Free DR lookups need no key. Sign in for $100 in credit.
Hypothetical
Every figure here is invented to show the method. No real client, vendor or result is described.
The agency, the vendor, the domains and every number below are made up. They exist to show arithmetic and structure. Your batch will differ.
Imagine an agency that bought 40 placements for one client site, at a fixed price per link, with a promised DR floor of 30 and a 60-day replacement window. The vendor sheet lists 40 source URLs and 40 anchors.
Nothing here is unusual. The question is how to turn that sheet into a pass-fail table. We use three checkpoints: delivery day, day 30 and day 60.
| Checkpoint | Live page fetched | Host in refdomains list | Rel as ordered | Notes |
|---|---|---|---|---|
| Day 0 | 38 of 40 | 0 of 40 (not indexed yet) | 36 of 38 | 2 URLs 404, 2 rel=sponsored added |
| Day 30 | 37 of 40 | 29 of 40 | 35 of 37 | 1 more removed, index refreshed on day 22 |
| Day 60 | 36 of 40 | 34 of 40 | 35 of 36 | Replacements pending for 4 |
Day 0 shows zero hosts in the index. That is normal. A zero at delivery is not a defect; a zero after two refreshes is.
By day 60 four links are gone. Two more changed, because they carry a sponsored label that the order did not mention. Notice the difference. Whether the label counts as a defect depends on the contract, so the sheet should say what was ordered in plain words. The glossary lists the rel values you can quote.
The rating floor is checked once with bulk_dr. One call. In the example, 40 domains in one call cost 1 credit, and 3 domains score below 30 using our formula. Those three go into the ticket first because they are objective and cheap to prove.
Of the 4 missing pages, assume 3 were removed by the publisher and 1 moved to a new URL. The move is easy to miss. A redirect from the old address would show as a 301 in the saved header dump, which is why headers go into storage beside the HTML.
The 2 sponsored labels are a different class of problem. Not broken. They are not broken links, and the vendor may well say the label is the honest option. Decide in advance whether a label counts, because a dispute settled by taste wastes a week.
| Step | Calls | Credits | Dollars |
|---|---|---|---|
| bulk_dr on 40 domains | 1 | 1 | $0.0001 |
| refdomains at day 0, 30 and 60 | 3 | 6 | $0.0006 |
| anchors at day 60 | 1 | 2 | $0.0002 |
| Total | 5 | 9 | $0.0009 |
The live-page column comes from curl on the tester's machine, not from our API. We do not offer that check, and pretending we did would hide the most laborious part of the job.
Credits are almost beside the point. The labour is reading 40 pages. If you want to cut that, script the grep from the mistakes list and keep your eyes for the 4 or 5 rows that fail.
For a month-by-month view of a whole client portfolio, move on to the cost table. If rankings moved after the batch, the traffic-drop runbook is the next stop.