Reach drop diagnostic

Treat a drop as a question, not a verdict

This diagnostic cannot inspect your account. It has no connection to a platform, no access to analytics, and no way to establish a hidden restriction from a public handle. Instead, your answers produce an ordered list of things to investigate. The ranking is a transparent set of priorities, not a probability estimate. A first-place suggestion means “check this first given what you entered,” not “this is definitely the cause.”

Start by choosing when the drop began, the main format, the account type, whether you received an actual warning, and any recent change. If you are unsure about a notice, say so. Uncertainty is useful input. Guessing No to reach a more reassuring result removes the very question you should investigate before interpreting performance.

Why this is not a shadowban checker

A tool that asks only for a handle has not established what happened inside an account. Frenzlab therefore does not produce a green badge, a restriction score, or a claim that an invisible penalty was detected. A claim about exactly which restriction data a platform exposes would require checking that platform's current developer documentation. [VERIFY: Meta for Developers — Instagram Platform API reference — https://developers.facebook.com/docs/instagram-platform/; TikTok for Developers — API documentation — https://developers.tiktok.com/doc/]

The honest distinction is between evidence you can review and a label you cannot substantiate. An actual warning, a changed reporting period, or a different post format is something you can document. “The algorithm hates this account” is not a useful working measurement. If a platform provides account-status or eligibility information, review its official explanation rather than treating this tool's checklist as a substitute. [VERIFY: Instagram Help Center — Account Status — https://help.instagram.com/]

How the investigation order is calculated

An actual warning receives the highest priority. An uncertain warning answer still puts notice review near the top. A very recent drop raises the priority of comparing matching reporting windows. A reported schedule, topic, or format change raises the priority of reviewing that change. The checklist also asks you to compare the selected format and build a baseline appropriate to your workflow.

These priorities are deterministic. Enter the same answers and you get the same order. No machine-learning model is making a secret judgment about your account. The displayed list intentionally omits numeric confidence scores because the tool has no evidence from which to estimate them. Several explanations may remain possible after the first pass.

Three worked examples

Example 1: Two days compared with a month

Imagine a new post has 800 recorded views after two days and an older post has 4,000 after thirty days. The difference is real arithmetic, but the comparison has mismatched observation windows. Select 1–3 days and begin with the matching-window checklist. Compare what each post had after its first two days if that information is available. If the older post had 700 at that point, the original comparison told a misleading story. This example demonstrates a measurement issue, not a platform distribution rule.

Example 2: A new format and fewer posts

Suppose a creator published eight image posts in one month and three short videos in the next. Total interactions fell from 800 to 450. That is a 43.75-percent decline in the total, but average interactions per post changed from 100 to 150. Neither average alone proves the new format is better; the examples also differ in content and sample size. Select Format as the recent change and separate those questions before concluding that account-wide reach was suppressed.

Example 3: One warning among otherwise similar posts

Imagine six similar posts are available for comparison and an actual account notice appeared before the last two. Select Yes for warnings. The checklist moves notice review first because it is concrete evidence to read. That does not prove the notice caused every change in performance. Write down exactly what it says, which content it references, and the official next step. Then compare the six posts using the same reporting window instead of replacing analysis with the notice alone.

Use a compact comparison sheet

Create a private table containing post date, format, topic, measurement age, reach, impressions, and interactions. Add a column for paid promotion and a short change note. Choose columns you can populate accurately; a blank cell is better than a guessed number. The sheet is a recommended working method, not a required platform export format.

Use the engagement rate calculator to make the denominator explicit. If you compare engagement by followers in one row with engagement by reach in another, the resulting percentages answer different questions. Likewise, a count of interactions is not automatically a count of unique people. Keep the metric labels attached when sharing your findings with a client.

Turn the checklist into a small experiment

Choose one hypothesis you can examine without making sweeping changes. For example, compare several posts of the same format before changing the format again. Record what you changed and what stayed stable. Decide in advance which reporting window you will compare. This is a practical discipline, not a promise that a particular test will establish causality.

Avoid deleting many posts, changing the topic, changing the schedule, and rewriting the bio all at once just to “reset” the account. That bundle of changes makes your own observations harder to interpret. If an actual notice requires action, follow the official instructions for that notice; separate that response from your content experiment.

The most common mistake: calling correlation a diagnosis

A drop after a change can suggest a question, but timing alone does not settle the cause. The diagnostic's wording is deliberately cautious: compare, review, record, investigate. Do not translate those verbs into “confirmed.” If the available records cannot distinguish two explanations, say so in your report rather than adding certainty to make the conclusion sound useful.

What this tool deliberately does not do

It does not scrape a profile, access private analytics, reveal audience identities, guarantee a reach recovery, or recommend engagement automation. Your answers remain in the browser. If account access itself is the issue, use the account recovery builder. For a longer example of the measurement workflow, read the reach-drop diagnostic guide. The useful result is a better next investigation, not an invented verdict.