Link profile: authority, trend and competitive gap

The page answers the questions a consultant opens first in front of a client: where the domain's authority comes from, what it gained and lost over the past year, what its anchors give away, and which sources link to its competitors but not to it. Everything is computed on the project's main domain — the one set in the project settings — from the DataForSEO Backlinks API. One rule runs through the whole page: a figure DataForSEO did not measure reads « not available », never 0. Zero is a measurement; mistaking it for a missing one costs you in a client meeting.

The six profile cards

• Domain authority, scored out of 1000, not out of 100. That is the DataForSEO Backlinks API scale, and it is logarithmic: going from 200 to 300 takes far more work than going from 100 to 200. Do not compare this number to a DA or a DR read elsewhere — the scales have nothing in common.
• Backlinks: the total number of inbound links recorded, across all pages.
• Referring domains: how many distinct domains link to you. This figure, not the previous one, tells you how wide a profile is — a thousand links from three sites weigh less than a hundred links from a hundred sites.
• Spam score, out of 100. A profile-wide indicator computed by DataForSEO for the whole domain.
• Broken links: inbound links pointing at a dead page on your site. They are the easiest wins available — the authority is already earned, you only have to restore or redirect the target.
• Referring IPs: how many distinct IP addresses sit behind your referring domains. A wide gap between referring domains and referring IPs points to links hosted in the same place, and therefore often to a single network.

Followed and nofollowed links

• The dofollow / nofollow split is shown at two levels: referring domains and the links themselves. They do not tell the same story — one single domain can carry hundreds of nofollow links and swamp the link-level reading.
• Only dofollow links pass authority. A heavily nofollow profile often explains authority that stalls despite a respectable link count.
• When the measurement is missing, the page writes « not available » and shows no share at all. DataForSEO does not return the nofollow counter for every domain; without it, subtracting would produce a « 100% dofollow » that was never measured. That was the old behaviour and it is fixed: an unmeasured domain no longer reads as a perfect profile.

Links won and lost over twelve months

This is often the most actionable section on the page. Twelve months of movement, month by month: green bars are links won, red bars links lost, and the blue line is the monthly net. A numeric summary sits above it — won, lost, net links and net referring domains. Read it this way: a profile that wins three hundred links and loses three hundred and twenty is not stable, it is receding while everyone assumes it is being maintained. Conversely, a single dip right after a site rebuild almost always points to redirects worth checking rather than to lost favour.

Anchor text and over-optimization

• Anchors are grouped into families: branded, bare URL, keyword, generic (« click here », « read more ») and no anchor. The branded family is recognized from the project name, the brands and the domains declared in settings — a thinly configured project will see its branded anchors counted as keywords.
• Shares are computed on the fifty most frequent anchors the API returns, not on the entire profile. The page says so under the section title, and you should say it the same way to a client: « 22% of the recorded links », never « 22% of your backlinks ».
• A warning appears when a keyword anchor passes 10% of the recorded links (concentration worth watching, amber) or 20% (likely over-optimization, red). An exact-match anchor that dominant is rarely earned naturally.
• The share held by the top three anchors is repeated under the chart. That is the number to check before ordering a batch of exact-match links, not after.
• An anchor with no letter or digit (« ] », « → », for instance) is classified « Empty / symbol », not as a keyword: it no longer triggers a false over-optimization warning. Stored snapshots are reclassified the same way when displayed.
• Generic anchors: « View website », « click here », « learn more » and their common variants are recognised as generic. An anchor counts as your brand only if it contains the brand name as a whole word: an anchor that only contains it inside another word is classified elsewhere.

Competitive gap

• The list of domains that link to your competitors and not to you. Each row names the competitors that source already cites, its authority and how many links it gives them: it is an outreach list, not a ranking.
• The ordering is deliberate: sources citing the most competitors come first. A site linking to four of your five competitors is a source the whole industry depends on — and you are not on it.
• You need competitors with a domain in the project settings. Without them the section says so and points back to the project instead of staying blank: the gap is computed from their domains, there is nothing to guess. Ten competitors at most are queried.
• Your competitors are recognised by their domain, whether it is entered with or without « www », as a full address, or already collected elsewhere in the project. A competitor without a domain is named, with a link to Settings to add one: without a domain, there is no way to know who links to it.
• Your own domains and the competitors themselves are stripped from the list: one competitor linking to another is not a link opportunity.
• When no gap is found, the page announces it as good news rather than emptiness — every domain linking to your competitors links to you as well.

Referring domains, link competitors and top backlinks

• Referring domains: the hundred most authoritative ones, with their authority (0-1000), their link count and the date the first link was seen. A histogram splits them by authority band — 0-99, 100-299, 300-499, 500-699, 700-1000 — which shows at a glance whether the profile rests on a few large sources or on a long tail.
• Link competitors: domains earning links from the same sources as you, whether or not you listed them as competitors. DataForSEO identifies them from shared sources, and the list sometimes surfaces rivals the client never named.
• Top backlinks: the highest-rated individual links, with their source page, anchor, authority and dofollow or nofollow status.

Data freshness and the Refresh button

The collected profile is kept and shown whatever its age, with its date: opening the page never triggers a paid call. Until a profile has been collected, the page shows a Run analysis button. The Sunday scan then refreshes the profile every week. The Refresh button, top right, forces a fresh reading right away — handy the day before a client meeting, after a link campaign or after a site rebuild. An empty response is never cached: a passing outage will not freeze the page for a week. And if a single block fails (the anchors, say), the others stay on screen and a « Partial data » line names what is missing, instead of blanking everything. The comparison between two periods shows faster; if it takes too long, a Retry button replaces the endless loading.

No market selector, and that is on purpose

Unlike the rank tracking, search volume or press pages, this one has no market selector. The DataForSEO Backlinks API accepts neither a country nor a language: a link is a link whether you look at it from Canada or from France, and a domain's profile is the same everywhere. Adding a selector would multiply paid calls without changing a single line of the result. If you miss the setting here, it is because it has no business being here.