Case sketch
Competitors backlink analysis on a roundup: before and after
An invented comparison page, a measured gap and a small order. Every figure below is hypothetical, so use it as a method, never as a forecast.
No real site, customer or result is described. Numbers were chosen to be round and plausible. Your own gap, prices and timing will differ.
Before: the gap, measured
| Page | Position | Referring domains to the URL | Notes |
|---|---|---|---|
| Our roundup | 11 | 9 | Updated in September 2026 |
| Rival A | 4 | 21 | Strong brand |
| Rival B | 6 | 15 | Older, thin copy |
| Rival C | 8 | 12 | Same topical blogs |
Reading the table
Competitors backlink analysis starts with one URL, not one site. Rival C holds 12 referring domains to its page, against our 9. That is a gap of three.
Rival B interests us more, because its copy is thin and it still sits above us at 15. Fifteen links are carrying a weak page, which tells us links are a real lever in this particular result, at least for now.
We would pull the list of linking domains from Ahrefs or Semrush, mark those on readable, relevant blogs, and note which ones link to two rivals at once. Those overlap sites are the likeliest to accept a link to us, and a cleaner donor choice in the marketplace is the filter for niche and traffic.
The order
Refresh the page first
Update prices, add test notes and photos, and fix the title. Wait a week so Google can recrawl.
Buy three placements
Three relevant kitchen and coffee blogs at about $30 each, plain anchors, written in the automatic mode.
Write the date down
Position 11, 9 referring domains, the exact clicks of the last 28 days.
Leave it alone
No further tweaks for eight weeks, otherwise you cannot tell what worked.
Four ways competitors backlink analysis misleads
Site-wide numbers
A rival with a Domain Rating of 70 may have a single weak link to the roundup you compete against. Count links to the URL.
Sitewide footer links
Hundreds of links from one blogroll count as one referring domain for most purposes. Do not copy them.
Dead donors
Some linking pages in the export are gone or deindexed. Open a sample before you trust the total.
Brand effect
Rival A ranks partly because readers search its name. No link order replicates that, so do not chase it.
After: three imagined outcomes
| Outcome | What it suggests | Next move |
|---|---|---|
| Position 7, clicks up | Links were part of the gap | Order two more, same style |
| Position 10, clicks flat | Small effect | Look at title, snippet and rivals' content |
| No change | Links were not the bottleneck | Stop buying and revisit the decision tree |
What would make us distrust the result
A Google core update inside the window. Two changes made at once. A rival that also gained links, or a SERP feature such as an AI Overview pushing the page down. Any of these muddies the picture, so write down what else changed.
That is why the budget plans cap every wave at a few orders: one change at a time, logged with a date, so a surprise has fewer suspects.