Social Media Report Metric Definitions

Apply these checks in the Social Media Report Builder before entering manual source observations. Our example is Lantern Arts, a fictional organization operating Dock Theatre, Foundry Gallery, and North Hall. Its reporting problem is not a shortage of numbers: different venue teams have supplied snapshots, period totals, and lifetime post measurements as though they were the same evidence.

Define the observation before choosing the field

For every proposed value, ask what was counted, where it was counted, which account it describes, and when the observation applies. Then ask which source screen or export supplied it. These questions form a working metric definition. A label such as views alone cannot tell the report reader whether a figure concerns a channel, a particular content group, or a lifetime observation.

Lantern Arts receives three files. Dock Theatre supplies a monthly account activity summary. Foundry Gallery sends current lifetime results for its exhibition posts. North Hall sends a screenshot of its follower count. Each file can contain useful evidence, but they answer different questions. The reporting lead should not choose identical dates in the builder and assume that this makes their underlying observations equivalent.

Match the source observation to an allowed field only when its meaning fits. The builder provides fixed platform catalogs, not a custom metric designer. A ticket purchase count does not belong in link clicks, and an average viewing duration does not belong in total watch hours. If the source measurement has no suitable field, explain it separately rather than making the closest sounding label carry a different unit.

Write a short source definition before copying the number. For Dock Theatre, this might identify the account summary, selected September activity, included formats, time zone, and capture date. Keep the wording readable and specific enough for a colleague to repeat the selection. A source note is an explanation of your supplied evidence, not proof that the builder verified the account or retrieved that evidence itself.

Use the account label and scope to prevent avoidable confusion. Dock Theatre Instagram activity and Foundry Gallery exhibition cohort are clearer than simply Instagram one and Instagram two. If a record mixes a necessary ending audience snapshot with activity metrics, explain that distinction in the note. A single record level scope cannot erase the individual meaning of the metrics inside it.

Distinguish missing values, explicit zeros, and signed scores

An empty observation means the figure is not known in this report. The source may not provide it for that account, the team may lack a matching export, or the measurement may not apply to the selected material. Keep it empty. Missingness is information about evidence coverage, not an instruction to assume that no activity occurred. You can explain the reason briefly in the account note.

Zero is a supplied observation that the source reported none of the measured event or quantity. Suppose North Hall’s selected content has an explicit shares count of zero. Enter zero. If its screenshots count was never supplied, leave that field empty instead. Replacing both with zero would make an unavailable observation look like a measured absence and would change which calculations the report can legitimately show.

The distinction applies independently to current and previous values. A known current figure with an unknown previous figure is still worth reporting, but it has no usable period difference. A known previous figure with an unknown current figure cannot establish current performance. The builder does not fill either gap, infer a trend from neighboring fields, or treat the report’s overall completeness as a substitute for a missing observation.

Most catalog fields are nonnegative counts entered as whole numbers. Watch hours can use decimals, while Reddit net vote score is a signed integer. A negative score is a valid observation in its own unit; it is not a missing number or a count of negative people. Do not turn a source upvote ratio into score, or rename score as an exact number of upvotes.

During CSV preparation, preserve this distinction in the cells themselves: an empty value stays empty and a measured zero stays zero. Spreadsheet cleanup that fills all blanks can silently change the report’s meaning. After import, inspect a record with missing data and one with an explicit zero. They should remain visibly different, with comparisons shown only where the required observations are actually known.

Separate ending snapshots, period activity, and content cohorts

An ending snapshot describes a quantity at a point in time. Followers and subscribers use that meaning in this report catalog. If Dock Theatre has 4,200 followers at the current checkpoint and 4,100 at the earlier checkpoint, the difference is one hundred in the observed audience balance. It is not a direct count of new follows, because gains and losses can both occur between snapshots.

Period activity describes observations occurring inside a defined interval, according to the source. For example, a source may supply comments recorded during September on the selected account material. Posts published before September can contribute if the source includes activity on them during that window. Do not restrict the interpretation to newly published posts unless that restriction was part of the source selection.

A content cohort is a selected group of items followed under a stated observation rule. Foundry Gallery might select eight exhibition posts and record their lifetime views as of October sixth. That describes those eight posts at that checkpoint, including any activity outside September. Calling the record September posts does not transform the lifetime observations into views occurring during September. The cohort and the measurement window are separate ideas.

