Amazon PPC Ad Report Audit

What it does

Amazon PPC Analyzer by mguozhen reviews exported advertising evidence rather than building a new campaign. It focuses on Sponsored Products Search Term Reports, campaign performance exports, keyword reports, and small metric snapshots. The skill separates observed spend, clicks, orders, and attributed sales from assumptions, then highlights wasted spend, converting search terms, bid opportunities, negative-keyword candidates, campaign structure issues, and budget pressure. It can also turn the reviewed period into a consistent weekly report. It does not prove that a recommendation improved an account.

When to use

Use it after exporting data from Seller Central Campaign Manager, during a weekly PPC review, or before deciding whether a search term should be harvested or excluded. It is useful when the question is "what happened in this report and what should we review next?" It is intentionally different from the existing campaign builder: it audits supplied evidence and prepares decisions; it does not design a fresh campaign structure or apply changes.

If a CSV is already available, skip the Amazon Ads report downloader and begin here. After validation, use the Amazon Ads HTML Report when you need a self-contained offline view.

Inputs

• A CSV export such as a Sponsored Products Search Term Report, Campaign Performance Report, or Keyword Report.
• The reporting period, marketplace, target ACoS or profit guardrail, and any launch or optimization context.
• Optional campaign screenshots, a verbal account summary, or a few metrics when a full export is unavailable.

The following is a redacted demonstration input, not an execution result:

date,campaign,search_term,clicks,spend,sales,orders
2026-09-08,SP-Exact,insulated bottle,12,18.00,60.00,2
2026-09-08,SP-Auto,water bottle,8,12.00,0.00,0
2026-09-08,SP-Broad,steel bottle,20,30.00,90.00,3

Outputs

The audit returns an executive summary, quick wins, structural issues, a priority table, and a next-audit checklist. It can identify terms with spend but no orders, converting terms worth testing in exact match, possible negatives, over- or under-funded campaigns, and data that is too thin for a strong conclusion. ACoS is calculated as spend / attributed sales * 100. In the first demo row, 18 / 60 * 100 = 30%; in the third row, 30 / 90 * 100 = 33.33%. A row with zero sales has no calculable ACoS and must not be silently treated as zero.

Example workflow

1. State the marketplace, date range, target ACoS, and whether the account is launching or optimizing.
2. Paste an approved, redacted report and label columns whose attribution window or currency is uncertain.
3. Recalculate the visible ACoS examples, separate facts from inferences, and ask for missing guardrails.
4. Review waste, winners, harvesting candidates, negatives, and budget pressure in that order.
5. Return proposed bid or budget changes as review items only. The CSV above is illustrative input; no Amazon call or account change was executed for it. After validation, pass the reviewed CSV to the Amazon Ads HTML Report if an offline report is needed.

Supported agents and setup

The catalog identifies this Skill for OpenClaw. Other clients may support the same SKILL.md format, but that compatibility requires independent verification. Evidence is available in the SkillHub directory listing, the fixed source SKILL.md, and the author repository. The repository metadata does not declare a license, so this entry records Not declared.

Account, API, and permission requirements

No Seller Central account or API key is required by the source Skill. The seller must provide the report or metrics and confirm currency, attribution window, and date range. The Skill declares Bash access and this catalog entry therefore records filesystem and shell permissions; local report history may contain sensitive sales and search-term data. Ecommerce Data API cannot fetch a seller's private advertising reports on its behalf. Its keyword research, reverse-lookup, and traffic-source APIs can only supplement the supplied report with marketplace context. Official Amazon Ads budget, bid, or campaign writes are not a function of this site.

Limits and human review boundary

ACoS, CTR, CVR, ROAS, and bid formulas are decision aids, not guarantees. Ask for the target ACoS or profit margin before any bid recommendation, flag small samples, and avoid aggressive changes when a term has fewer than ten clicks. Check whether launch-stage spend, promotions, stockouts, currency, attribution windows, or delayed orders distort the comparison. A human must approve every budget or bid change, preserve a rollback point, and perform any official Amazon Ads write outside this site.

Pair with Ecommerce Data API

Use amazon.keywords.research, amazon.keywords.reverse-lookup, and amazon.traffic.sources to add current keyword, reverse-ASIN, or traffic context after the seller supplies the advertising export. Keep report facts and API observations in separate columns with source dates. Do not present supplemental keyword evidence as a substitute for the private report, and do not imply that these APIs can read or write the seller's Amazon Ads account.

Direct routes: Amazon keyword research, traffic sources, and create an Ecommerce Data API key.