You have a domain’s Spam Score in one report and a different number in another export. Before deciding which is worse, check whether the numbers describe the same metric. The product, target, scale and collection date belong beside the value.
This is particularly important when an older tutorial describes a 0–17 scale. Applying that scale to a current percentage, or converting a separate API field into a percentage yourself, can produce a confident but unsupported conclusion.
Identify the score before interpreting it
Start with the current report or export heading. Record the provider, product, queried address and unit. Check whether you are reading the queried site’s score or a breakdown of scores for sites linking to it. Those are different observations.
Moz’s current Spam Score documentation describes a percentage based on sites sharing selected features with sites found penalised or banned in its dataset. It does not confirm a penalty on the particular site you queried.
Use our quick website score-check guide for an initial lookup. The additional task here is to reconcile the evidence before comparing reports or acting on a change.
Keep documented scales in separate columns
| Report or field | Documented presentation | Comparison rule |
|---|---|---|
| Current Moz Link Explorer Spam Score | Percentage-based metric. | Keep the target, report definition and date with the percentage. |
| DomDetailer API v2 mozSpam field | The API reference documents 0–17. | Preserve this field’s documented unit; do not relabel it as the current Moz percentage. |
| Missing evaluation or failed lookup | No confirmed measurement. | Record unknown or failed, rather than inventing a zero. |
DomDetailer is part of the same publishing family as DomainNexus. Its v2 field reference is the source for that specific schema. The fact that a field contains “Moz” in its name does not justify silently equating its scale or freshness with a different current report.
If provenance or compatibility is material to your decision, verify it with the provider’s current documentation before comparison. Preserve the original export. Changing its column label does not solve a disagreement about what the data represents.
Create a measurement record you can explain
- Write the exact queried address and the scope selected in the interface.
- Save the report or export date and any separate crawl date exposed.
- Retain the original field name, unit and provider definition.
- Mark unavailable evaluations and request errors separately.
- Add the question the result raises, instead of labelling the domain simply good or bad.
In a hypothetical comparison, a current report shows 30% while a separate API export returns 6 on a documented 0–17 field. Dividing six by seventeen does not establish a comparable Spam Score percentage. The useful finding is that the measurements need clarification, not that one report proves an improvement.
Our metrics guide explains similar scope problems with domain, page and link measures. Do not average unlike values to create a reassuring middle figure.
Turn the result into a concrete investigation
Inspect the site and a useful sample of important referring pages. Record whether their content, links and destination URLs support a relevant concern. A statistical similarity signal is not evidence that every unfamiliar link is harmful or that adding cosmetic features will improve search performance.
For a domain with history, our pre-purchase review adds archived-use and acquisition checks. A low number does not certify a clean history, current registration availability or rights to reuse the former website.
Use official findings for official-status questions
For an authorised property, Google’s Manual Actions report provides manual search-policy findings. Its separate Security Issues report concerns harmful or compromised content. An email reputation listing is another system again.
Do not bulk-disavow links to make a vendor number smaller. Google’s disavow guidance limits that advanced action to a considerable problematic link situation combined with a manual action or likely manual action. Most sites do not need it.
Finish with the recorded measurement, the evidence inspected and the justified next step. Improving the website should serve users and address actual problems; lowering a dashboard value alone is not a search-ranking strategy.