Instagram Reach Drop Diagnostic

Turn a reach drop into a comparison you can use

An Instagram reach drop is easier to work on when you replace a general impression with a specific comparison. Which posts changed, which metric are you comparing, and how old was each post when you recorded it? A disappointing afternoon and a weaker set of mature posts are different situations. Start with the numbers behind the concern, then decide whether the next job is improving the comparison, reviewing a recorded account notice or testing one content change.

The Instagram Reach Drop Diagnostic compares two manually entered groups of per-post counts. Choose Reels, photos, carousels or Stories and use one matching metric across both groups. The baseline describes an earlier reference; the recent sample describes the posts you want to review. The tool calculates their medians and supporting summaries, then compares the median decline with a threshold you choose. It gives you a structured starting point for a practical review.

This is a browser-only calculator, not a connection to your account. It does not require login, fetch Insights, check hashtag positions, inspect activity logs or diagnose a hidden algorithm change or shadowban. You provide the observations and the account context. That separation matters because the same numerical decline can accompany a different topic, an unusually strong reference sample, younger posts or an explicit notice that needs its own follow-up.

Write a review question before collecting values. For example, ask whether recent product demonstration Reels recorded at the same age are lower than earlier demonstrations made on the same promotion basis. Keep the question narrow enough to answer with the available posts.

Choose comparable posts and one matching metric

Select reach or views according to the metric you can consistently obtain in your own reports. For reach, use the native reach or viewers figure only as defined in that report. Views describe plays or displays and can include repeats; they are not interchangeable with a distinct-account measure. Preserve the original label in your notes. The reach versus views guide helps you choose a comparison without quietly changing what the numbers represent.

Measure every included post at the same chosen age after publication. You might use a point that suits your reporting routine, provided both samples use it. Earlier lifetime totals and recent first-day totals answer different questions. If you lack historical snapshots at the chosen age, mark age alignment as different or unknown.

Keep format, topic and promotion basis comparable as well. A paid product launch, a regular unpaid tutorial and a collaboration with another account do not become equivalent simply because all three are Reels. Either make separate groups or describe the mixed scope honestly. For Stories, compare consistent first frames or sequences at a consistent position and reporting window. Comparing the start of one sequence with the final frame of another introduces a separate difference before any content test begins.

Choose an inclusion rule before looking for favorable numbers. Include consecutive qualifying posts from a defined period, for example, rather than selecting the strongest baseline and weakest recent posts. Record missing observations and why they are missing. The tool accepts different sample sizes, but similar coverage is easier to explain.

Enter the counts and record the comparison context

Enter between three and thirty values in each list. Every value must be a whole count from zero to one billion and must describe one post, not an account-period total. Separate values with spaces, commas or new lines. Do not include thousands separators: enter twelve hundred as 1200. Check the resulting sample size against your source notes, especially when copying a list that originally used punctuation to format large numbers.

Use the optional review label to identify the project or comparison, within sixty characters. The optional observation-age note allows forty characters and can state your measurement point. These labels do not fetch dates or filter the sample. Keep fuller details separately: publication dates, original metric labels, the selection rule, promotion changes and any missing snapshots.

Select whether observation ages are matched, different or unknown and whether scope is comparable, mixed or unknown. Record account visibility as public, private or unknown. The Account Status choice reflects what you personally observed: unchecked, clear, notice or pending. These are context labels, not live checks. If a colleague supplied the figures but nobody inspected status, unchecked is the accurate choice; a normal-looking profile is not a substitute for that observation.

The decline threshold can be set from one to one hundred percent and starts at twenty percent. Choose it as a review trigger that fits your project. It is not an Instagram benchmark or a statistical significance test. Set it before interpreting the current result where possible, and document any later adjustment.

