Instagram Unfollowers Checker: Compare JSON Lists Accurately

Use the Follower List Checker with Instagram selected and the comparison mode chosen before adding files. This guide follows a ceramics studio checking its supplier relationships, then reviewing two follower snapshots around an exhibition. You provide the records; the browser compares them locally without an Instagram login, profile lookup, or account action. Start with the relevant extracted JSON files, not the entire downloaded archive.

Choose the question before choosing the files

For a current relationship review, select relationships mode. Source A is Following: the accounts the studio follows. Source B is Followers: the accounts recorded as following the studio. The A-only bucket is labeled Not in follower list. It is a present-list mismatch, not a list of confirmed recent unfollowers. Some of those accounts may never have followed the studio at all.

For a change review, select snapshots mode instead. Both sources must describe the same type of list for the same Instagram account. Put the earlier follower list in A and the later follower list in B. A-only becomes Absent from later list, B-only becomes Added to later list, and the intersection becomes Present in both. Following snapshots can also be compared, but they answer a different question.

Write the question in your working notes before importing anything. For example, ask whether the studio follows suppliers that are absent from its current follower record, or whether its recorded audience changed around the exhibition. Do not put last month’s following file against this month’s followers file and describe the result as audience retention. Both the relationship type and the collection period would have changed.

The followers versus following guide explains the relationship view in more detail. For Instagram specifically, remember that a perfectly parsed export can still answer the wrong question if its sources are mislabeled. Before proceeding, check the account owner, list direction, and intended period together. A filename is a useful clue, but your source notes establish what the file actually represents.

Request the right Instagram information

Meta provides account-information downloads through Accounts Center. In Instagram’s settings, look for Accounts Center and Your information and permissions, then the export or download information option. Menu wording can vary with the interface available to your account. Follow the choices shown in your own account rather than assuming every device uses identical labels. The checker itself does not start this request or connect to those settings.

Choose the Instagram profile you intend to inspect, especially when several profiles appear together. Request the followers and following information where the category selection allows it, and choose JSON rather than HTML for this workflow. Use the broadest relevant date range, commonly All time, when preparing a full-list comparison. A restricted period should not be assumed to contain every relationship you want to measure.

Download the completed package to your device and extract it locally. Only the relevant supported files belong in the checker. It does not accept ZIP archives or HTML pages, and renaming either format to end in JSON will not convert its contents. If the package contains readable web pages rather than JSON data, return to the export choices or prepare a supported identity list separately.

Keep a short record of the selected profile, requested categories, date-range setting, and download date. If the export provides its own generation information, preserve that too. The moment you download a file is not necessarily the moment every relationship inside it was recorded. Do not promise an exact delivery time or describe the export as a live mirror; inspect the actual package before relying on its coverage.

Load every follower part and the following file

Look through the extracted connection data for follower files such as followers_1.json and any additional numbered parts matching followers_*.json. The folder arrangement can change, so search the extracted package rather than relying on one permanent directory name. For relationships mode, choose all follower parts from that one export for source B, and the corresponding following.json for source A. Keep exports from different requests in separate folders.

The supported Instagram follower shape is an array whose entries contain string_list_data. The adapter reads identity values or profile href fields from that structure. The supported following shape contains a relationships_following array. You do not need to turn those supported structures into spreadsheets first. However, the existence of a JSON extension alone does not mean every possible Instagram export structure has an adapter.

Suppose the studio’s package has three follower parts. Selecting only the first would compare a complete following list with an incomplete follower source, making some valid followers look absent. Add all parts belonging to the same follower record, then check the combined unique count and cleanup information. Do not add a second export just to make the total resemble the profile’s current public count; that would merge different records.

For snapshots, repeat that collection separately for each date. Earlier follower parts all go into A; later follower parts all go into B. Leave following.json out of this particular comparison. Preserve the original files unchanged so another person can reproduce the preparation. If the tool rejects a structure, inspect the error and use a prepared list of handles rather than deleting unfamiliar fields until something appears to work.

Check cleanup without confusing rows with accounts

The checker normalizes Instagram handles by removing a leading at sign and matching case consistently. Recognized Instagram profile URLs can resolve to the same handle, with tracking details discarded for comparison. Thus a pasted @KilnRoom and the corresponding lowercase handle can match. This is text normalization, not confirmation that a profile is currently available or that its owner has remained the same person over time.

Do not use display names as substitutes for handles. Two suppliers can share the visible name Clay Studio, while a single supplier can change that name without changing its handle. Email addresses, biography text, post links, and captions do not establish the profile identity needed for this comparison. If a copied row contains several kinds of information, prepare a clean handle or supported profile URL before adding it.

In a small preparation exercise, imagine 18 candidate entries: 14 distinct valid handles, three additional occurrences of those handles, and one invalid entry. The meaningful source size is 14, with three duplicates and one invalid entry reported separately. It is not an audience of 18 accounts. Duplicates can arise across numbered export parts or when someone combines a pasted list with files containing the same records.

Review both sources independently before interpreting any mismatch. If invalid entries exist, correct their preparation or explain their exclusion in your notes. An empty second box is not proof that the studio has zero followers. A genuinely empty source must be explicitly marked as intentionally empty; an unfinished import should remain unfinished. This distinction prevents a missing file from producing a dramatic but meaningless list of apparent losses.

