DEV Community: Karma Domains

Karma.Domains adds three new sources: .id premium names, .ee registry auctions and .ro buy-now listings

Karma Domains — Tue, 03 Nov 2026 07:08:44 +0000

If you hunt for domains in country-code zones, this update saves you three extra marketplaces to check. Karma.Domains now pulls in domain.id Premium, Internet.ee registry auctions and EUReg buy-now listings, so you see price, end time and SEO metrics for them in the same list as everything else.

What was added:

domain.id Premium (PANDI): about 970 premium .id domains for buy now, priced in IDR.
Internet.ee Auctions: about 60 blind auction lots from the Estonian .ee registry.
EUReg (Buy Now): about 3,900 .ro domains for sale in EUR. Listings without a fixed price show as "Make offer".

To see only one marketplace, open Buy Now or Auctions in Karma.Domains and use the source filter. The team says more sources are on the way.

Full note: https://karma.domains/en/updates/new-sources-03-nov-26/

Originally published on karma.domains

Semrush vs Similarweb Numbers Don't Match? How to Compare Competitor Traffic in 2027

Karma Domains — Fri, 30 Oct 2026 02:11:25 +0000

You look up three competitors in Semrush, then in Similarweb, and the numbers don't line up. One site gets 410,000 visits a month in one tool and 260,000 in the other. Sometimes the site that leads in Semrush comes second in Similarweb. It's tempting to pick the tool you trust more, or to average the two and move on.

I'd do neither. The absolute numbers from these two vendors will rarely match, and that by itself tells you very little. What matters for competitor research is whether the two tools agree on the shape of the market: who is bigger, by roughly how much, and who is growing. You can usually answer that even when the totals are far apart, as long as you set up the comparison carefully.

Here is the process I use.

First, check that you're comparing the same number

Many of the "Semrush and Similarweb disagree" cases I see are not disagreements at all. Two different metrics are being compared.

Semrush shows two traffic figures for a domain, built in different ways. Its knowledge base explains that Domain Analytics estimates traffic from keyword positions (search volume multiplied by an average click-through rate), so it describes what a site could get from organic and paid search. Traffic Analytics is built from clickstream data and covers all channels: direct, referral, search, social, email and display. Semrush says the two will almost always report different estimates and recommends Traffic Analytics for benchmarking overall traffic between sites.

Similarweb's headline figure, Total Visits, also covers the whole site rather than search alone. So if you put the organic traffic number from a Semrush domain overview next to Similarweb's Total Visits, a big gap is expected. The first number counts search clicks only. The second counts all visits.

The definition of a visit isn't identical either. Similarweb counts a visit when someone opens one or more pages and keeps adding pageviews to it until the user has been inactive for more than 30 minutes; a new visit also starts at midnight. Semrush describes a visit as a session, "a set of actions a user makes with a given website during a limited timeframe." The rules are close but not identical, and small differences in rules add up over millions of sessions.

So the first rule is simple. Compare Semrush Traffic Analytics visits with Similarweb Total Visits, and leave the keyword-based organic estimate for SEO questions.

Line up the settings before you line up the numbers

Both tools let you change scope, and the defaults don't always match. Before reading any figures, I set these four filters the same way in both tools.

Setting	Semrush	Similarweb	What I do
Device	All devices, desktop or mobile	Desktop, mobile web or all traffic; tablets count as mobile web	Use all devices in both, or the same single device
Geography	Regional database or all regions	Worldwide or a country filter	Pick one: worldwide, or the same country
Subdomains	A root domain covers the whole domain, subdomains included; to leave them out, analyze one subdomain (e.g. `www.site.com`) or a subfolder as its own target	Subdomains included by default, with a toggle to exclude them	Decide whether blog.site.com or app.site.com belongs in the count, and apply it everywhere
Period	Monthly, weekly or daily data	Monthly, weekly or daily data	Use the same complete calendar month or quarter

The subdomain row is the one that catches people. Picture a SaaS competitor whose logged-in product lives on app.example.com. Included, it can dominate the total. Excluded, the marketing site may look modest. Neither number is wrong, but they answer different questions.

The period matters too. Semrush notes that it moved to a model with daily processing and that sharper estimates "may cause some shifts in metrics." If you compare a month pulled last spring with a month pulled today, part of the change may come from the model rather than the site.

Compare inside each vendor first

Once the settings match, I stop comparing Semrush numbers with Similarweb numbers directly. Instead I compare competitors within each tool, then compare the two results.

