Scoring sheet
Backlinks test: score a link from API fields before you bid
A backlinks test is only useful if two people running it get the same number. This page gives you a four-signal sheet you can rebuild in a spreadsheet and rerun every month.
Appraiser's rule
Pay for breadth, check the wording, ignore the sticker
Referring-domain breadth deserves the most weight, because it is the one signal that is hard to fake cheaply. Anchor variety is the cheapest tell of a bought profile. Everything else is a tiebreak.
- Breadth: 40 of 100 points
- Anchor variety: 25 points
- Page strength: 20 points
- Overlap with your rivals: 15 points
Why four signals and not forty
Short lists get used. A scoring sheet with forty columns dies in a week, and the person who built it is the only one who trusts it.
The four signals below each map to one call you can run with LinkMidas, and together they cost 6 credits, or $0.0006, per candidate when you pull all of them. The free domain call handles the first screen, and only the survivors earn the paid calls.
A caveat on the weights. They are our suggestion as appraisers, not a law of search, and Google has never published a link-value formula. Change them once you have your own results to compare against.
The points sheet
| Signal | Source | Full marks when | Zero marks when | Weight |
|---|---|---|---|---|
| Breadth | refdomains (2 credits) | 1,000+ referring domains, spread over several TLDs | Fewer than 20, or nearly all on one TLD | 40 |
| Anchor variety | anchors (2 credits) | Brand and URL anchors lead; no phrase above 10% | One commercial phrase above 40% | 25 |
| Page strength | toppages (2 credits) | The page you would sit on has links of its own | All links point at the homepage | 20 |
| Rival overlap | intersect (2 credits) | Source links to two or more of your competitors | Links to nobody relevant | 15 |
A worked example, clearly hypothetical
Take three invented offers. None of them is a real seller. The numbers exist only to show the method.
Offer A asks $60 and sits on a domain with 1,000 referring domains, so its DR is 41. Offer B asks $25. Its 100 referring domains give a DR of 27. Offer C asks $140 at 10,000 referring domains, a DR of 55.
Per DR point, B is cheapest at about $0.93, A costs about $1.46 and C about $2.55. That ranking misleads. DR is a log scale: 14 points separate A from C, yet C has ten times as many referring domains. Price per point flatters the weak site and punishes the strong one.
Score them on the sheet instead. It is quick. If A scores 78, B 41 and C 90, the cost per sheet point is $0.77 for A, $0.61 for B and $1.56 for C. B looks like the bargain until you notice it fails the breadth floor. A is the buy, and C is a stretch you take only for a page that must rank.
Rebuild it in a spreadsheet in an evening
- 1
One row per candidate
Columns for domain, asking price, DR, referring domains, top anchor share, and your four points.
- 2
Fill the free columns first
Use the free appraisal routine to drop weak candidates before spending a credit.
- 3
Call the paid actions only for survivors
The recipe page has the requests. Save each response with its date.
- 4
Divide price by points
Compare dollars per sheet point, not dollars per DR point, and re-score each month when the index refreshes.