Dataset Audit — what you actually paid for

You paid per result. This tells you what you got.

Point it at the dataset of any run — from any Actor — and it returns the receipt: how many records were duplicates, how many ignored the filters you asked for, and what a usable record really cost you.

Why

Across the Apify Store the same complaint keeps coming back, in the reviews of tools that otherwise work fine:

"I expected 2,000 results for $10. I got 748 unique ones. That's $0.01337 per unique job, not the $0.005 advertised — 2.67 times the price."

That user did the arithmetic by hand, in a review box. This Actor does it for anyone, in about a second.

What it gives you

• Duplicates — how many, and exactly which keys repeat
• Unmet criteria — records that ignored the filters you set (date windows, minimum thresholds, required fields)
• Empty fields — how often each field came back blank
• Cost per usable record — what you paid divided by what you can actually use
• A live Excel receipt — input cell at the top, real formulas underneath. Change the amount you paid and every figure recalculates. It is a working model, not a screenshot.

Input

Field	Meaning
datasetId	The dataset to audit. Pick it from the list — that grants read access to that dataset only.
dedupeKeys	Fields that identify one record, e.g. ["jobId"]. Empty compares whole rows.
filters	What you had asked for, e.g. [{"field":"postedAt","op":"after","value":"2026-07-01"}]
amountPaidUsd	What that run cost you, from your Apify console
locale	en or it — the language of the Excel receipt

Operators: nonEmpty, equals, oneOf, min, max, after, before, matches.

A field that is missing counts as a failure, not as a case to skip: that is precisely the record you believed you had excluded when you paid.

What it does NOT do

Said plainly, because the absence of this section is why tools get 3 stars:

• It does not read your account. It runs with limited permissions and can only see the one dataset you select. Not your other runs, not your other storages, not your billing.
• It cannot read the run's cost by itself. Apify does not expose that to a limited-permission Actor, so you type the amount in. If you leave it out, you still get duplicates and unmet criteria — just no cost per record.
• It does not judge whether the data is true. It checks the data against the criteria you declare. If a scraper invents a plausible email, this will not know — it will only tell you the field was not empty.
• It does not compare across runs. Duplicates are counted inside one dataset. If the same record was sold to you in two separate runs, that is not caught yet.
• It does not store anything. Your rows are read, counted, and dropped. The only output is on your own account.

Cost

Free. If it ever stops being free you will read about it here first, with the notice period the platform requires.

---

Built by ledgerworks — financial tools that get the numbers right. Same workshop as 13-Week Cash Flow.