A social media report that is assembled by hand every month will eventually not be assembled. The person who builds it gets busy, the numbers arrive late, and the report quietly stops being the thing anyone trusts. The fix is not a prettier template — it is deciding what the report contains, how often it goes out, and which parts of the pull can be automated so the human only does the interpretation.

This guide covers all three, with the data limits stated up front so the report is honest before it is useful.

Decide the audience before the format

Three audiences read social reports, and they want three different things.

The executive summary is for someone who has thirty seconds. One paragraph: what moved, what it means, what you are doing next. No charts. If they want charts, they will ask.

The client update is for someone paying for the work. Output, movement, top posts, exceptions, next actions — in that order. The client wants to know that the work happened, that it is moving in the right direction, and that you are on top of the exceptions.

The team review is for the people doing the work. Post-level detail, format comparisons, what the calendar produced against what was planned. This is the only audience that wants the raw numbers.

One report cannot serve all three. Pick the audience, then build the report for them — and if two audiences need the same information at different resolutions, build the detailed version and cut an executive summary from the top rather than maintaining two documents.

What every report includes, and what to leave out

The structure that survives every stakeholder relationship has five blocks.

Output — what was published: posts per network, per client, per label. This is the block that proves the work happened, and it is the one clients read first when they are unsure whether the retainer is producing anything.

Movement — what changed: reach, engagement, followers, clicks — each with a direction and a comparison window. State the comparison: "versus last month" or "versus the same month last year." A number without a comparison is a number without meaning.

Top posts — the three to five posts that performed, ranked by the metric that matches the goal. For a brand-awareness month, reach. For a conversion month, clicks. For a community month, saves or shares. Say which ranking you used and why.

Exceptions — what went wrong: failed publishes, accounts that needed reconnecting, a post that was pulled, a platform that did not report. This block is the one that builds trust, because it proves the report is not a highlight reel.

Next actions — what you are changing. One to three items, each tied to something in the movement or exceptions block. A report without a next action is a status update; a report with one is a plan.

What to leave out: vanity metrics that answer none of your questions, per-platform engagement rates compared across platforms, and any number that cannot be sourced. The metrics guide covers which numbers earn a place and which are decoration.

The cadence that survives

Three cadences cover most teams, and they serve different purposes.

Weekly — fifteen minutes, internal only. Pipeline check: did everything publish, is every account connected, is the data fresh. Then one sentence of movement and one decision for next week. This is the reading pass described in full elsewhere; it exists to catch problems while they are still small.

Monthly — the stakeholder report. Output, movement, top posts, exceptions, next actions — the five-block structure above, written for the client or the executive. This is the one that gets automated, because it is the one that stops happening when it is manual.

Quarterly — the strategy review. Three months of movement, format-level analysis, what the calendar produced against the plan, and the decisions for the next quarter. This one should not be automated: it is the one place where the interpretation is the entire product.

The pattern: automate the pulls, not the judgement. The weekly and monthly reports can be largely assembled by a script or a scheduled query; the quarterly review requires a person who knows what the numbers mean.

What you can actually automate today

PostHub exposes three analytics endpoints, each accepting a date range in the form from and to as YYYY-MM-DD:

• Summary — the aggregate metrics for the period: reach, engagement, clicks, the totals that go in the movement block.
• Trend — the time series, for the chart that shows the month's shape rather than its total.
• Top posts — the ranked list, for the block that shows what worked.

Each endpoint is org-scoped to your workspace and carries the same authentication as the rest of the API. For a scheduled pull — a script that runs every Monday morning and assembles the month's numbers into a document — you need an API token with the analytics:read scope. Tokens are created per person, expire on a schedule you set, and show their last-used date so you can retire the ones that stopped being needed.

The request quotas are plan-dependent: Starter allows three thousand requests per month with one API key; Professional allows seven thousand five hundred with three keys; Agency allows fifteen thousand with five keys. A weekly pull of three endpoints is well inside any of those budgets; a per-minute poll is not, and would not tell you anything the hourly sync has not already captured.

The honest gap: there is no built-in scheduled report email and no CSV export in the product today. The automation path is the API — a script that pulls on a schedule and writes into whatever document your stakeholder actually reads. It is a small amount of setup, and it is the difference between a report that arrives and a report that gets assembled when someone remembers.

The data limits to state in the report

Every automated report inherits the platform's reporting limits, and the report is more honest when it names them than when it leaves blanks for the reader to interpret.

Absent is not zero. A metric the platform cannot provide arrives as null — a placeholder, not a measurement. Telegram and Google Business Profile have no post-level insights at all, so every metric for those networks is absent by design. A null averaged into a total as if it were zero drags every average down, which is how a healthy month starts looking bad.

Some metrics are gated. X reports impressions only on accounts with the paid metrics tier. LinkedIn engagement arrives, but impressions generally require Marketing Developer Platform access. YouTube reports views and little else. Tumblr's note total mixes likes, reblogs and replies into one figure. The analytics documentation has the full support matrix.

Freshness has a floor. Metrics are swept periodically rather than streamed, and only posts published within a recent window are actively refreshed. A figure that has not moved in two days is a stored value, not a live one — say so in the report rather than presenting it as this morning's number.

Label coverage is your responsibility. If a campaign was not labelled at compose time, the monthly report cannot separate it from the rest of the output. The label filter is the difference between "the launch produced these twelve posts" and "some of these posts were probably the launch."

Three habits cover most of this: state the date range on every chart, note how many posts in the set actually reported, and write "not reported" where a metric is absent rather than leaving a blank.

Where PostHub fits, and where it does not

PostHub gives you the three endpoints with date-range filters, the API tokens with scoped access and visible last-used dates, the org-scoped architecture that keeps a client's numbers inside their workspace, and the label system that makes campaign-level reporting possible when the labels were applied at compose time. The analytics features page and the product overview cover the surface in full.

What it does not do is send the report for you or export it as a file. The pull is automatable; the delivery is yours to build — a script, a shared document, a channel message, whatever the stakeholder actually reads. And the interpretation stays yours entirely: no endpoint can tell you that this month's dip follows a two-post week rather than an algorithm change, or that the campaign label you forgot to apply is why the report cannot separate paid from organic. The numbers are the input; the argument is yours.

Key takeaways

• Pick the audience before the format: executive, client and team want three different reports from the same numbers.
• Five blocks cover every stakeholder report: output, movement, top posts, exceptions, next actions — and the exceptions block is what builds trust.
• Automate the pulls, not the judgement: the monthly report can be assembled by a scheduled query; the quarterly review should not be.
• Three endpoints accept date-range filters; an analytics:read token is all a scheduled pull needs; quotas run 3,000 to 15,000 requests depending on plan.
• State the data limits in the report: absent is not zero, some metrics are gated by the platform, freshness has a floor, and label coverage is your responsibility.

FAQ

Can PostHub email the report to my stakeholder automatically?

Not yet — there is no scheduled report email or CSV export in the product today. The automation path is the API: a token with analytics:read, a script that pulls on a schedule, and a delivery into whatever document your stakeholder reads.

How much data can the analytics endpoints return?

Each endpoint accepts a from and to date range and returns the metrics for that window. The sweep that populates the metrics runs periodically for posts within a recent window, so a very old range may have gaps where the sync did not reach — check the last-captured date before quoting a number from a distant month.

What if a client wants a report format you do not use?

Build the detailed version and cut the executive summary from the top. The five blocks — output, movement, top posts, exceptions, next actions — are the structure; the presentation is yours to adapt. The dashboard reading guide covers which tiles are real and which are still placeholders, which is worth knowing before a client asks why a number is blank.