Ten-minute link audit for a client you have never seen
Use this reference of five fields to check website backlink data for a client you have not met. Read them in order, write one line per field, and you have a defensible note before the call ends.
LinkSkimmr research team ·
What the note should say
Three lines beat a twelve-page export
Line one: the size of the profile against two rivals. Line two: the one thing that looked odd. Line three: the next check you would run, or the statement that you would leave it alone.
Date the lookup, because the index refreshes about monthly
Name the tool and metric, since our DR is not Moz DA
Never write safe or penalty-proof anywhere
Field reference: what to read and what it flags
Order matters. Fields one and two cost nothing, so read them before spending a credit.
#
Field
Where it comes from
What it flags
1
Domain rating (0-100)
overview, free
Size class of the profile, compared with rivals
2
Referring domains
overview, free
Breadth; the input of the DR formula
3
Biggest linking domains
refdomains, 2 credits
Networks, dead blogs, odd TLD clusters
4
Anchor text
anchors, 2 credits
One money phrase repeated; brand share too low
5
Top linked pages
toppages, 2 credits
Value parked on a 404 or on a redirect chain
Reading fields one and two without fooling yourself
DR is a log scale. Going from 40 to 50 needs roughly five times the referring domains, which is why a rival at DR 52 is not 'a bit ahead' of a client at DR 41.
Our formula is round(100 times log(ref_domains plus 1) divided by log(2 times 10 to the 7th)), so you can back out the referring-domain count from any DR and sanity-check what a salesperson quoted you. The threshold table does the arithmetic for you.
One more habit: compare the client to rivals in the same call family. Mixing our DR with a screenshot of someone else's metric produces a number that means nothing, however convincing the screenshot looks.
Fields three to five as three calls
Each call is billed once, whatever the number of rows, up to 1,000 rows. Ask for the biggest slice first.
bash
BASE="https://linkskimmr.com/api.php"
KEY="sk_live_..."
# field 3: who links most
curl "$BASE?action=refdomains&q=clientsite.com&per=100&key=$KEY"
# field 4: how the anchors read
curl "$BASE?action=anchors&q=clientsite.com&per=100&key=$KEY"
# field 5: where the value sits
curl "$BASE?action=toppages&q=clientsite.com&per=50&key=$KEY"
Three cheap reads that catch most problems
Read the top 100 linking domains aloud, quickly. A cluster of near-identical names, one odd TLD repeated thirty times or a pile of expired-looking blogs deserves a line in the note.
Anchors come next. A natural profile is mostly brand names, bare domains and generic words such as 'here'. If one commercial phrase fills a quarter of the list, say so, but remember that a pattern is a prompt for a deeper look and never a verdict.
Last, open the top three linked pages in a browser. If the best-linked page redirects twice or returns a 404, you found the cheapest win of the day, and the redirect page explains the fix.
What to leave out of the ten minutes
Skip the long export. A spreadsheet of 40,000 links proves effort and answers nothing.
Skip toxicity scores too. Semrush and Ahrefs publish their own, they disagree with each other, and a client will treat any number as a verdict. If a pattern bothers you, write what the pattern is and what you would check next.
Google's March 2024 spam update and the earlier link spam update of December 2022 are the useful context here: Google says it neutralises many spammy links rather than punishing the receiving site, which is a reason to describe patterns carefully and not alarm anyone.
When a link is missing from the list
Our crawler and our index refresh about monthly, so a recent link may simply not be there yet. Absence in our data does not mean Google ignores a link. Re-run next month before you tell a client something is gone.