Cohort comparisons need a consistent opportunity to accumulate observations. Eight recent exhibition posts measured shortly after publication are not directly equivalent to eight older posts measured after several months. The builder can label cohort scope and retain your notes, but it does not standardize item age or recalculate lifetime values. Describe the limitation rather than making the chart imply a fair exposure period.

Published items is the number of explicitly selected items, not a measure of all content the organization has ever published. Its meaning follows the documented selection. It also says nothing about whether SMTasker supports publishing to that platform. For Lantern Arts, keep the content inventory decision separate from the activity window and the audience snapshot so all three can be understood without guessing.

Check source windows before comparing reporting periods

Write down the reporting dates and the actual source window. These are not always identical. A screenshot may show a rolling twenty eight day summary while the report title says September. A source may restrict the period available for a particular measurement. Enter the observation only when you can explain that relationship; the title and date fields cannot force an incompatible source to represent a calendar month.

The builder uses inclusive start and end dates, with the previous period ending before the current period starts when comparison is enabled. Inclusive means both boundary dates belong to the interval. A September first to September thirtieth window includes thirty days. Verify the source time zone and filters as well, because a local midnight boundary can differ from the boundary applied in another source view.

Consider North Hall comparing 14,000 views in a thirty one day interval with 12,000 in a twenty eight day interval. The raw difference is +2,000 and the positive baseline gives about +16.67%. That describes the entered totals, not a length adjusted improvement. The longer window provides more time for observations, so the report needs an explicit unequal period warning rather than a confident performance claim.

Unequal windows are allowed, but the builder does not silently normalize values per day. Nor would daily scaling automatically fix changes in event timing, content mix, audience eligibility, or observation definitions. If Lantern Arts needs a fairer review, the reporting lead can collect matched windows or narrow the question. The important step is deciding what comparison is defensible before giving its percentage a business interpretation.

Capture timing can matter even for matched dates. Source figures may update after the activity period, and a lifetime post observation continues to change by design. Record the capture date and repeat the same selection method next time. When the earlier source cannot be reconstructed reliably, leaving its value empty is preferable to combining a new observation with an old screenshot whose scope is no longer clear.

Keep exposure, audience, and response metrics separate

Impressions describe counted displays under the source’s rules, while views describe a source defined viewing observation. Neither label automatically means unique people. YouTube thumbnail impressions, for example, concern eligible thumbnail displays rather than every possible place someone might watch. Preserve the source context and identify whether Shorts or longer content is included instead of assuming that similarly named fields form a universal exposure measure.

Reach or accounts reached describes a source defined audience observation, not an interchangeable substitute for impressions. Unique viewers also require their own source scope. The same person can see several posts, visit several venues’ accounts, or use multiple platforms. Adding those figures does not produce a unique audience for Lantern Arts, even when all the values are individually accurate and share the word people in an informal explanation.

The same caution applies within an account. Three exhibition posts can each reach overlapping audiences, so summing their individual reach does not establish account level reach. Use a suitable account observation when the selected field requires it. If that observation is unavailable, keep the field empty. The builder does not deduplicate people across posts, accounts, venues, or platforms and does not present combined reach or follower totals.

Response fields also retain their own meanings. Likes, reactions, comments, replies, saves, reposts, quotes, bookmarks, and shares describe different observed actions. LinkedIn content clicks are not automatically visits to an external ticket page. Snapchat evidence needs its Public Story or Spotlight context, and a source’s post level new followers figure must not be substituted for the catalog’s ending follower snapshot. Availability varies with the source and account.

Do not create a new engagement total by adding every response field, and do not infer an engagement rate or click through rate from whichever observations happen to be filled. This builder does not calculate those measures. Its useful chart compares one account’s selected focus metric in the same unit across the entered windows. That constraint keeps a watch hours bar from competing visually with a comments or audience bar.

Calculate deltas only with an appropriate baseline

With comparison enabled, an absolute delta is current minus previous when both values are known. If Dock Theatre records 300 saves against 240 previously, the difference is +60 saves. The unit stays saves. Because the previous value is positive, divide sixty by two hundred forty and multiply by one hundred to obtain +25%. This is observed change, not a claim about its cause.