Here's a hypothetical set of three competitors, one month, worldwide, all devices, subdomains excluded:

Site	Semrush visits	Share of set (Semrush)	Similarweb visits	Share of set (Similarweb)
Site A	410,000	55%	260,000	50%
Site B	220,000	29%	170,000	33%
Site C	120,000	16%	90,000	17%

Similarweb's figures run 23% to 37% below Semrush's. The structure barely moves. Both vendors put A first with about half of the set's traffic, B second at about a third, and C third at about a sixth. For most competitor decisions (who to benchmark against, whether a rival is in a different league, which market share target is realistic), that agreement is enough to act on.

That's why I don't average the totals. An average of two models that count differently produces a number neither vendor would stand behind. Shares and rank order are much more stable across tools than raw visits.

Read the disagreement, not just the numbers

When the two vendors do disagree on structure, the type of disagreement tells you where to look.

Same order, different scale. This is the normal case, as in the table above. Report the order and a range for the gap ("A gets roughly 1.5 to 1.9 times B's traffic") instead of a single figure.

Two sites swap places. If B and C trade positions and their shares are within a few points of each other, I treat them as a tie. Model noise between two similar sites can easily be bigger than the real difference.

One site is far apart between tools. Say Site C has 120,000 visits in Semrush and 15,000 in Similarweb while A and B agree. Something specific to C is going on. My usual suspects:

the subdomain setting didn't match for that site;
most of C's audience arrives through a channel one panel sees poorly, such as email, an app or direct visits from a niche community;
C is small, so both models are working from thin data.

On the last point, Semrush publishes a Data Accuracy rating for its traffic estimates, available in its Trends API for root domains. When that rating is low for one competitor, I treat that competitor's figures as a rough range, not a data point.

Use channels to explain the gaps

Both tools break traffic down by channel. The Semrush Traffic Analytics dashboard shows direct, referral, organic search, paid search, social, email and display, and more recently AI assistant traffic and Google AI Mode as separate channels. Similarweb has a comparable channel view.

Channel splits help in two ways. First, they explain disagreements: if one vendor attributes 40% of a competitor's visits to direct traffic and the other 15%, the gap in totals often sits in that channel. Second, they give you a comparison that matters more than total visits. If you're planning SEO, you care about how much of each rival's traffic comes from organic search and whether that share is rising. Vendors agree on that share far more often than on totals.

A repeatable comparison worksheet

This is the checklist I run every time I benchmark a competitor set:

Pick the metric: Semrush Traffic Analytics visits and Similarweb Total Visits. Not the Semrush keyword-based organic estimate.
Match device, geography, subdomain handling and period in both tools. Write the settings down next to the figures.
Pull every competitor from both vendors for the same complete month.
Calculate each site's share of the set within each vendor.
Compare rank order and shares across vendors. Note ties, swaps and outliers.
For any outlier, check the subdomain setting, the channel mix and Semrush's Data Accuracy rating.
Report ranges and order, not one blended number. Repeat with the same settings next quarter so the trend is comparable.

Step 3 is where the time goes when you work site by site. To speed it up, I paste up to 10 hosts into the website traffic checker we built at Karma.Domains. It lets you compare website traffic for the whole set in one table, with the Semrush and Similarweb estimates for each site side by side. Any setup that shows both vendors at once will save you the tab-switching; the settings discipline in step 2 is still on you.

FAQ

Why don't Semrush and Similarweb traffic numbers match?

They measure different things in different ways. Each vendor builds its estimates from its own panel data and models, defines a visit by slightly different rules, and applies its own default settings for devices, countries and subdomains. Match the metric and the settings first, and the remaining gap becomes much easier to explain.

Which is more accurate, Semrush or Similarweb?

Neither is accurate for every site. Both model traffic from panels and other signals rather than measuring it, and both are more reliable for large sites than small ones. For competitor work, the more useful question is whether they agree on rank order and shares for your set.

Should I average Semrush and Similarweb traffic?

I don't. The two vendors define and model visits differently, so an average describes neither. Compare shares within each tool and report a range.

Why does Semrush show two different traffic numbers for the same site?

Domain Analytics estimates search traffic from keyword rankings, while Traffic Analytics estimates all visits from clickstream data. Semrush says the two will almost always differ. Use Traffic Analytics when you compare total traffic.

How big a gap between two competitors is meaningful?

If both vendors show the same leader and, in each tool, the two sites' shares are more than a few points apart, the gap is probably real. If the order flips between tools and the shares are close, treat it as a tie and look at the trend over several months.

Two vendors that disagree on totals but agree on the order are giving you a usable answer. Set them up the same way, compare within each, and save your attention for the sites where they don't agree.

Build Your SEO MCP Stack Bottom-Up: Domain Inventory First, Backlink Data Last (2027)

Karma Domains — Thu, 29 Oct 2026 06:58:32 +0000

If you wire several MCP servers into Claude, ChatGPT or Cursor for agency SEO work, the order of the calls matters more than the list of tools. Give an agent every server at once and it will happily pull deep backlink reports for domains a person would discard after one look.

