---
title: "How to read a site report — mxverify"
url: https://mxverify.uk/guide
description: "What each figure in an mxverify site report means — BV, link levels, relevance, links, traffic, AI assistant visits — and what it cannot tell you."
site: mxverify
generated: 2026-11-12T05:30:35.958Z
---

# How to read an mxverify site report

Written by [Palash Bagchi](https://mxverify.uk/about), who maintains mxverify. How each figure is worked out, rule by rule, is on the [methodology](https://mxverify.uk/methodology) page.

Every mxverify site report shows the same parts, in the same order. This guide says what each part means, what it cannot tell you, and how to use it when you judge a site for a link. A report's first fold, "How to use this report", applies this guide to that site's own figures, where they are on record.

## Reading backlink and traffic reports

### How to read a backlink report: the Ahrefs guide, and this one

Ahrefs' own guide, [How to Do a Basic Backlink Audit (in Under 30 Minutes)](https://ahrefs.com/blog/backlink-audit/?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide), reads a backlink profile in three steps: a high-level view of backlink and referring-domain counts and how they grew, a deeper look at the best and worst referring sites, then the full Backlinks report. An mxverify report answers a narrower question: what one new link from the site is worth, and why. Its "Links to this site" card gives a third-party backlink index's counts; Ahrefs' own counts are in Ahrefs' tools.

### How to interpret a backlink report: the Moz guide's metrics, and mxverify's

Moz describes [Domain Authority](https://moz.com/learn/seo/domain-authority?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide) as a score from 1 to 100 that predicts how likely a site is to rank compared with competitors, and [Spam Score](https://moz.com/help/link-explorer/link-building/spam-score?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide) as the percentage of sites with similar features that Moz has found penalized or banned by Google. mxverify shows neither. To interpret an mxverify report, start from the BV score, then the link level the site usually gives, then how its links behaved where links placed on it were audited.

### How to read a traffic report: Google Analytics, and a beginners guide to the estimate

Google Analytics' [Traffic acquisition report](https://support.google.com/analytics/answer/12923437?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide) shows where a site's new and returning users come from, grouped by channel, from the site's own data, which only its owner sees. An mxverify report shows a third-party modelled estimate instead: monthly visits, the channels they came from and the top search keywords, to compare sites rather than measure one. BV uses only the organic search figure.

### Backlink metrics explained: Domain Rating, Domain Authority, referring domains, Spam Score and Majestic

Domain Rating is Ahrefs' 0 to 100 score of the relative strength of a site's backlink profile, on a logarithmic scale; Domain Authority is Moz's 1 to 100 ranking-likelihood score; Majestic's Trust Flow and Citation Flow score quality and link equity from 0 to 100. Referring domains are the websites linking to a site, each counted once, which a report shows as how many websites link to it. The spam score on a report is the backlink index's own score of the pages linking to the site, not Moz's Spam Score. mxverify reads only Domain Rating, as half of Site Strength, and does not show the rating itself.

### How to read a backlink report on mxverify

Read it from the top: the BV score is what one new followed link is worth, the level says what kind of link the site usually gives, and the relevance range shows what your topic does to it. Then check "How its links behaved" for audited links, "How it links out" for the site's nofollow share, and "Links to this site" for who links to it. A report's "How to use this report" fold applies these steps to the site's own figures.

### How to read a website traffic report for SEO

For SEO, read the organic search share first: it is the traffic search rankings send, and the only traffic figure BV uses. The monthly chart shows the trend over the months the estimate covers, and the top search keywords show what the site is found for — searches and CPC describe the keyword's whole market, while Visits (est.) describes the site. A top search keyword is not a position in Google.

## The BV score

The box at the top of a report is the site's BV score: what one new followed link from the site is worth to rankings, in US dollars, under BV 2.9. It is not a price anyone charges, and it does not predict what a link will do for a particular site. It comes from two things: how strong the site is (Site Strength) and what kind of link it usually gives (its level).

