Referring Domains: Reading the Metric Right
Referring domains are the number of distinct websites linking to you – not the number of links. The difference sounds small but decides whether your link analysis is worth anything. This guide shows how to read the metric: its ratio to backlinks, healthy growth patterns, and what to do about lost domains.
Scope: this article is about the metric in daily use – how to interpret it, its trends and its pitfalls. How domain, link and IP popularity are defined as measures is covered under domain popularity.
1. Backlinks versus referring domains
An example makes the difference tangible. Two sites, both with 1,000 backlinks:
| Site A | Site B | |
|---|---|---|
| Backlinks | 1,000 | 1,000 |
| Referring domains | 12 | 430 |
| Links per domain | 83 | 2.3 |
| Reading | Few sources, often site-wide footer or sidebar links | Broad spread, typical of earned links |
To Google, B is considerably stronger. The reason lies in how link value works: the first link from a new domain carries the most, each additional one from the same source considerably less. Holding 83 links from one domain leaves you with little more than one.
Hence: referring domains are the more meaningful number. A raw backlink count is a vanity metric – a single site-wide footer link can push it into the tens of thousands.
2. The ratio as a diagnostic
It gets interesting when you set the two against each other. Links divided by referring domains produces a figure that says a lot about where a profile came from:
| Links per domain | Typical cause | Reading |
|---|---|---|
| 1 to 3 | Editorial mentions | Healthy |
| 4 to 10 | Repeat mentions, blogrolls, partner pages | Unremarkable |
| 11 to 50 | Site-wide elements such as footer or sidebar | Look closer |
| over 50 | Widgets, templates, bought placements | Warning sign |
A high figure is not proof of manipulation – a popular theme with a credit link in the footer produces the same pattern. But it is always a reason to inspect the sources individually.
3. Reading growth patterns
The trend over time says more than any snapshot. Three patterns come up repeatedly:
- Steady growth. New domains arriving continuously, small spikes around publications. The normal picture for an active site.
- A sudden spike. Hundreds of new domains within days. Plausible during genuine press coverage – otherwise a hint of bought links.
- Flat domains, rising backlinks. The link count grows while the domain count does not: you are only accumulating more links from sources you already had. That growth is effectively worthless.
The third pattern is the most dangerous, because it looks like success in a report. Only the domain count exposes it.
4. Segmenting by quality
A bare count of referring domains is not yet analysis. It becomes meaningful once you split it. Four cuts that earn their keep in practice:
| Cut | The question behind it |
|---|---|
| By dofollow/nofollow | How many domains pass any value at all? |
| By topical proximity | Do links come from your field or from random corners? |
| By link placement | Are links inside the content, or in footers and directories? |
| By first seen | How many domains arrived in the last 12 months? |
The combination of topical proximity and placement is especially useful: 40 genuinely relevant domains with in-content links are worth more than 400 arbitrary directory entries. The raw number hides exactly that.
5. Making sense of lost domains
Every link profile keeps losing domains – pages get rewritten, articles deleted, companies close. Losing a few percent a year is normal and no cause for concern.
Before reacting, check the cause. It is often mundane:
- The linking page no longer exists. Nothing to do.
- The link was edited out of the text. A polite follow-up is worth it if the mention still stands.
- The link now points at a 404 on your site. The most common self-inflicted case – redirect it to the right target with a 301.
- The link was switched to
nofollow. Usually a policy change at the source. - The tool simply stopped crawling the source. Not a real loss, just a gap in the vendor’s index.
The most rewarding case is number 3: an external link pointing at a URL that no longer exists on your site. Those links are already earned – you only have to route them to a valid target again. After any relaunch this is the fastest recovery available.
6. Benchmarking against competitors
An absolute figure says little. It only becomes useful compared with the sites ranking for the same queries. A sensible approach:
- Pick three to five real competitors from the search results – not the ones sales talks about.
- Compare referring domains, not backlinks.
- Build the intersection: domains linking to several competitors but not to you. That is the most concrete target list any link analysis can produce.
- Compare monthly gains rather than totals – that shows who is catching up right now.
If your domain count is level with theirs and you still rank worse, link building is not your bottleneck. Then it is worth looking at content, search intent and internal structure.
7. Why every tool shows different numbers
The same site has three different domain counts across three vendors. That is not a bug but unavoidable:
- Different indexes. Each vendor crawls a different slice of the web.
- Different freshness. Some hold links for months, others drop them quickly.
- Different filters. What counts as spam and gets discarded is each vendor’s own call.
- The Search Console links report shows Google’s own view, but truncated and without quality data.
The practical consequence: use one tool for trends and stay with it. Switching sources produces a jump that is only a change of measuring instrument.
Pro tip: before ticking off a referring domain in a report, open the linking page.
Rank-O-Saur shows you whether that page is indexable, how many links it emits in total, and which
rel attribute your link carries. A non-indexable page with 200 outbound links counts as a
referring domain in every statistic – and does nothing for you.
8. Common mistakes
- Reporting backlinks instead of domains. The bigger number looks better and says less.
- Optimising for the domain count. Make a number the goal and you end up in cheap directories – many domains, no effect.
- Comparing numbers across tools. Different indexes, incomparable results.
- Treating every loss as a problem. Natural churn is normal; only 404 targets are a real action item.
- Counting nofollow domains without flagging them. It distorts every analysis.
- Skipping the intersection analysis. It is the most valuable part of a competitive comparison and the most frequently skipped.
9. Checklist
- Report referring domains rather than backlinks as the headline metric.
- Calculate links per domain and investigate outliers.
- Look at the trend across at least twelve months.
- Segment by dofollow/nofollow, topic and placement.
- Check lost domains for 404 targets and recover them with a 301.
- Benchmark against three to five real competitors.
- Build the intersection of competitor domains as a target list.
- Stay with one tool so trends remain comparable.
- Use new domains per month as the progress measure.
10. Frequently Asked Questions
What is the difference between backlinks and referring domains?
Backlinks count every individual link, referring domains only the distinct websites behind them. If one domain links 500 times through its footer, that is 500 backlinks but one referring domain. The domain count is therefore the more meaningful metric.
How many referring domains do I need?
There is no universal number. What matters is the comparison with sites ranking for your queries. If your domain count is comparable and you still rank worse, the bottleneck is not links but content.
Are multiple links from the same domain useless?
Not useless, but sharply diminishing. The first link from a domain carries the most, each further one considerably less. Ten links from ten domains are therefore far more effective than ten links from one.
Is losing referring domains a problem?
Some churn is normal, as pages get rewritten or deleted. Action is mainly needed where the link points at a 404 on your site – a redirect to the right target recovers that link.
Why do tools report different numbers?
Because each vendor runs its own index with its own coverage, freshness and spam filtering. None of the numbers is Google’s view. For trends you should therefore stick with the same tool throughout.
Do nofollow links count as referring domains?
In most tools, yes, they are included. For assessment you should look at dofollow and nofollow separately though, because only the former passes value. A profile with a high nofollow share looks stronger in the raw number than it is.
How do I find new link prospects with this metric?
Through the intersection: identify domains linking to several of your competitors but not to you. Those sources demonstrably link within your field and are the most promising candidates for outreach.
Read next