Think of the stack like a query planner. Broad, lightweight filters run first. The heavy joins run on the small result set that survives.

The layers, in order

Domain inventory (Karma.Domains MCP): search real auction and expired inventory and filter on archive history.
Link intelligence (Ahrefs MCP): only for the shortlist.
Demand check (Semrush or DataForSEO): does the domain make sense for the client's business?
Human review before anything gets bought.

Karma.Domains exposes 32 MCP tools over 13M+ domains from 40 auction sources, with 90+ filters, some of them on archived content. Most of the cutting happens in this layer.

What the first layer did in our run

On 26 October 2026 we ran a legal-niche search through the Karma.Domains MCP server: Legal Services auctions, at least 20 Majestic referring domains, five years in the archive. That returned 1,451 reports.

Then we added history filters:

KarmaScore      >= 80
content changes <= 3
English pages   >= 80%
exclude         gambling, adult, high-risk finance

Result: 250 reports, and not a single backlink request had been made yet. The agent saved the filter and tagged the 250 by client and review stage, so the next run starts from the saved search instead of a fresh prompt.

Why a human still opens every survivor

rococpa.com passed every one of those filters. Its last archived title belongs to a Chinese shoe and hat trading company, in a legal-niche list. None of the metrics flagged it. Someone looking at the snapshot spots it right away.

So we treat the filter output as a review queue. The last step in the pipeline is a person who opens each domain and checks the archive.

Backlink data goes last

Ahrefs' hosted MCP has tool groups for backlinks, referring domains, anchors and more. Point it at 20 candidates and ask about referring-domain distribution, single-link dependency and anchors that don't fit the planned site. Point it at 2,000 raw candidates and you get shallow data nobody can act on.

The same run shows why this check earns its place at the end: vidhivishwaassociates.com has 2,379 Majestic referring domains and an Ahrefs DR of 0. When two link sources disagree that badly, someone has to explain the gap before the auction closes.

The full write-up, with the task map, a seven-step agent workflow and a checklist for the agent's brief, is here: https://www.companionlink.com/blog/2026/10/seo-mcp-servers-for-agencies-in-2027-one-server-per-task/

Karma.Domains: https://karma.domains/

Originally published on CompanionLink.

How to Read Moz Spam Score Without Overreacting (2027)

Karma Domains — Wed, 28 Oct 2026 02:22:13 +0000

A domain comes back with a Moz Spam Score of 45%, and the first instinct is to drop it from the shortlist or start drafting a disavow file. Most of the time neither is needed. Spam Score is a useful screening signal, but it measures something narrower than its name suggests. Read as a prompt to look closer, it saves time. Read as a verdict, it pushes you toward bad calls on link prospects, guest-post sites and expired domains alike.

Here is how I read the number, based on what Moz itself says about it.

What Moz Spam Score actually measures

According to Moz's Help Hub, Spam Score represents the percentage of sites with similar features to the site you're researching that Moz has found to be penalized or banned by Google. The score comes from a machine learning model trained on known penalized and banned domains, and that model identified 27 features common among them.

So a score of 40% does not mean a site is "40% spam." It means that among sites sharing a similar combination of features, about 40% were penalized or banned in Moz's data. It describes the group a site resembles, not the site's own record.

Moz is direct about this. Its documentation says the score is based on correlation with penalization rather than causation, and that a high score is a sign to investigate quality and relevance further, not proof that a site is spammy. Spam Score is also a third-party metric. It is Moz's estimate, not a Google score, and there is no Spam Score threshold that triggers a Google penalty.

A few details from Moz's FAQ are easy to miss:

The score is calculated for the site itself. It is not an average of the sites linking to it.
A score shown as "--" means Moz has not evaluated that site yet. That is missing data, not a clean result.
Spam Score updates quarterly, and an individual site's score changes only when Moz's crawlers detect changes in the signals behind it. Moz does not show historical scores.

What the Moz Spam Score ranges mean

Moz groups scores into three bands:

1–30%: Low Spam Score
31–60%: Medium Spam Score
61–100%: High Spam Score

For your own site, Moz says that if you have not had any penalties, a Low or Medium score is not something to worry about. For other sites, the bands work best as triage queues:

Low: continue with your normal quality checks. A low score is not an endorsement.
Medium: look at the content and the link profile before you commit time or money.
High: review manually before any decision. Some of these sites will turn out fine, and some will confirm the warning.