## Link levels A to D

The same site is worth very different amounts depending on the link it gives. A report prices a link at the level the site usually gives, and shows what each level would be worth there:

- **A — Counted** (×1): An editorial citation: the site linked because it is useful to its readers.
- **B — May be discounted** (×0.5): A guest, sponsored or placed article, or a product or tool directory listing.
- **C — Usually neutralised** (×0.1): A press release or syndicated copy, a forum post, profile, footer or sidebar link — and any nofollow link.
- **D — Network** (×0.05): A site on a server whose sites show footprints of one owner, a link farm, or machine-generated pages.

"Worth up to" is the level A figure: a followed editorial link at the site's strength, the most a link from it can be worth before relevance. Two caps can lower the level a site is priced at, and the report names them after the level ("level A · nofollow"): **nofollow**, when a share of the site's links carry nofollow, sponsored or ugc, which BV counts as passing no ranking credit (search engines treat them as hints), and **high-risk niche**, when the site is likely an adult, gambling, pharmacy or similar site, whose links to unrelated sites search engines discount.

## Relevance

The BV score is worked out at ×1, as if relevance were neutral. A link is worth more from a site on the same topic as the site it points to, and less from an unrelated one:

- Unrelated ×0.5
- Loosely related ×0.75
- Adjacent ×1
- Closely related ×1.1
- Same topic ×1.25

Each report prints both ends of that range beside the score. A same-topic link is worth 2.5 times an unrelated one from the same site, so a smaller site on your topic can be worth more than a stronger one that is not.

## Site Strength

Site Strength runs from 0 to 100, in steps of 5: half the site's authority rating (Ahrefs Domain Rating, which mxverify uses and does not show), half its estimated organic search traffic on a log scale. It sets how much a link from the site can be worth; it is not a verdict on the site.

## How its links behaved

A site is marked audited when links placed on it have been checked by an audit on [bklink.uk](https://bklink.uk/?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide): each page fetched and the link read — still there or removed, followed or nofollow. These are the report's only measured figures about links placed on the site. Percentages are of the links that could be read, and appear from 3 audited links; below that the counts are shown alone. A high removed-or-dead share says links were taken down or their pages lost, not that the site marks its links nofollow. The audited tick is a record of checks, not an endorsement of the site or of who runs it.

## How it links out

mxverify reads a site's homepage and up to ten of its articles and counts every link they make to other sites: followed, nofollow, sponsored or ugc. The share that is nofollow, sponsored or ugc says how often the site marks its links to other sites as passing no credit. Each reading is graded strong, fair or thin by how many articles and links it found. The reading is a sample of what the site publishes, not every page.

## Links to the site

"Links to this site" counts the links other sites have made to the site, from a third-party backlink index: how many websites link to it, how many links in all, how many of those websites link with nofollow, and how many IP addresses they link from. Many linking websites on few IP addresses can be a sign of a link farm.

The **spam score of its links** is the index's own score, from 0 to 100, averaged over the pages that link to the site. It describes those linking pages, not the site itself, and a low score does not clear a site: it says only that the pages linking to it look ordinary to the index. "Where the links sit" is where on the linking pages the index found the links, and describes placement only. None of these figures moves BV.

## The traffic estimate

Traffic figures are modelled estimates from a third-party traffic source, for the whole site and all its visitors in one month. No one outside a site sees its analytics, so treat them as an estimate to compare, not a measurement. Where the estimate has no data on a site it gives zeros; a report shows "no data" there, never a zero as a figure. BV uses only the organic search figure.

The **top search keywords** table has four figures per keyword:

- **Searches / mo**: how often the keyword is searched a month, by everyone, on every site.
- **CPC**: what an advertiser pays for one click on a search ad for it.
- **Ad value / mo**: searches × CPC, what a month of all the keyword's searches would cost as search ads. It is the whole market's figure: not the site's traffic, not what the site earns, and nothing about how strong the site is.
- **Visits (est.)**: the estimate's own figure for the keyword on this site, which behaves like visits a month. This is the column that says something about the site.

