Referring domains vs. backlinks

Referring domains vs backlinks is a difference in counting units. A backlink describes a link from a source to a destination. A referring-domain count groups source records by domain. One domain can contribute many pages and links, so a larger backlink total does not necessarily mean more distinct publishers.

Choose the count that answers your question before comparing competitors. The competitor backlinks hub explains how these inventories support research rather than a universal ranking score.

Keep four layers visible

A raw link row can preserve a source URL, target URL, anchor, link attributes, and observation time. Several rows may come from the same source page. Several source pages may share a hostname such as news.example.com. Several hostnames may belong to the same registrable domain.

Do not assume that a vendor's “domain” label matches your own grouping rule. Read its report documentation and retain the original field. If you derive another grouping locally, give it a separate column rather than replacing the supplier's value.

The Public Suffix List documentation explains why suffix boundaries cannot be handled reliably by taking the last two labels of every hostname. For analysis, document the rule and list version you use. A grouping decision can change counts without changing any links on the web.

Count a fictional source set

Imagine a furniture repair guide with the following synthetic observations. These rows are teaching examples, not a measured backlink profile.

Source group	Source pages	Observed link instances
Workshop journal	3	4
Community repair directory	40	40
Regional newsletter	2	2

The table contains 45 source pages and 46 link instances across three named groups. The workshop journal links twice on one page, which explains the difference between pages and instances. The directory repeats a link across 40 pages.

If the next question is “How many individual pages should we inspect?”, 45 is the relevant starting count. If it is “How many source relationships might require a conversation?”, the three groups are more useful. Neither total tells you the quality of the passages or who controls the pages.

Now suppose the directory removes its repeated footer from all 40 pages. The link-instance total falls sharply, while the number of independent source groups changes by one. Both observations matter, but they tell different stories about what happened.

Avoid equating domains with publishers

Domain grouping is a useful technical view. Editorial ownership is a separate question. A single organization may control several domains, while a hosting platform may contain pages from many unrelated authors.

Keep an ownership note only when you have evidence. Do not collapse domains into one publisher merely because they share a design or network address. Equally, do not assume every subdomain represents a distinct editorial decision.

When researching a candidate, return to its exact source page. The surrounding text and destination give you a better basis for review than the aggregate count alone. A broad domain total can hide repeated templates, archived pages, or references unrelated to the campaign.

Choose grouping before calculating a gap

A domain-level gap asks whether a source group appears in the compared inventories. A page-level gap asks whether a particular source page appears. These can produce different shortlists from the same raw records.

A competitor might have a citation on one article while your site appears elsewhere on that domain. The domain-level comparison may show no gap, but the article could still be worth reviewing for a distinct reader need. Record which question you are asking so the result is interpretable.

The backlink gap worksheet uses explicit membership states. Our export comparison guide covers mismatched units, and link intersect shows how shared groups become a review queue.

Preserve the detail behind the total

Keep a route from every aggregate back to its source rows. A reviewer should be able to open a group, inspect the relevant pages, and see when the observations were collected. Store unresolved rows separately if their addresses cannot be parsed or grouped reliably.

Report the unit in the heading: source pages, link instances, source hosts, or grouped domains. Add the grouping rule and date range nearby. These small labels prevent a report from claiming dozens of new publisher relationships when it actually found dozens of repeated links on one site.

Sources

1. Learn more about the Public Suffix List · Public Suffix List · Not stated; accessed September 12, 2026