Advertiser activity dataset
Understand advertiser records in an Instagram export
Advanced Insights organizes supported advertiser names, matching-method groups, targeting categories, and interest labels while keeping exported facts separate from deduplicated counts and summaries.
- Full Instagram JSON export
- Direct and derived labels
- Distinct-count logic
- No ad-delivery claims
Supported scope
What this page is designed to help with
Advertiser matches
Read advertiser names grouped under the matching mechanisms represented in the export file.
Targeting categories
Show categories exported alongside audience and advertiser information when that dataset is present.
Inferred interests
Organize activity-associated interest labels without treating them as verified personal attributes.
Duplicate-aware totals
Compare raw list entries with distinct names when one advertiser appears under more than one mechanism.
Source records
What is directly exported
The advertiser importer reads advertisers_using_your_activity_or_information.json when it appears under the Instagram ads and businesses area. Each supported label can contain a vector of advertiser names tied to a matching description.
It also reads other_categories_used_to_reach_you.json and supported interest-category records. The direct facts are the strings and grouping labels present in those files, not an explanation of how a specific company obtained information.
Calculations
What Advanced Insights derives
The report counts entries per matching mechanism, normalizes non-empty names, and deduplicates the combined list to produce a distinct advertiser count. A repeated name contributes to the raw mechanism totals but only once to the distinct total.
Targeting and interest totals are counts of unique non-empty labels in the supported files. They are summaries of the export, not scores of advertising intensity.
Raw entries across mechanisms: 42
Distinct advertiser names: 36
Repeated names across mechanisms: 6
Targeting-category labels: 9The subtraction is a derived duplicate count; the names and mechanism labels are direct records.
What it does not prove
A match is not the same as an ad impression
An advertiser name in an audience-match list does not prove that you saw an ad, clicked an ad, bought something, visited a store, or supplied a particular identifier. The file can describe a matching relationship without exposing its complete origin or outcome.
The current dataset may also lack event timestamps. Without dated source records, the report does not assign a match to a particular day or claim it is still active.
Coverage
Missing, empty, and changed files remain different states
Personal, creator, and business accounts can receive different ad-related files. A missing advertiser section can mean the category was not requested, Instagram did not supply it, the file was empty, or the structure changed.
Coverage reports the available state. It does not turn absence into a zero-advertiser claim.
Sensitivity
Names and inferred labels deserve careful handling
Advertiser names, locations of interest, possible identifiers, and inferred labels can be sensitive even when they are platform-generated. Advanced Insights presents counts separately from detailed values and keeps private reports out of sitemaps and search indexing.
Review a label as a record associated with the export, not as a verified statement about identity, intent, health, politics, finances, or other personal traits.
Questions
Common questions
Why can raw advertiser entries exceed distinct advertisers?
The same normalized name can appear under multiple matching mechanisms. Raw totals count those placements; the distinct total counts the name once.
Does an advertiser match mean I saw its ad?
No. A match record and an impression are different facts. The export must contain a separate viewed-ad record to support a viewing statement.
Are inferred interests confirmed facts about me?
No. They are labels associated with activity in the export and should be described as platform inferences.
Keep exploring
Related guides and tools
Review the broader supported dataset library and coverage model.
Privacy and account history data
Understand profile facts, change history, inferred information, and time limits.
Instagram export troubleshooting
Fix missing, partial, or unrecognized Advanced Insights categories.
Review processing, retention, and private report controls.
Inspect supported advertiser evidence
Upload a fuller Instagram JSON export to see which advertiser, targeting, and inferred-interest records are actually present.
Analyze Advanced Insights →