Auditing a 200-post archive: a worked example with made-up numbers
Your blog has two hundred old posts and nobody remembers what they link to. This example, with invented figures, shows how many calls it takes and what the credit bill looks like.
SEOBlogBot research team ·
This is an example
The blog, the post count and the results below are hypothetical. Replace each figure with your own before you plan anything.
The setup
Imagine one blog, 200 posts.
A crawl of your own site lists 1,150 outbound links, which reduce to about 410 unique hosts after you strip duplicates and your own domain. A manual Ahrefs free backlink check on 410 hosts is an afternoon of pasting and a headache, so you decide to script it.
The plan is three phases: rate the hosts, inspect the bad ones, then look at the blog itself. Expect the whole run, from export to a cleaned archive, to fit inside a working week for one editor and one developer. Then one more pass.
Phase by phase
1
Rate all 410 hosts in one call
bulk_dr accepts up to 2,000 domains, so one call at 1 credit does it. Sort the rating column ascending.
2
Inspect the lowest slice
Say 38 hosts come back at 0 or 1. Open each cited page by hand; most are dead sites or parked domains. Plan on 20 minutes per ten pages.
3
Pull the blog's own link data
refdomains and anchors for your domain are 2 credits each, 4 in total. Compare the anchors with the posts you worried about.
4
Fix, then log
Replace or remove the bad citations, record the date and the credits spent in the audit log.
The bill for this hypothetical run
Call
Credits
Dollars
bulk_dr on 410 hosts
1
$0.0001
refdomains on the blog
2
$0.0002
anchors on the blog
2
$0.0002
Total
5
$0.0005
Reading the result without fooling yourself
Zero is a clue, not a verdict.
A host with no referring domains in our index may be a very new site, a private intranet page or a genuine parked domain, and only a human can tell which. Treat the rating column as a sorting key for your attention, never as a score that decides on its own.
What would surprise you
The cost is not the problem.
The time goes to reading 38 pages and deciding what to do with each, and no API removes that step. Plan for an editor day, and remember that our data refreshes roughly monthly, so a host that looks empty may simply be new to the index.
To keep the archive clean afterwards, put the gate in front of new posts and schedule the monthly audit. The cost table lists every action's price.
Before you start your own sweep
Export outbound links from rendered HTML, not Markdown.
Strip your own domain and common social hosts first.
Keep the raw JSON of each response next to the date.
Write the floor down before you look at the results.
Budget editor time, not credits.
Variations on the same example
Change the size, keep the method.
A blog with 2,000 posts has perhaps 3,000 unique hosts, which still fits in two bulk_dr calls at 1 credit each. A network of ten blogs repeats the plan ten times and costs about five cents, so the budget never decides the project; the review time does.
Mind the rendering step. On WordPress, export the rendered HTML through the REST API, and on Hugo or Eleventy read the built folder, since the output already has every link a visitor will see. Ghost sites can use the Content API for the same job.