A top search keyword is not a position in Google. The monthly chart shows how visits moved over the months the estimate covers; a report's "How to use this report" fold names the trend from 3 months.

## Potential display-ad revenue

A report estimates what a site could plausibly earn a month from display ads: monthly pageviews ÷ 1,000 × a page RPM for its niche, shown as a range. Pageviews are the traffic estimate's visits × its pages per visit. The page RPM (dollars per 1,000 pageviews) is a benchmark by niche: finance, business and tech $15–$40+; health and wellness $3–$10; e-commerce and shopping $2–$7; travel and lifestyle $1.50–$5; entertainment and news $1–$4. Those rates are for visitors from the US, the UK, Canada and Australia; pageviews from other countries are counted at 20% of the rate, since display ads there earn several times less.

The niche is our reading of the site's articles where we have one, else its content category. A report offered to search engines always gets an estimate, and says what it stands in for: a site whose niche we have not identified is priced at a general $1–$5; a company or platform site that runs no display ads, at what it could earn if it did; a site with no country split, with every pageview at the other-countries rate; a site in a niche mainstream ad networks do not serve, at the lowest band. Any other report gets one only when none of those applies. A site with no traffic gets none. The estimate is worked out once a month, on the 10th. It is a benchmark, not the site's earnings: it does not know which ad networks the site uses, how viewable its ads are or how many a page carries, and real rates vary widely by network and season. It is also not the keyword table's ad value, which is the whole market's search-ad spend. BV does not price on it.

## AI assistant visits

The AI assistants section counts visits: people who clicked through to the site from an AI assistant's answer. It gives the share of all visits that came from assistants, and each assistant's share of those visits month by month. It is not a count of how often an assistant cites, names or recommends the site. The report's only citation figure is Google AI Overviews, under "Links to this site".

At small volumes the split swings hard from month to month: a handful of visits can move an assistant from 0% to 100%. A report's "How to use this report" fold mentions AI assistants only when they send at least 2% of the site's visits and at least 1,000 visits a month, and says when one assistant's share moved 20 points or more between months. BV does not price on these visits.

## Server and network

Each site is resolved to the address it serves from. Publications on the same address are listed together, and where 10 or more share one, the server is checked for footprints of one owner: shared private nameservers, registrations inside one 90 days, one registrar, one site theme, the same pattern of audited links. A server is called a network only when at least two of them agree, one of them nameservers, registration dates or theme, and BV values a link from a network's site at level D. Links placed across one network are one relationship, not many. A shared address alone is ordinary on shared hosting, and CDN addresses, which front unrelated sites, are never grouped.

## Page speed

Page speed is how the site loads for real Chrome users over 28 days, from the Chrome UX Report (Google), CC BY 4.0: LCP (how long the main content takes to appear), INP (how quickly the page responds to a tap or click) and CLS (how much the layout shifts while it loads), and whether it passes Core Web Vitals. It is a fact about the experience of the site's readers. BV does not price on it, and it says nothing about what a link from the site is worth.

## Domain and hosting

The registration date, expiry and registry status come from the public registration record; the content category is mxverify's own classification of the site's pages. They describe the site. BV does not price on a domain's age.

## Similar sites

Under its cards a report lists up to six similar sites, chosen once a day from what the reports themselves hold. A site is listed only when it shares a fine subject with the report's site (payments, grain elevators, product launch platforms, local TV stations and some sixty more), read from the words of its homepage, its top search keywords and its pages' titles, or when the two use much the same words. A shared category alone never makes a pair, and nor do language, country or technology. Where the report's site has words of its own, a model then reads them beside the words of about twenty candidates (those sites, and the sites closest to it in words or profile), and the report lists only the candidates the model judges to be on the same subject or industry as the report's site, whatever their market or kind of customer, the most likely first; until it has read them, the report keeps the list above. Each card names the subject the two share, or says "Same subject" where the model judged them one subject and their words name none in common. Among sites equally close on subject, the one nearer in BV comes first. A news site with fewer than three such matches lists other news sites instead, each card marked "Also a news site"; any other site with fewer than three shows none.