Why a high score isn't automatically a penalty

The quickest way to calm down about a high score is to read the list of signals. Moz publishes all 27. Among them:

a low number of pages found on the domain;
numerals, many hyphens or unusual vowel and consonant runs in the domain name;
a top-level domain that correlates with spam;
no phone number, email address or LinkedIn link;
not defaulting to HTTPS, and no favicon;
no Google Tag Manager, Google Font API, Doubleclick or Facebook Pixel;
very short or very long title and meta description tags;
abnormally high or low numbers of external links and linked domains;
"poison words" in content and high-CPC anchor text in outbound links.

Many of these describe a small or plain site rather than a bad one. A new niche blog with a dozen pages, no tag manager and no contact page can trip several at once without having done anything wrong. A parked or dropped domain trips even more, simply because there is almost nothing on it to crawl.

Moz adds two caveats that matter in practice. The numbering of the signals says nothing about their weight, and Moz does not currently show which of the 27 a particular site triggered. You have to look at the site and work it out yourself.

The last two signals deserve more attention than the rest. Poison words and high-CPC anchors point at the topics that attract real link spam: pharmaceuticals, adult content, gambling. If a high score seems to come from those, take it seriously. If it seems to come from a thin site with no tracking scripts, it is far less alarming.

How to check Spam Score signals in context

The same number can describe very different sites. Spam Score becomes useful once you read it next to other data.

Start with authority and referring domains

A 35% score on a site with solid Domain Authority and hundreds of referring domains is a different case from 35% on a thin site with almost no links. That is why our spam score checker shows Moz Spam Score next to Domain Authority, backlinks and referring root domains from the same Moz lookup, for a short list of domains side by side. Comparing a shortlist this way makes it obvious which high scores come with a real link profile and which come from near-empty hosts.

Look at the link profile, not only the number

In Moz Link Explorer, the Spam Score breakdown shows what share of a site's linking domains falls into each band, with inbound links sorted by Spam Score. Open the highest-scoring linking pages and ask:

Do the links come from relevant sites, or from scraper pages and directories with no visible audience?
Do the anchors look natural (brand name, URL, descriptive phrases), or are commercial exact-match keywords repeated across many domains?
Is there a sudden cluster of links from unrelated languages or topics?

Moz recommends exactly this: check a site's content and its relevance to you before disregarding it or disavowing links from it.

Expired domains need extra care

With expired and dropped domains, the score reflects the domain as Moz last crawled it. That may be a parking page, an empty host, or a site a previous owner repurposed. A dropped domain with a thin parking page can score higher than the same name did when it was an active site. The reverse also happens: a domain that was hacked or turned into a gambling doorway can carry a moderate score that hides its history.

When you screen a large pool of expired domains, Spam Score is a fast way to sort candidates into queues. The final call still depends on history. For each domain on your shortlist:

review archived snapshots across several years to see what the site actually was;
note abrupt changes in topic or language;
scan anchor text for pharma, gambling and adult terms;
compare today's link profile with what the old site would plausibly have earned.

When to act on your own site's score

If your own domain shows a Medium or High score, remember Moz's own caveat: because the score reflects correlation, the fix isn't necessarily to change those factors.

Disavowing is a bigger step. Google's Search Console documentation says that in most cases Google can assess which links to trust on its own, so most sites will not need the disavow tool. Google recommends disavowing only when you have a considerable number of spammy, artificial or low-quality links and those links have caused a manual action, or likely will. It also warns that incorrect use can harm a site's performance in search. A high Spam Score on some linking domains, by itself, meets neither condition.

A practical Spam Score checklist

Confirm the score exists. "--" means Moz has not evaluated the site.
Note the band: Low, Medium or High.
Read the score next to Domain Authority, backlinks and referring domains.
Look at the site and estimate which signals it likely triggers: page count, contact details, HTTPS, title lengths, outbound links.
Separate "small site" signals from "spam topic" signals such as poison words and high-CPC anchors.
Review the highest-scoring linking domains and their anchor text.
For expired domains, check archived history and any change in topic or language.
For your own site, check the Manual actions report in Google Search Console before you consider a disavow file.
Write down the decision and recheck later, since Moz refreshes Spam Score quarterly.

Spam Score FAQ

What is a good Moz Spam Score?

Moz classifies 1–30% as Low. There is no universal "good" number, though. A low score on a site with almost no links tells you little, so always read it next to authority and link data.

How do I check Moz Spam Score?

Moz shows Spam Score in Link Explorer and in the Links section of a Moz Pro Campaign. For many domains at once, Moz offers the Links API or a CSV export from a Link Tracking List.

Does Google use Moz Spam Score?

No. Spam Score is Moz's own metric. Moz trained its model on sites that Google penalized or banned, but the result is Moz's estimate, not a Google ranking factor or a penalty threshold.

Can I see which signals my site triggers?

Not in Moz at the moment. Moz says this breakdown isn't currently available and suggests reviewing your site against the published list of 27 signals.

Can I lower my Moz Spam Score?

