Know what your
data is saying.

The web analyzer builds relationship, activity, content, and audience reports from supported Instagram export files. It does not connect to Instagram, fetch live information, or infer private activity.

Which files we read

ZIP imports select Instagram relationship files from connections/followers_and_following, supported activity and media JSON from your_instagram_activity, and audience and post insights from logged_information/past_instagram_insights. Message metadata and account identity support interaction reports. Only files under the selected Instagram export root are used. Photos, videos, Threads data, and unrelated entries are skipped. Direct JSON selections provide a relationship-only preview.

As in the iOS export analyzer, follower names come from all string_list_data.value records; following names come from each record’s title. Older supported formats can use a profile URL or username value when the primary field is absent. Malformed required records stop the web comparison instead of silently dropping accounts.

How the relationship lists work

What makes a comparison useful

Select complete files from one account and one export, including every numbered follower file. The tool can detect missing inputs and gaps between selected parts, but cannot prove that the export is complete or that every file belongs to the same account.

Relationship preview lists are ordered by username. Their position does not indicate recency, importance, or interaction level.

Activity, content, and audience reports

New Followers and Lost Followers compare two imports from the same account. A first import is a baseline; a relationship timestamp does not identify an unfollow event. Use Add earlier ZIP in the Relationships section. The website checks account identity and export dates, but cannot prove account ownership or export completeness.

Activity-based reports, including Top Followers, DM Soulmate, and One-Sided Connection, need their corresponding likes, comments, story activity, searches, or message records. These are interpretations of the activity that was exported, not proof of another person’s feelings. Your outgoing likes, story views, and profile searches must not be read as incoming activity. Post and audience reports require the relevant exported insights. The website uses the native app’s scoring weights and eligibility thresholds, with up to five entries shown in each preview.

Users Blocked means accounts you blocked. It is different from Who Blocked Me. Missing activity or insight files mean that a report may be unavailable; they do not prove zero activity.

What this website does not measure

Relationship lists cannot identify visitors, blocks against you, exact unfollow times, or feelings. Exported post insights can include profile-visit totals, and audience insights can include period follow/unfollow totals. These aggregate counts do not identify individual people. Lost Interest and Secret Admirers are not calculated from these records.

How scores and availability work

People scores combine recorded messages, chat days, comments, likes, story activity, and searches using the iOS weights. Top Followers includes only followers with a positive score. DM Soulmate needs at least 10 messages across 5 days. Featured winners are excluded from later featured categories in the same order as the app.

Group and unidentified conversations contribute to message totals but are excluded from individual rankings. Direct-message attribution uses sender, account, and thread labels as best-effort signals. One-Sided Connection compares outgoing actions with recorded incoming DMs; incoming likes and views are not available. It excludes accounts with unattributed message direction.

The native engagement estimate is a capped weighted activity score relative to follower count. It is not a measured received-engagement rate. Missing sources remain unavailable, and invalid optional files do not prevent other report sections from loading.

Posting-time and hashtag scores use likes + comments × 3 + saves × 4. Time recommendations use the browser’s time zone and eligible timestamped posts. Fewer than 10 samples means low confidence, 10–29 means medium, and 30 or more means high. Empty heatmap cells have no samples. The time recommendations show overall time ranges; media-type-specific recommendations remain in the app.

Local processing and limits

Files are parsed in a browser worker. The analyzer reads only supported JSON and skips unrelated archive entries. Selected files are limited to 100 MB, each supported JSON file to 12 MB, and total extracted analysis data to 32 MB. If an export exceeds these limits, select the relationship files directly.

Reports are held in memory for the current page session. No history is saved by this website. Clear the report, reload, or close the page to end the session.

Preview and app access

The free tool provides summaries and up to five entries per list, ranking, post table, hashtag list, or demographic category. The app provides full lists with an in-app subscription. The preview limit is a user-interface boundary, not secure access control over your own local data.

Sample reports are fictional and marked as samples. Website files are not transferred through the App Store. Keep your original export to import in the app.