A site is never listed beside one on the same server, address block or listed hosting network, one with the same private nameservers or registered domain, one built from the same homepage template, a page on a free host, or a site placed in the same customer campaign. A site on a network's server, or valued at level D, is listed only on the reports of other such sites, and a site in a high-risk niche only on another's. No site is listed on more than 40 reports, every site of one registered domain counted together. Each card shows the site's own BV and level from its report. A similar site is not a recommendation to place a link there; its report is the evidence to weigh. The line under the cards, "More like it", links the lists of sites that share this report's level, BV band, category, CMS or registration decade.

## Using a report to judge a placement

- **Price check.** Take the report's figure for the level of link on offer and for how close the site is to yours: the same-topic end if it is on your topic, the unrelated end if it is not. Where a quote is far above BV's same-topic figure for that level, the difference pays for something BV does not measure, such as the site's readers or the vendor's margin. That can be worth paying for. On a site in a high-risk niche, the report's figures are for a linked site outside the niche; a site in the same niche is not capped, and the report gives its figure too.
- **Ask for the level.** BV prices the level the site usually gives. An editorial link (level A) is worth the "worth up to" figure before relevance; a sponsored or guest post is level B at best, and any nofollow link level C. Where BV sets the level by rule (a network's site, a platform, a press-release wire, a search engine), no level A link is on offer.
- **Qualify the site.** Read the level, the caps, the audited shares (followed, nofollow, removed or dead), the reading's nofollow share, the server and the spam score of its links together. None of them alone is a verdict on a site, and none of the links-to-the-site figures moves BV.
- **Check the traffic claim.** Set the estimate beside what the site or a vendor claims, and ask about a large gap. The estimate is modelled, not the site's own analytics.
- **Check the link once it is placed.** Followed, still on the page, in the article: an audit on [bklink.uk](https://bklink.uk/?utm_source=mxverify.uk&utm_medium=referral&utm_campaign=guide) checks each one, and its results join the site's report.

## Limitations

- BV is a model of ranking value, worked out the same way for every site. It is not a price, and it does not predict what one link will do for one site.
- Traffic, links to the site and AI assistant visits are third-party estimates, each dated in the report's "Where each figure comes from" fold. They can be far off for a small site.
- A reading of a site is a sample of its homepage and up to ten articles, not every page it publishes.
- A report is evidence to weigh, not an endorsement of the site or a check of who runs it.

## Primary sources

What a report says about how search engines and browsers treat links and pages rests on these, in the words of their publishers:

- [Google Search Central: qualify your outbound links to Google](https://developers.google.com/search/docs/crawling-indexing/qualify-outbound-links) — nofollow, sponsored and ugc as hints about how a link should be treated, the basis of the nofollow cap and the reading's nofollow share.
- [Google Search Central: spam policies for Google web search](https://developers.google.com/search/docs/essentials/spam-policies) — link spam and what Google treats as an attempt to manipulate rankings, the background to the network and high-risk niche rules.
- [web.dev: Web Vitals](https://web.dev/articles/vitals) — what LCP, INP and CLS measure and the levels Google calls good.
- [Chrome UX Report documentation](https://developer.chrome.com/docs/crux) — how the real-user page speed figures are collected.
- [RFC 9083: JSON responses for the Registration Data Access Protocol (RDAP)](https://www.rfc-editor.org/info/rfc9083/) — the form of the public registration record behind a domain's registration date and expiry.

These describe the platforms' own rules. BV is mxverify's model of ranking value, not Google's, and none of these pages says what a link is worth.

To report an error in a report, write to [support@bklink.uk](mailto:support@bklink.uk) with the report's address.

---

Source: mxverify, "How to read a site report — mxverify", https://mxverify.uk/guide (retrieved 2026-11-12). How every figure is worked out: https://mxverify.uk/methodology