Read the median alongside the rest of the sample

The median is the middle value after sorting a sample. With an even number of observations, it is the average of the central pair. This gives a reference for the middle of the selected posts without letting one very large observation determine the whole result. Median change is recent median minus baseline median. When the baseline median is above zero, percentage change divides that difference by the baseline median and multiplies by one hundred.

Read the mean, minimum, maximum and sample size beside the median. The mean responds more strongly to unusually large values, while the range shows how spread out the observations are. The comparison table keeps the baseline range and highest observation visible so you can notice when a viral reference may be shaping your expectations. A high maximum is not automatically a reason to delete a post from the baseline; first ask whether it belonged under your original inclusion rule.

The below-baseline count reports how many recent values are strictly below the baseline median. A value equal to that median is not counted. This helps distinguish a broad shift across the selected recent posts from a change involving only part of the group. It is descriptive, not a probability, confidence score or prediction for the account.

If the baseline median is zero, percentage change is unavailable because the formula has no usable denominator. Inspect the absolute medians and individual values instead. Comparison readiness is also limited when ages or scope are not confirmed comparable. A comparable sample label means those inputs are aligned, not that the sample has high statistical confidence. Do not add per-post reach values and call the sum unique period reach, because accounts can overlap across posts.

Work through the example from counts to next check

Use the example baseline values 800, 1000 and 1200, with recent values 400, 600 and 800. Each list contains three observations. Sorting does not change their order, and the middle values are 1000 for the baseline and 600 for the recent group. The median difference is minus four hundred. Dividing minus four hundred by the baseline median of one thousand gives a median change of minus forty percent.

The baseline mean is also 1000 and the recent mean is 600 in this deliberately balanced example. Baseline observations range from 800 to 1200; recent observations range from 400 to 800. All three recent values are strictly below the baseline median of 1000, including the recent high of 800. This makes the selected pattern easy to describe: the middle is lower and every recent observation sits below the earlier middle value.

For this run, confirm matched observation ages and comparable scope, record public visibility and a clear user-observed status, and retain the twenty percent threshold. A forty percent decline is at or beyond that chosen trigger. The result identifies a comparison worth reviewing. The clear status observation remains separate; it neither explains the decline nor reverses the arithmetic. Start with differences in the actual posts and their preparation rather than searching for a verdict the sample cannot provide.

If compatible reports also show a baseline non-follower share of seventy percent and a recent share of forty-five percent, enter those optional values. The change is minus twenty-five percentage points. It is not a twenty-five percent reach decline and does not replace the minus forty percent median change. A useful follow-up is to compare how clearly the recent posts introduce their subject to unfamiliar viewers while keeping format, topic, promotion basis and observation age consistent.

Use non-follower share without inventing audience counts

The optional share comparison needs both the baseline and recent non-follower percentages. Each can range from zero to one hundred. Use compatible reports describing the same metric and scope as each other. If one number concerns account-wide views and the other concerns a selected set of post reach observations, leave the fields blank for this comparison. A missing percentage means unknown, while a recorded zero means the compatible report actually showed zero.

Subtract the baseline share from the recent share to find the change in percentage points. This describes a change in audience composition within the supplied report basis. It does not tell you the number of non-followers reached, and the tool does not multiply a median by a share to invent that count. The median of per-post values and a percentage from another aggregation do not automatically share a denominator.

Read share and overall delivery together without blending them into a score. A lower non-follower share can accompany a smaller, similar or larger delivery total, depending on the actual compatible totals. A stable share can also sit beside a substantial drop in the selected median. If discovery is your concern, the non-follower reach guide develops that specific review rather than treating every lower percentage as the same problem.

Audience relevance is a separate question. A broad demonstration may introduce the account to unfamiliar viewers without producing relevant inquiries, while a narrow explanation may serve existing interested people. Use the Audience Quality Estimator for a separately prepared manual account sample when fit matters. Keep actual questions, inquiries and outcomes in their own records.

