the right filter
Two cases, and it is worth being honest about them rather than pretending the metric is useless.
As a floor, not a target. Very low domain rating alongside low traffic and no publishing history is a reasonable first-pass exclusion when you are working through a list of six hundred candidates. It is cheap to compute and it removes obvious junk. Just do not mistake the floor for the criterion.
When the category genuinely is horizontal. If you sell project management software to everyone, the general-business title's readers really might be your buyers, and the relevance argument weakens. Horizontal categories are the exception where high-authority general placements earn their cost — and they are also the categories where everyone else is buying the same placements.
How to weigh the two in practice
Order the checks by what they cost you to run and what they can tell you.
Start with page-level organic traffic, because it is the cheapest disqualifier and it catches manufactured authority immediately. Then topical overlap, because it decides whether the placement is worth anything at all. Then the byline and publishing history, because they are quick and they catch inventory. Then reachability, because a perfect target you cannot contact is not a target. Domain rating comes last, as a sanity check on what has already survived.
Run in that order, domain rating rarely changes a decision. Run first, it makes most of them, and it makes them wrong. That inversion is the single most expensive habit in this market, and it persists because one number is easier to put in a proposal than four judgements.