Work through a current supplier relationship review

The fictional ceramics studio records 144 unique accounts in Following and 210 in Followers from its current pair. Exactly 96 handles occur in both sources. Relationships mode therefore returns 48 accounts Not in follower list, 114 Followers not in following list, and 96 In both lists. The first check is arithmetic: 48 plus 96 equals 144, and 114 plus 96 equals 210.

Across the two lists there are 258 distinct handles: 48 plus 114 plus 96. Adding the source totals directly would give 354 and count the shared handles twice. The chart uses the three disjoint result buckets, not two overlapping source totals. This matters when writing a summary for the studio owner: the chart describes the supplied relationship records, not a larger pool of verified people reached by the account.

If the studio calculates reciprocity among accounts it follows, the denominator is 144. Dividing 96 by 144 gives about 66.67%. Dividing the same overlap by 210 gives about 45.71%, which instead describes the share of recorded followers that also appears in Following. These percentages answer different questions and are manual interpretations of the counts, not evidence that either audience segment is more valuable.

Review the 48 mismatches according to the studio’s purpose. A materials supplier, local gallery, or specialist educator can be worth following without appearing in Followers. The 114 accounts in the opposite bucket might include customers the studio has no reason to follow. No bucket is an automatic cleanup instruction. Keep any account decision separate from this local checker, which cannot follow, unfollow, message, or assess the relationship’s quality.

Measure exhibition-period changes with two snapshots

For the exhibition review, the studio saves an earlier follower record with 210 unique handles and a later record with 224. There are 196 handles present in both. Snapshots mode returns 14 Absent from later list and 28 Added to later list. The totals reconcile: 196 plus 14 equals the earlier 210, while 196 plus 28 equals the later 224. These are fictional records for illustrating the method.

The net change is positive 14 because 28 additions minus 14 absences equals 14. Looking only at the source totals would reveal that net increase but hide the larger movement underneath it. The union contains 238 distinct handles across both records. This is not the number following at either date; it is the number represented at least once in the two snapshots you supplied.

A manually calculated recorded-list retention figure is 196 divided by 210, or about 93.33%. Net recorded-list growth is 14 divided by 210, or about 6.67%. Use the earlier list as the denominator when describing change from that baseline. Neither calculation proves that the exhibition caused the additions. There may have been other posts, offline referrals, different collection coverage, or account changes during the same period.

Enter ordered snapshot dates when they are known, and keep the original date notes alongside the files. A handle missing later might reflect an unfollow, a renamed account, unavailable data, or another change the files cannot explain. The result supplies no exact event time. Use the snapshot comparison guide when designing a repeatable schedule, and preserve uncertainty in the wording of your exhibition report.

Investigate suspicious differences before reporting them

If the missing bucket seems unexpectedly large, first count the follower parts you selected. Next check that both sources belong to the studio, not a personal profile managed from the same login. Confirm that the files describe followers rather than requests, blocked accounts, or another connection category. A large result is not itself a parsing error, but these preparation checks often explain why a technically valid comparison is misleading.

If counts differ from the profile display, do not edit the files to force agreement. The display and export may represent different moments or coverage, and your prepared source may contain rejected entries. Record the discrepancy and inspect the cleanup counts. If confidence remains low, obtain a fresh comparable record and rerun the check. A smaller trustworthy result is more useful than a larger number created by mixing sources.

Keep within the import limits: 5 MB per file, 10 MB per source, and 100,000 candidate rows per source. Handles are limited to 100 characters and profile URLs to 500. Multiple supported files can form one source within those limits. Do not cut a large list arbitrarily and call the remainder complete. A deliberately limited cohort needs matching scope on both sides and an explanation in the report.

After correcting a source, review the newly calculated result before copying or downloading it. Changes and rejected imports must not be treated as permission to reuse an older result. Also check direction: swapping earlier and later reverses the meaning of additions and absences. Switching to relationships mode changes the meaning again. Read the bucket labels every time instead of assuming that the first colored bar always represents historical losses.

Save a useful report and a reproducible baseline

Select a result bucket and search for a particular handle when reviewing a question from the studio owner. Results are sorted and displayed in pages of 50 rows. Search and pagination help inspect the output; they do not reduce the full comparison counts. Do not report only the first visible page as the complete result. Check which bucket is selected before copying its list for a private review worksheet.

Use the CSV download when you need the full categorized validated results, including identifiers. Use the PDF for an aggregate handoff: it includes the chart, source and cleanup counts, comparison context, optional dates, and practical next steps, but not raw identifiers. The PDF is therefore suitable for the exhibition summary, while a restricted CSV can support the owner’s closer review. Treat copied lists and CSV files as separate identity-bearing records.

The checker does not persist your comparison as an ongoing monitoring account. Save the relevant original files and your preparation notes locally if you want a future baseline. Record the profile, relationship type, scope, and date meaning in the folder name or accompanying notes. Avoid keeping unrelated account-export material in the working folder just because it arrived in the same package. Your next comparison should start from identifiable evidence, not memory.

Translate the result into a concrete planning question, such as whether the exhibition brought a larger recorded audience while the studio’s profile still explains its products clearly. The Social Media Profile Audit can structure that separate review from information you provide. SMTasker automation is a different workflow; this free checker does not connect its output to SMTasker, import results automatically, or turn a list difference into bulk actions.

Questions about Instagram Unfollowers Checker

Continue with another list guide

Return to the Follower & Connection List Checker to compare your prepared data, explore a related guide below or browse all free tools.