Follow the result state in a sensible order

When the result says to align evidence first, repair the observation ages or scope before planning a content response. This is often a concrete reporting task: locate compatible snapshots, separate paid and unpaid posts, or narrow the sample to one topic. Label the corrected run separately so the change in readiness remains understandable.

When baseline percentage is unavailable, use the absolute values and build a usable reference rather than assigning a dramatic percentage to a zero denominator. When a comparable decline reaches your threshold, choose a focused next check. A smaller decline still deserves attention if it affects a business decision, but it does not meet the trigger you selected. Stable or higher medians mean this particular comparison does not show a median decline, even if an individual post disappointed you.

Account Status and visibility create separately labeled next checks. Meta has described Instagram Account Status as a way to understand recorded content decisions. Read any actual notice in your account and save its wording; if a review option is offered, use the instructions attached to that decision. The Account Status and reach guide explains how to keep a first-party notice distinct from your performance calculation.

After comparison issues and any observed notice, examine changes you can describe: subject choice, opening clarity, sequence structure, collaboration or promotion. Meta’s ranking explanation describes multiple predictions and personalized recommendations, not one universal switch that a reach calculator can expose. Treat your content explanation as a hypothesis to test. A useful next step names an editable feature and a measurement plan instead of assigning a hidden cause to every low observation.

Choose one change and define the next review

Choose one variable that addresses the clearest observed difference. A maker whose recent demonstrations begin with a long setup could test showing the finished result first. Keep the practical topic, format, approximate length, promotion basis and measurement age comparable. Define exactly what the new opening contains before publishing. Changing the subject, opening, cover, posting pattern and promotion together may produce an interesting post, but it leaves the review unable to separate those changes.

Decide what evidence will support keeping, revising or dropping the change. The primary measure can be the next comparable median, with range and below-baseline count as supporting context. A separately recorded engagement measure or useful inquiry count can protect the actual goal from being reduced to delivery alone. The Engagement Rate Calculator can support a distinct response calculation when its inputs match your question; do not substitute it for reach.

Keep a change journal outside the calculator. Record the hypothesis, changed element, unchanged conditions, publication references, observation age and decision criteria. Use the Content Calendar Generator to organize the intended sequence separately. If an unexpected promotion or account change occurs, note it before interpreting the result. Practical tests are not perfectly controlled experiments, but a clear journal makes them more useful than a memory of trying several things.

Review when your planned comparable observations are available, not after a universal waiting period. A tentative improvement earns a repeat under similar conditions; a mixed result may need a tighter hypothesis rather than more simultaneous changes. The reach recovery plan develops the journal and review criteria. Recovery is a planning goal, not a promised outcome. End each review with one decision and the next observation that could change it.

Save the evidence and keep operational work separate

Copy the aggregate review for a concise handoff, download the CSV for sample rows and summary values, or save the two-page PDF with its chart and next-test notes. Check that the labels, optional shares and threshold match the discussion before exporting. The report footer points to this exact tool. Save your source notes alongside the output so a later reviewer can distinguish the calculations from the decisions behind sample selection.

Inputs are not sent or persisted by this tool, and it does not maintain your review history. Retain the exported report and change journal yourself. When correcting a value, regenerate the report and identify the correction. When changing the sample definition, create a separate comparison rather than presenting it as an improvement in the same series.

SMTasker supports nine platforms through Android phones and a local Windows or macOS client. Its supported Instagram workflows, schedules and activity logs can help you review operational changes separately. Consult the Instagram automation overview for that work. Logs can establish what a supported routine recorded or scheduled; they do not reveal hidden ranking data. This free diagnostic does not import its inputs into SMTasker or apply the test you describe.

Skye uses your own OpenAI or Anthropic key with separate provider billing, and external AI can connect through MCP. The Skye getting-started guide covers that separate setup. Explore a five-day SMTasker trial when your next task involves a supported account routine. Keep the numerical review as your decision record, and use the free tools directory when the next question needs a different calculation.