Social Metrics Reconciler: Map and Compare Business Reports

USD 0.25 per completed report (one per run).

Combine customer-provided business-account CSV or JSON metric reports into a consistent JSON dataset. Map each report's own headers to canonical fields; the Actor returns row records, per-account period summaries, and deltas for compatible windows.

This Actor reads only files you provide. It does not sign in to Instagram, TikTok, or other platforms, fetch account URLs, verify a source, or promise compatibility with a platform's native export format. The included sample is synthetic and is not a native export.

Input

Add one report per CSV or JSON file in reports. For each report, set format, provide CSV text or JSON rows in data, and map exact source header names to canonical fields. Required mappings are business_account_key, platform, period_start, and period_end. Map at least one metric: impressions, views, reach, likes, comments_count, shares, or clicks. post_key is optional. Periods use YYYY-MM-DD.

You may combine multiple business accounts and periods. Use a stable opaque business account key, a platform label, and a stable post key where post-level rows are supplied. Metric counts must be non-negative integers. Blank and missing metrics remain null; zero remains zero. Unknown source columns are discarded and do not appear in output.

Limits are 5,000,000 UTF-8 bytes for the complete input, 20 business accounts, and 1,000 source rows per business account across all supplied periods. Duplicate account/platform/period/post keys are rejected with a safe error code.

Do not include personal profiles, usernames or handles identifying private people, follower lists, comment text, names, email addresses, audience demographics, or contact records. Supply business aggregate counts only. The Actor does not log source records or mapped values.

Output

The complete JSON result is stored in the default key-value store at key OUTPUT. The default dataset contains one compact index item with record counts and result_key: "OUTPUT". The output schema links directly to both locations. This stores results larger than Apify's 9 MB per-dataset-item limit without splitting the JSON arrays.

The OUTPUT record contains:

• canonical_records: only the allowlisted canonical fields and mapped metric values.
• account_summaries: account/platform/period totals. Impressions, views, likes, comments, shares, and clicks are summed only when all rows are present. Account-level and post-level rows cannot be mixed in one account period. Post-level reach is never presented as account reach; aggregate reach is available only from a single account-level row.
• period_deltas: contiguous adjacent periods are compared only within the same account and platform, for equal-length non-overlapping periods and matching metric definition IDs. Gaps, incompatible periods, missing values, or mismatched metric definitions produce an unavailable delta with a reason.

Metrics on different platforms remain separate. A platform label is user-supplied metadata, not source verification. By default, a metric definition ID is scoped to the platform and canonical metric name. Set an explicit stable metric_definition_ids value when the metric's meaning changes; mismatched IDs prevent a delta.

Pricing and support

USD 0.25 per completed report (one per run).

A successful run produces one combined report, covering all supplied accounts and periods within the input limits. The charge is once for that completed report. There is no Actor-start fee; standard Apify platform usage is included. A run that fails before producing its report is not charged.

Example

The prefilled two-period sample uses fabricated business data and two different source header sets. The credential-free Apify SDK 4.1.0 local-storage smoke produced this result summary:

{
  "canonical_record_count": 2,
  "account_summary_count": 2,
  "period_delta_count": 1,
  "sample_impressions_delta": 4,
  "result_key": "OUTPUT"
}

It does not establish any Instagram or TikTok native export compatibility.