Chronometrica vs Fortuitous Pirate: How to Check Empty Google Ads Results

Choose the outcome contract your work needs

The main choice is whether your process requires declared results for each target. Here, an outcome contract means the documented information a program provides about what happened during a check.

Chronometrica documents `targetOutcomes` and separates `no_results` from an unavailable public source. It also documents useful creatives saved, duplicate or unusable creatives skipped, and final stop reasons. Choose it when those records are required to review a zero-row result. This recommendation rests on its documented contract. Chronometrica Actor

Keep Fortuitous Pirate when its single-query input, `advertiserId`, and `proxyConfiguration` controls meet your workflow. Its checked schema accepts the historical Germany query and both blank platform and format values. Inspect its actual current logs and output before deciding whether to switch. Its short README does not establish that logs or target outcomes are absent. Fortuitous Pirate Actor

These are selection rules, not measured performance results. The evidence does not establish a speed, price, accuracy, coverage, or reliability advantage.

Separate the saved row count from the source state

An empty dataset answers one narrow question: how many creative rows were saved? The source state answers a different question: what could the program reach and use?

Keep three possible explanations separate during review.

First, a check may reach the public source and find no useful results for the target and filters. The next step is to inspect the target identity and search scope. A domain, advertiser, region, or date mismatch can change what you expect to find. Do not select a cause until the evidence supports it.

Second, the public source may be unavailable to the check. In that case, a zero-row dataset does not describe the advertiser’s ad activity. Review source access before drawing a conclusion about the search scope.

Third, a creative page may be visible while its public assets are unusable for collection. Chronometrica’s result rule can skip a creative that lacks usable public media or a preview. A visible page and a saved creative row therefore do not have to correspond one for one. Chronometrica Actor

This distinction also changes how you report the result. Write “zero useful rows saved under these settings” when that is what you know. Do not shorten it to “the advertiser has no ads.” The shorter statement removes the filters, source conditions, and asset limits that give the result its meaning.

The configured cap is another separate fact. A request for at most one row does not promise one row. It sets a maximum. If the check saves zero, raising the cap alone does not explain the earlier result.

Prepare the Germany input before you run

A Task is a saved input preset for an Actor. The linked regional research Task is titled Canada. Use it as a starting point, then inspect its actual saved input. The checked public page was available, but that page check did not verify its saved field values. Regional research Task

For the Germany case in the support issue, use this input:

{
  "targets": [
    "ritzenberger.de"
  ],
  "region": "DE",
  "platform": "ALL",
  "adFormat": "ALL",
  "includeDeliveryDetails": false,
  "maxItems": 1
}

This input passed a check against the saved public JSON Schema. A JSON Schema defines allowed input fields, types, and values. All fields shown above exist in the checked schema. This confirms the input shape; it does not confirm live source access or promise a result. Chronometrica Actor input

Use these steps before the small check:

1. Preserve the domain exactly as reported. Keep any separately known advertiser identity with your review notes.
2. Open the Task input and replace the region with `DE`. Do not rely on the Task title or assume its saved region already matches.
3. Set `platform` and `adFormat` to `ALL` for this check.
4. Inspect every saved field. Remove any uncertainty about filters before comparing the output with a browser view.
5. Keep `maxItems` at `1` for the initial check shown here.

Record the date scope used in your public-source comparison. If the Actor does not expose that scope in the evidence you inspect, mark it as unknown. Do not add an input field that you have not checked against the schema.

The checked schema permits 1–100 entries in `targets` and an integer from 1–5,000 for `maxItems`. If your original requirement is a single run configured to return more than 5,000 results, this Actor does not meet that requirement. A smaller check does not satisfy a larger all-results request. Chronometrica Actor input

Review the outcome beside the dataset

After an intended small check, review the saved rows and the outcome records together. Start with the target. Then inspect the filters, source state, skipped creatives, and stop reason. Chronometrica documents these through its target context and outcome accounting. Chronometrica Actor

Use a short review ledger. A ledger here means a written record of the evidence for one target.