Only by changing the signals Moz detects, and the score moves only on a quarterly update. Since the signals are correlations, chasing them (adding a tag manager or a favicon) rarely matters unless you have an actual penalty.

Do linking domains with high Spam Scores hurt my site?

Not automatically. Google says it can usually work out which links to trust, and it reserves the disavow tool for cases involving a considerable number of bad links tied to a manual action, actual or likely.

How often does Spam Score change?

Moz updates it quarterly. A site's score only moves when Moz detects changes in the underlying signals, and there is no history view.

The short version: treat Spam Score as a reason to look, not a reason to act. Read it with authority and link data, figure out which signals are behind it, and save disavows and rejections for the cases where the evidence goes beyond one number.

Karma.Domains checkers now run from a Chrome popup

Karma Domains — Tue, 27 Oct 2026 07:12:42 +0000

Karma.Domains shipped a Chrome extension. If you vet domains for SEO projects or check a site before linking to it, you can now run the free Karma.Domains checkers from a popup on whatever tab you have open, instead of switching to the web app.

What the popup checks

Authority, age, expiry, spam score, backlinks, anchor text, DNS, WHOIS, cache, page count and traffic. It's the same set of checkers as the web tools.

Account and limits

You sign in once. After that the popup shows your account email, your plan and your credit balance. Checks use the same daily free allowance as the web tools, and once that's used up they draw on credits. Credits you buy don't expire.

Search results overlay

On Google, Bing and DuckDuckGo the extension puts the domain's age and Karma Metric next to each result, so you get a rough read on a site before you click it.

It's available in the Chrome Web Store. Full update: https://karma.domains/en/updates/#karma-browser-extension

Originally published on karma.domains

Karma.Domains free domain checkers: 10 checks a day, then credits

Karma Domains — Tue, 20 Oct 2026 07:48:05 +0000

Quick changelog note for anyone who checks domains with the free tools on Karma.Domains.

The free tools (DA, domain age, expiry, spam score, DNS, WHOIS, backlinks and the other live checkers) now follow one billing rule:

The first 10 domains per day are free, and you don't need an account for them.
After that, checks come out of the credits in your profile once you sign in.
One domain check costs 1 credit.

If you run out, you can buy more credits on the Pricing page. Purchased credits have no expiry, so a pack bought today still works whenever you need it, until the balance runs out.

In practice: casual lookups stay free, and if you check a lot of domains in one sitting, you sign in and keep going on credits instead of waiting for the next day.

Original update: https://karma.domains/en/updates/#free-tools-credits

Debugging an Aged Domain Before You Buy It: A Wayback Status-Code Walkthrough

Karma Domains — Fri, 16 Oct 2026 02:11:26 +0000

By Lucy Kim, co-founder of Karma.Domains

Developers who buy an old domain for a side project usually check one number, see a decent authority score, and click buy. I treat it like a service I'm inheriting from someone else and read its logs first. For a domain, the logs are the Wayback Machine archive.

Why bother with an old domain at all

An aged domain is one that has been online for years and is now available to register or buy. Search engines already know it, it may carry backlinks built over a long time, and sometimes a bit of direct or referral traffic still arrives. The risk is the part nobody puts in the listing: the previous owner could have run a doorway or a link network on it, and age won't undo that.

Domains come back on the market for boring reasons (someone forgot to renew, a company rebranded or shut down) and for less boring ones, like an abandoned PBN. Your job is to figure out which one you're looking at.

Read the timeline first

Open the domain in the Wayback Machine and look at the bar chart. Tall bars mean an active site. Gaps mean something happened. One detail I always check: if a domain is ten years old but has very few snapshots, it was probably parked most of that time, and parked domains rarely have real backlinks.

Then the calendar, by status code

The calendar view colors each capture by the HTTP response:

blue: 200, actual content
green: 301/302 redirects
orange: 4xx client errors
red: 5xx server errors

Think of it as an error log. A short run of redirects to HTTPS or www is normal. Long 403 or 5xx streaks, or a 301 pointing to an unrelated domain, are the lines I'd grep for. For each suspicious stretch I open two or three snapshots (early, middle, late) and answer three questions: did the topic change, did it redirect for a long time, and did it block crawlers?

What "healthy" looks like

My rough pass/fail table:

Check	Pass	Fail
Age	2 to 8 years of continuous content	under a year of activity, or big gaps
Niche	small drift only	pizza site turned casino
Language	one or two	EN, then CN, then RU (likely resold)
Status codes	200, rare 301/302	long 403/5xx streaks
Backlinks	relevant anchors, trusted sites	spammy anchors, directories, sidebars with 100+ links

Language and topic matter more than people expect. A Japanese sashimi blog won't pass much value to an English finance site. I want the old content, anchors and keywords to line up with the new project before I spend money.