A previous zero supports an absolute difference but not a percentage baseline. Suppose Foundry Gallery has eighteen shares after an earlier observation of zero. The delta is +18 shares. There is no valid percentage result from dividing that difference by zero. Describe the two observations plainly rather than writing infinite growth or substituting a tiny invented denominator to produce an impressive percentage.

A negative prior score is also excluded from this percentage calculation. If Lantern Arts’ Reddit score changes from -8 to 12, the absolute delta is +20 score points, but no percentage comparison should appear. The earlier score is not a positive baseline. Keeping the absolute change lets the report describe movement across zero without making the sign and scale of a percentage tell a confusing story.

A positive prior score can support a percentage even if the current score is negative. Moving from 2 to -6 gives a delta of -8 and a percentage change of -400%. That large negative result follows the stated formula; it does not mean four times as many people disliked something. Score is not a direct people count, and its focus chart needs a zero axis to show negative values honestly.

Comparison disabled means the report should not claim a period delta from the entered pair. Missing current or previous observations likewise prevent that calculation. A percentage is never required just to make a report look complete. Use the builder’s account table to check the known values first, then decide whether the baseline and source scope support the interpretation you intend to write.

Match targets to units and interpret progress literally

A target is an optional positive reference for the selected focus metric, entered in that metric’s unit. If North Hall selects watch hours and enters a target of 20, that means twenty hours, not twenty minutes or a twenty percent improvement. The current observation must also be hours. Converting a source duration correctly happens before entry; the builder does not infer the original unit from the number.

With 12.5 current watch hours and an entered target of 20 hours, progress is 12.5 divided by 20, multiplied by one hundred: 62.5%. That describes the observed current quantity relative to the supplied target. It is not a forecast, a deadline assessment, or a probability. The organization still needs its own reasoning for why twenty hours is an appropriate reference for that particular review.

A known current zero against a positive target gives zero percent observed progress. An unknown current observation gives no progress result. A negative current score does not qualify for this target progress calculation, even though it is valid in the score table. Keep those cases distinct so that an unavailable measurement or a signed observation is not presented as ordinary nonnegative progress toward a count target.

Progress can exceed one hundred percent without becoming a prediction. A current observation of 25 hours against a 20 hour target stands at 125% of that reference. It says the observed quantity is above the entered target, not that the account has become universally successful. The target may be modest, outdated, or concerned with only one part of the organization’s broader objectives.

Changing the focus metric changes what a target would mean. Do not carry a saves target into a watch hours discussion or reinterpret a follower quantity as a views goal. Review the focus choice and its target together before exporting. When a target is not useful, leave it unset rather than entering zero or borrowing an unrelated benchmark to complete a progress display.

Keep a repeatable metric dictionary with the report

Lantern Arts can keep a compact working dictionary in its own collection process: account, field, unit, source label, source selection, activity or cohort meaning, capture date, and known limitations. This is a manual reference, not a separate feature added to the builder. It helps the three venue teams supply consistent evidence without asking the reporting lead to rediscover every definition each month.

Use the builder’s source note for the details that must travel with each account. The field allows three hundred characters, so prioritize the selection, measurement window, capture checkpoint, and any important exception. Put the interpretive consequence in the takeaway rather than cramming a full glossary into the note. For example, explain that Foundry’s lifetime cohort cannot establish September activity, then identify the question it can support.

Before approval, ask a colleague to reconstruct one observation from the definition alone. Can they find the account, choose the same window, recognize the unit, and explain why an empty field is empty? If not, improve the definition before adding more metrics. Save the complete JSON project explicitly when the report is ready to keep; there is no automatic saved history to recover the explanation later.

For the next review, repeat the collection method and inspect changes in source labels or availability before copying the previous template. Keep account operations separate from measurement interpretation. SMTasker’s 5-day trial is available for evaluating supported account workflows, while this free reporting builder remains independent. Neither a trial nor a downloaded report converts manual source observations into connected analytics.

Questions about Report Metric Definitions

Continue with another reporting guide

Build your own review in the Social Media Report Builder, or choose another reporting workflow below. Visit all free tools for other preparation tasks.