1. Target resolved: Record the advertiser identity the check resolved. If it differs from the intended advertiser, keep that difference visible.
2. Filters requested: Record the region, platform, format, date scope where known, and requested cap.
3. Source state: Record whether the outcome supports no useful results or an unavailable public source.
4. Useful rows: Record the actual number saved. Use zero only when the output reports or establishes zero.
5. Skipped creatives: Record duplicate and unusable creative counts as provided. Preserve unknown or missing values.
6. Final stop reason: Record why the check ended. Do not infer that reason from dataset size alone.
7. Unresolved cause: State what the available evidence still cannot explain.

Keep the ledger tied to one target and one set of settings. Otherwise, a later filter change can make an earlier result appear inconsistent when the two checks asked different questions.

Read skipped counts with care. A duplicate skip and an unusable-asset skip support different interpretations. A duplicate concerns repeated material. An unusable asset concerns whether a creative can become a useful saved row. Use only the detail the output actually provides; do not invent a breakdown when it gives a combined count.

If a value is missing, write “unknown.” Missing source-state evidence is not the same as a confirmed source failure. Missing skip counts are not confirmed zero skips. This keeps the record useful for a later review.

A fictional example: equal rows, different next steps

The following example is fictional. It is not an Actor test, a live dataset, or a customer result.

An analyst checks one target twice. Both checks save zero creative rows.

Check A reaches the public source. Its outcome reports no useful results under the declared filters. The analyst’s next step is to compare the resolved advertiser and filters with the public view that created the expectation of ads. The analyst checks the region, platform, format, dates, and source time.

Check B cannot use the public source. The analyst first inspects source access. The check does not support a conclusion about the advertiser’s ads because it could not use the source.

The arithmetic is simple: Check A saved 0 rows, and Check B saved 0 rows. Their row counts are equal. Their source states are different, so their next steps differ.

Both checks used a one-row cap. That cap did not require either check to return one row. The cap limited the maximum result count; it did not guarantee a result.

Neither outcome proves that the advertiser has no ads. Check A remains limited by its target, filters, and useful-result rule. Check B remains limited by source access. The example supports a review method, not a fee estimate or a reliability claim.

Use a small acceptance check to make the decision

An acceptance check is a small test with a written pass condition. Its purpose here is to decide whether the output can answer your empty-result question.

Set the pass condition before the run: “For this target, I can identify the resolved advertiser, requested filters, source state, saved rows, skipped creatives, and final stop reason.” A positive row count is not required to pass this check. An explained empty result can be useful evidence.

Follow this method:

1. Write the expected public evidence. Note the domain or advertiser, the region, the visible creative reference, and when you observed it. Keep the original input unchanged in your notes.
2. Validate the input shape. Check allowed fields and values against the Actor’s current schema. Separate a schema error from a search that is valid but produces no useful rows.
3. Run the intended small check. Use the explicit settings you reviewed. Do not change several filters merely to obtain a positive count.
4. Complete the ledger. Read the outcome evidence beside the dataset. Mark every unresolved value as unknown.
5. Choose the next action. Review identity and filters for no useful results, source access for an unavailable source, or asset limits for unusable creatives.

If a further check is needed, change one relevant setting and record why. This makes the new result easier to interpret. A larger dataset is not automatically a better explanation.

For unresolved cases, prepare a support report with sanitized input, expected public evidence, the observed outcome, and the remaining question. Sanitized means that private or sensitive details have been removed. Do not include proxy credentials or other secrets.

This process also gives you a fair basis for keeping an existing tool. Judge whether its actual evidence meets your acceptance condition. A historical support report alone cannot decide that question for a present run.

Once the empty-result question is resolved, use the regional creative comparison guide to compare saved creative sets across regions. That is a separate analysis. Establish what the empty check means before using it as evidence of a regional difference.