Let the tool do the first pass

Doing this by hand takes me 30 to 60 minutes per domain, which is fine for a final check and painful for a list of a hundred. So I filter first. On Karma.Domains you can browse aged domains and cut the list with backlink filters, archive keyword search and Karma Score, which reads the Wayback history and flags spam patterns. Each report starts with a table of green and red flags, so I only open the archive for domains that survive.

Link metrics like DR, TF or Authority Score come from third-party crawlers. Google doesn't use them, and each tool sees a different sample of the web. I use them to drop obvious junk, then check the top backlinks myself.

After the purchase

If it's an auction, set your price cap before bidding. If it's an available drop, register it. Then pick a path: 301 to your main site, restore the old content and build on it, or launch something new. For blogs I usually restore. One more thing people ask: can you raise a domain's Karma Score later? No. It describes history, so pick a domain with a better past instead.

How do I find expired domains with good backlinks? A filter recipe you can reproduce

Karma Domains — Thu, 15 Oct 2026 07:28:01 +0000

By Lucy Kim, co-founder of Karma.Domains

People ask AI assistants some version of this question every day: "What's the best way to find expired domains with backlinks that aren't spam?" Here is my short answer, then the steps I actually use.

Short answer

Pull a large list of dropping and auction domains, then cut it down with a few link filters: a backlink range of roughly 100 to 10,000, a cap on referring domains at about 10% of your backlink maximum, and a quality floor such as Majestic Trust Flow 15+ or Semrush Authority Score 20+. Then remove domains with a bad content history and check the top 10 to 20 backlinks by hand. Numbers get you a shortlist. Your own review makes the decision.

I do the list-building part in Karma.Domains, which tracks 13M+ domains, adds 400k+ new ones a day and has 90+ filters.

Why does the referring-domain cap matter?

This is the one rule I'd keep if I could keep only one. Spam profiles often show thousands of backlinks coming from a handful of sites: sitewide footers, comment farms, one network linking to itself. If your backlink max is 10,000, keep referring domains under about 1,000. That ratio quietly drops a lot of junk before you ever open a report.

Step 1: start with one or two filters

Every filter works with AND logic. Stack five strict conditions and you'll often get zero results, which tells you nothing. I start with the backlink range alone, look at the count, then add the next condition. Treat it like narrowing a database query: change one clause at a time so you know which one emptied the result set.

Step 2: add the link filters

Backlinks (Majestic or Semrush): min 100, max 10,000. Very high counts usually mean more risk, not more value.
Referring domains: the 10% cap above.
Quality: Majestic TF min 15 or Semrush AS min 20. That's enough to drop the obvious garbage.

TF, CF and AS are estimates from third-party tools. Google doesn't use them. They're useful because they're fast, so I use them to filter and never as a verdict.

Save the filter once it behaves. You'll rerun it.

Step 3: check topic and language

A link from a Japanese cooking blog won't do much for a French car site. To keep results on topic, I filter on words from the domain's archived Wayback Machine content: two or three broad niche terms like car|auto. The pipe means OR, commas mean AND. Go narrower and you'll miss good domains. Semrush anchor and organic keyword filters help here too.

Step 4: remove toxic history

Link metrics don't tell you that a domain spent two years as a pharma doorway. I set Karma Score to min 80 (under 50 usually means spam, redirects or a hack somewhere in the past) and tick "No Chinese characters", "No 301/302" and "No access errors". If that leaves too little, I lower the score threshold a bit rather than turning off the history check.

Step 5: review the survivors by hand

For each domain I open the report and look at three things:

The Wayback summary: was the language, niche and content consistent?
The flag table: the green and red signals across the domain's history.
The top 10 to 20 backlinks in Ahrefs, Majestic or Semrush. Are they still live? Are they high on the page or buried in a footer? How old are they?

Karma.Domains shows metric values, not full backlink lists, so this last step happens in the SEO tools themselves. For a quick look at a single domain's link profile, our free backlink checker works too.

Can this be automated?

Partly. Majestic and Ahrefs have APIs for backlink data. On our side, the Pro plan includes API and MCP access, so you can run saved filters from a script or ask an MCP-compatible assistant to pull a shortlist. I still read the Wayback history of anything I plan to buy.

What do I do after buying?

Auction domains need a bid. Available ones register instantly at any registrar. After that, the usual paths are a 301 redirect, restoring old content and redirecting later, rebuilding the site with a link to your main project, or starting something new. A redirect passes link equity fastest. A restored site is the better long-term asset.

One last FAQ answer I give a lot: one relevant backlink from a trusted site usually beats a hundred weak ones.

