Method
Reading a rival's backlinks for network footprints
A private blog network leaves marks in a link graph. None is conclusive alone. Together they make a case you can defend, and this page ranks them by weight.
Nobody labels a network. You infer it. The graph only shows you who links where, so every claim here is a hypothesis until you open the pages and read them.
We rank footprints. The order reflects how hard each is to fake and how often they mislead. A shared hosting address would be strong, but link data cannot show one, so we lean on what it can show: overlap, wording, ratings, targets and timing. To find backlinks to your website or a rival's, start with refdomains and keep the file.
Five footprints, weighed
| Footprint | What you see in the data | Weight | Innocent explanation |
|---|---|---|---|
| Exclusive overlap | A group of domains linking to the same 3 to 4 rivals and almost nobody else | Strong | A niche directory or a widely syndicated press release |
| Anchor reuse | Identical commercial phrases across many unrelated domains | Strong | A template in a sponsor widget |
| Rating band | Referrers all sitting between, say, DR 12 and 18 | Medium | Any young niche has a band |
| Single-target focus | Each referrer sends links to one page only | Medium | A resource page about one product |
| Burst timing | Many new referrers in one refresh window | Weak | A press hit or a viral post |
The examination in four passes
Pass 1: overlap
Run intersect on three or four direct rivals. Sort the result by the number of rivals each domain links to. The top of that list is where networks sit, together with the honest authority sites.
Pass 2: remove the famous
Strike out domains that obviously earn links naturally: newspapers, universities, government. What remains is the tail. A hundred tail domains linking to every rival is odd.
Pass 3: read the words
Run anchors on each rival and compare the commercial phrases. If three competitors share an odd exact-match phrase, ask keyword who else is linked with it.
Pass 4: open ten pages
Visit ten referrers at random. Real sites have a human logic: an author, an about page, posts that predate the link. Networks have thin posts and links to casinos, loans and a rival.
Say rival-a.com, rival-b.com and rival-c.com each have about 400 referring domains. An intersect returns 62 domains that link to all three. Of those, 40 are newspapers and universities. The other 22 sit between DR 11 and 17 and share three anchor phrases. That group of 22 is your lead. The numbers are invented for illustration.
What to do with a suspected network
Do not copy it. Google's March 2024 spam policies named scaled content and expired domain abuse, and the December 2022 link spam update moved link evaluation toward ignoring manipulation. A rival's network may be sinking them rather than lifting them.
Report it only if you have more than a graph. Otherwise, wait. Two tools can disagree, and our index refreshes roughly monthly. Look at the troubleshooting table before accusing anyone, and keep your own house in order with the defence routine.
If you want to know what a clean competitor profile looks like for contrast, pull the same three calls on a rival you trust. The difference teaches more than any list of rules.
Evidence folder for one suspected network
- Intersect output, saved with the call date
- The anchor lists of each rival, side by side
- Ten page visits, with a one-line note for each
- A list of innocent explanations you ruled out
- A decision: ignore, learn from it, or report