← All articles

Measurement

A social media scorecard for solo creators: what to track and what to stop

A one-page monthly report for creators that keeps exposure, interaction, site results and effort in separate columns, with metric definitions, a synthetic example and a decision log.

By Ferryman··6 min read

Keep four kinds of number apart

Most creator reports are a list of whatever the apps show. That tells you what happened on each app, but it rarely tells you what to do next. The columns below each answer a different question, and a platform can do well on one and badly on another. That disagreement is the most useful thing on the page.

GroupQuestion it answersTypical metricsWhere it comes from
ExposureDid the post get shown?Views, playsEach platform's own analytics
InteractionDid anyone respond?Replies, saves, shares, likesEach platform's own analytics
Site resultsDid anyone act off the platform?Sessions from tagged links, signups, salesYour website analytics or store
EffortWhat did it cost you?Hours spent adapting and postingYour own log

Why you shouldn't combine them into one score

A blended score needs weights, such as one reply being worth ten likes. Those weights are invented, and the score moves whenever you change them. Worse, it can rise while the thing you care about falls: a jump in views can cover a drop in signups, and the single number shows only that things improved.

Keep the columns separate and compare them by eye. If you want one comparison across platforms, use a ratio with a real unit, such as hours per signup. You can say exactly what it means and recompute it from the row.

Write the definitions on the report

Each platform defines its own numbers, so copy the definition into a notes tab and link to where you found it. These are the published descriptions at the time of writing, and platforms do revise them.

SourceWhat its documentation saysWhat to note on the scorecard
Instagram media insightsViews are the total number of times the media was played. Total interactions counts likes, saves, comments and shares on a reel, minus unlikes, unsaves and deleted comments.Data can be delayed up to 48 hours. When data doesn't exist or is unavailable, the API returns an empty set rather than 0, so don't record it as zero.
YouTube audienceMonthly audience is the total unique viewers who watched your content in the last 28 days.It is a 28-day window, so it won't line up exactly with a calendar month.
GA4 campaign linksCampaign values appear in the Traffic acquisition report, and parameter values are case sensitive.Use lowercase everywhere. Meta and meta are reported as two different values.

Synthetic example: one month, four platforms

These numbers are invented to show the method; they are not customer data or a benchmark. A solo creator publishes to four platforms for one month. The goal for the month is newsletter signups, and every link in every post is tagged by platform.

Platform (synthetic)Views reportedInteractionsTagged sessionsSignupsHoursHours per signup
Instagram9,60031042540.8
TikTok21,40052018155.0
X6,20014064630.5
Fourth networkUnavailable9522310.3
Total37,200 reported, three platforms1,06514615130.9

What the example tells you, and what it can't

TikTok has the most views and interactions and the fewest signups, at five hours for one signup. A views-only report would call it the winner. A signups-only report would call it a failure. Neither is safe yet: one signup is too few to judge, and TikTok may be introducing people who later sign up through another platform, which last-click tagged links won't show.

The fourth network shows no view count, and its row is still worth reading: three signups for one hour of work. Writing "Unavailable" instead of leaving the cell at zero is what stops the views total from quietly implying that nobody saw those posts. The 37,200 total also only counts views as reported by three platforms, so it is not a count of people.

With 15 signups in a month, differences of one or two between platforms can easily be chance. Treat a single month as a reason to run a test and look for the same pattern over two or three months before you stop anything.

End every report with a decision

A scorecard that changes nothing is only a record. Add a short decision log under the table, and limit it to one or two changes a month so you can tell which one had an effect. Each entry needs the evidence, what is still unknown, and a date to check. Continuing the synthetic example:

DecisionEvidenceStill unknownCheck on
Change: add one clear next step to TikTok posts, and ask new subscribers where they found youMost views, 1 signup, 5 hoursWhether TikTok viewers sign up later elsewhereNext month's report
Keep: X at the current pace6 signups, 0.5 hours per signupWhether it holds in a second monthNext month's report
Stop: nothing this monthNo platform has two months of weak resultsNot applicableNext month's report

A 30-minute monthly routine

Filling in the sheet should take less time than the decisions it supports. If it takes longer, remove columns.

  1. Set the goal for the period before you read any numbers, and name the one action that counts.
  2. Use the same date range for every platform, and note any source that uses a different window.
  3. Copy the numbers in, and link each one to the report it came from.
  4. Mark missing cells Unavailable, never zero.
  5. Log your hours per platform, even as a rough estimate labeled as one.
  6. Read across each row, write down the disagreements, and make one or two decisions with a check date.

Where Ferryman fits

Ferryman's analytics can supply the exposure and interaction columns for posts published through Ferryman: supported views, likes, replies and reposts by platform, over 7, 30 or 90 days or lifetime, with links to the posts behind the totals. It requires a paid plan, and metrics vary by what each platform reports.

It doesn't fill the other two columns. Sessions and signups come from your own website analytics and tagged links, and the hours come from your own log. Posts you published outside Ferryman aren't included, so check the app's native analytics for those.

FROM GUIDE TO WORKFLOW

Explore this workflow in Ferryman

Sources & further reading

Written by Ferryman. Sources checked Nov 9, 2026. Platform rules can change; consult the linked documentation for current requirements.