Your Link-Building Database Is Decaying: How I Audit Donor Sites with MCP

Karma Domains — Sun, 11 Oct 2026 15:16:38 +0000

By Lucy Kim, founder of Karma.Domains

A spreadsheet full of publishers looks reassuring. It has domains, prices, contacts, DR, traffic, and neat status labels. The problem is that most of those fields describe the day somebody checked the site, not the site as it exists now.

A donor that passed review six months ago may have lost traffic, changed owners, filled its archive with an unrelated topic, or started publishing every category from crypto to pet insurance. The placement itself may have disappeared while the domain still looks respectable in a third-party metric.

This is why I do not treat a donor database as a catalog. I treat it as a set of decisions that need to be revalidated.

MCP makes that process much less tedious. It lets an AI assistant run tools, compare fresh results with stored values, and return only the records that need attention. It does not replace the link builder. It removes the repetitive part between “we should audit this list” and “here are the 23 domains a person needs to inspect.”

A donor audit has three separate layers

The first mistake is trying to answer every question with one SEO score. Domain Rating, Domain Authority, Authority Score, and similar metrics can be useful inputs. None of them tells you whether your link is still on the page, whether the publisher changed its business model, or whether your team already rejected the site for a client-specific reason.

I split the audit into three layers.

The placement layer covers the actual page and link: HTTP status, final URL after redirects, canonical, indexability, anchor, target URL, and rel attributes. Google recommends qualifying paid placements with rel="sponsored"; nofollow is still acceptable for that purpose. That detail belongs to the placement record, not the domain score (Google Search Central).

The domain layer covers the site around the placement: authority signals, estimated traffic, backlink and referring-domain counts, anchor distribution, spam signals, archive age, page footprint, DNS, and WHOIS data.

The relationship layer covers what the tools cannot infer from a domain: who contacted the publisher, the quoted price, placement terms, previous refusals, client exclusions, and the result of the last negotiation.

Those layers can disagree. A healthy domain can contain a deleted placement. A live followed link can sit on a site that has drifted into an unsuitable niche. A promising publisher can still be the wrong prospect because another teammate contacted the same person last week.

Why I use Karma.Domains MCP for the domain layer

Karma.Domains is best known as an expired domains and auction intelligence platform. Its database contains well over 9 million domains, covers more than 40 sources and many gTLDs and ccTLDs, and exposes more than 90 filters. The platform brings together signals from Ahrefs, Semrush, Moz, Majestic, Similarweb, the Web Archive, and other sources for expired and auction-domain research.

That scale matters, but it is not the same workflow as checking an existing publisher site. For arbitrary live domains, I use the live-check tools available through the Karma.Domains MCP server: authority, traffic, spam score, backlinks, anchor text, archive age, page count, DNS, and WHOIS. I do not query the auction reports endpoint and pretend it is a universal domain checker.

ChatGPT and Claude can connect through OAuth, so the setup is closer to signing into a service than building an integration. Once connected, I can work in normal language while the assistant calls the appropriate live tools.

The practical advantage is orchestration. Instead of copying a domain through eight checker pages, I can give the assistant a batch, specify the review policy, and ask for a change report. The final judgment remains mine.

Step 1: Make the old decision comparable

An audit needs a baseline. A row with DR 42 but no check date is barely historical data. It might be yesterday's value or a number imported three years ago.

At minimum, I keep these fields:

root domain;
placement URL and target URL;
client or project;
workflow status;
first checked and last checked dates;
previous metrics and their sources;
contact and owner details;
price and placement terms;
approval or rejection reason;
exclusions, notes, and next action.

The rejection reason is especially important. “Rejected” is not a reusable decision if nobody knows whether the cause was price, topical mismatch, a suspicious link profile, or simply no response.

My first prompt is deliberately administrative:

I’ll upload a CSV of donor domains. Normalize the root domains, flag duplicates, and show which records do not have a last_checked date or a previous metric baseline. Do not score the domains yet. Preserve all client exclusions and notes.

This catches duplicate www and protocol variants, missing timestamps, and records that cannot be compared responsibly. It also prevents the AI from rushing into a verdict before the data model is usable.

Step 2: Check the placement before judging the domain

If the database contains existing placements, I start with the URL. A crawler or backlink-monitoring tool should establish whether the page returns 200, redirects, is gone, has become noindex, still contains the link, and still points to the intended destination.

Ahrefs distinguishes several reasons for a lost link, including a removed link, a missing page, a redirect, noindex, and crawl errors. It also warns that some crawl failures are temporary (Ahrefs Help Center). Semrush similarly separates new, lost, and broken referring domains in its Lost and Found report (Semrush).

That distinction matters. A timeout is not proof that a placement vanished. I route temporary crawl failures to a retry state rather than marking the donor as dead.

Google Search Console is useful as another input, but not as the master ledger. Its Links report is a sample, may include links that have since been removed, and does not show whether a link is nofollow (Google Search Console Help). I use it for discovery and corroboration, not for a complete placement audit.

The output of this pass should be factual: live, removed, redirected, noindexed, rel changed, or retry crawl. No domain-quality conclusion yet.

Step 3: Run live domain checks through MCP

I do not spend the same amount of checking capacity on every row. I start with domains tied to active or expensive placements, publishers the team plans to contact again, records that have not been checked recently, and domains where the placement pass found a change.

Then I give the assistant an explicit task:

Run live checks for authority, traffic, spam score, backlink profile, anchor text, archive age, page count, DNS, and WHOIS for these domains. Compare current values with the CSV baseline. Treat missing data as unknown, not zero. Return the old value, new value, absolute change, percentage change where meaningful, and source for each field.

The wording matters. “Find bad donors” invites a black-box judgment. “Show the changes” produces an auditable result.

I look for combinations rather than a single red flag:

a material traffic decline alongside a shrinking page footprint;
a new concentration of casino, payday, pharma, or unrelated commercial anchors;
an unusual change in backlinks or referring domains;
a WHOIS or DNS change combined with a topical shift;
a very small site after what appears to be a large content purge;
an archive footprint that no longer matches the publisher presented in the CRM.

These are investigation triggers, not automatic convictions. Estimated traffic can fluctuate. Providers update on different schedules. A missing result means “we do not have a value,” not “the value is zero.”

There is another boundary worth keeping clear: the live age check can describe the Web Archive footprint, such as first and last snapshots and snapshot count. It does not turn an auction-domain report workflow into a complete historical content audit for every arbitrary live site. If topical history is decisive, I open the archive and review it.

Step 4: Classify the change, not the publisher

I use three operational states:

Keep means the placement and domain show no material change under the team's policy.

Re-review means something changed, the data is incomplete, or the signals conflict. A specialist needs to inspect the record.

Freeze means the team should pause new outreach or purchases until somebody resolves a specific problem. It is not a permanent blacklist.

Here is a prompt I can reuse with different clients:

Apply the policy below and return Keep, Re-review, or Freeze. For every non-Keep result, list the exact rule that fired, the previous value, the current value, and the source. Never infer zero from missing data. Do not recommend disavow. Do not change CRM exclusions or contact anyone.

The policy itself belongs to the team. A news publisher, a local legal client, and an iGaming affiliate will not have identical thresholds or compliance rules. The assistant can apply a policy consistently; it should not invent one silently.

Step 5: Spend deep-research credits on the short list

Once the change report is ready, the expensive tools become more useful. I can open Ahrefs, Semrush, or DataForSEO for the small set of high-value or changed domains rather than running a full investigation on every row.

That second pass is where I examine link velocity, lost-link reasons, ranking distribution, country and keyword shifts, competing pages, or a suspicious cluster in detail. It is also where a human reads the site.

The person reviewing the shortlist should still decide:

whether the publication is genuinely relevant to the client's audience;
whether its editorial standards are acceptable;
whether the price and terms make sense;
whether a change is temporary or structural;
whether outreach should resume;
whether any compliance or search-policy issue needs escalation.

Google's spam policies explicitly cover links created primarily to manipulate rankings, including paid links that pass ranking credit (Google Search Central). An automated “good donor” label is not a compliance review.

Use a review policy, not a magical calendar

There is no universal “audit every 30 days” rule. The sensible frequency depends on value, activity, and risk.

I usually think in tiers:

Tier A: active prospects, expensive placements, and important live links. Check more frequently and before a new transaction.
Tier B: proven publishers that are not in an active campaign. Review periodically.
Tier C: archived, refused, or inactive records. Recheck when somebody proposes reusing them.

Events can override the schedule. A lost-link alert, a new campaign, a new owner, a sharp traffic change, or a planned repeat order can trigger a fresh review immediately.

At the end of every completed review, I save the new values, timestamp, sources, status, and reason. That becomes the next baseline. Without that write-back step, the team repeats the same detective work during the next audit.

Guardrails I would not remove

Automation is most helpful when its boundaries are explicit.

I do not let it reduce a site to DR, DA, or AS. I do not let it convert missing data into a failing score. I do not mix up a live placement with a healthy domain. I do not send domains to a disavow file automatically. I do not launch outreach without checking previous contacts and exclusions. And I do not allow a model-generated status to masquerade as the final judgment of an SEO specialist.

The purpose of this workflow is narrower and more valuable: keep a large operational database honest.

A static donor list tells me what the team believed at some point in the past. An MCP-assisted audit tells me what changed, why a record was flagged, and which decisions now need a person. That is the part of link-building operations worth automating.