The monthly report now groups what it counts by the service categories of your workspace, and every brand picks how it leaves the building: off, automatic, or reviewed by a person before it is sent. Per network analytics belong to every workspace; the monthly PDF report belongs to agency workspaces.
Service categories are your rows now, not our list
For a while, the category on a service was a closed list of five values we had written. It held up as long as every agency looked alike, and it stopped holding up the first time somebody sold community management, design, and a video production line that fitted none of the five boxes.
A category is now a row inside your workspace. It has a name you write and a position in the list you decide, and nothing else. A new workspace is seeded with a starting set, and from that moment the list is yours: renaming one touches nobody else's, and deleting one leaves the services that hung off it without a category rather than dragging them into the delete.
One detail is worth understanding before you reorganise anything. Quote, contract and invoice lines freeze the category name they were written with and never look the list up again. A document issued in March still says what it said in March even if you renamed the whole category in September, which is the only way a fiscal document can still mean the same thing a year later.
What changes inside the report
The report uses that list to group. Instead of a run of loose charts, the client receives the month sorted by the categories you sold the work in, which are usually the same ones printed on their invoice. The renewal conversation gets shorter when the report and the proposal speak the same language, and that is the whole idea.
It also changes what a bad month looks like on paper. When the work is grouped by what was sold, a quiet month on one category sits next to a strong one on another instead of averaging into a single flat line, and the client can see which part of the retainer moved.
What does not change is where the numbers come from. They are still the metrics of the six networks that serve analytics, under the same aggregation rules, which the analytics and reports page sets out in detail. A category sorts. It does not invent or recompute anything.
Three modes, and why reviewed is the one you start in
Every brand carries its own mode, inherited from whatever the workspace had set on the day the brand was created, and each brand can be moved without touching the others.
Reviewed is the starting mode for everybody, and that is deliberate. The draft is assembled on the day it is due, frozen exactly as it came out, and left waiting for somebody with permission to send it or discard it. While it waits it already holds that month's slot, so a second run of the job cannot build a duplicate or post a second copy to the client.
Automatic is what you switch on once you have watched three or four months come out clean. And off exists because some brands do not want a report at all, and forcing an empty one on them is worse than sending nothing. What to do with the document once it lands in the client's inbox is the subject of the monthly client report guide.
A gap is not a zero
This is the part that prevents the most arguments. Each network answers to its own vocabulary, and what one of them does not answer is not a zero: it is a gap. Printing a zero where the platform gave no figure claims nobody saw you, and the change indicator beside it then claims nothing moved either.
| Network | Reach | Impressions | Saves | Clicks |
|---|---|---|---|---|
| Yes | Yes | Yes | No | |
| Yes | Yes | No | Yes | |
| TikTok | No | No | No | No |
| YouTube | No | No | No | No |
| Yes | Yes | No | Yes |
LinkedIn deserves a footnote, because the product had this one wrong for months: it does report per post reach, and what it has no notion of is a video view count or a saved post. Saying that a platform withholds a metric is a claim about its API, and that one was false.
Demographics is the other place where a gap is the honest answer. Only Instagram and YouTube serve it, and each in a different unit: Instagram returns accounts per bracket and YouTube returns a percentage of signed in viewers. Those two cannot be added together, so the report shows them apart rather than manufacturing an audience total that does not exist.
Google Business is missing from that table on purpose. It does not measure the same thing as the rest: it measures discovery, how many people found the listing in search or on the map and how many of them then called, asked for directions or opened the website. Putting those appearances in the same column as an Instagram post's impressions would offer a comparison that means nothing, so its numbers are read in a section of their own.
Everything in this article happens in one place in GoFeed.
Try it freeThe rules that turn thirty days into one number
A monthly report is, technically, a decision about how to turn thirty daily values into one. The product documents around thirty metrics with the rule written beside each, and three families explain almost all of it.
- What gets added up. Counts of events: likes, comments, shares, posts. Adding the days is exact.
- What is never added up. Reach, and any figure that counts unique people. Somebody who saw you on Monday and on Tuesday is one person and not two, so thirty days of 1,200 reached is nowhere near 36,000. Only the platform can compute the period total, and when it does not serve one, the report leaves the gap rather than inventing a sum. Why that metric gets misread so often is the subject of the reach and impressions guide.
- What gets recomputed. Rates and averages, starting with the engagement rate, which is derived from the period totals and not averaged from the daily rates. The two numbers sit close enough together that nobody spots the error.
There is a fourth, smaller and more treacherous family: cumulative counters such as the follower count. There the period total is the last day's value, and adding them up means nothing whatsoever.
One calendar warning goes with all of this. Networks do not close the month at midnight on the 30th. Instagram and Facebook figures can take up to 48 hours to settle, Google Business two or three days, and YouTube does not answer for its most recent days until it has closed them. That is why the report prints the date of the data instead of hiding it, and why a draft read on the first of the month can move a couple of figures if you open it again two days later. That is not a broken document: it is the difference between a late number and a revised one.
Who gets what
It is worth saying plainly, because this is the most common confusion. Analytics belong to every workspace, agency or creator alike: the per network panels, the follower series, the best times, and demographics wherever the platform serves them, which is Instagram and YouTube and nowhere else.
The monthly PDF report is a different thing, and it belongs to agency workspaces, because it exists to be sent to a third party who hired you. A creator publishing for their own brands has nobody to send it to, and neither does an in house team. The difference between looking at your own metrics and delivering a document is worked through in the social media reporting article.
What to do in the first month
Three things, in this order. Sort the category list so it looks like what you sell rather than like what a network measures. Leave every brand in reviewed mode and read the first draft end to end, because that is where badly connected profiles and brands with no recipient show up. And when a client asks why a figure does not match what they see inside the Instagram app, show them the aggregation rule printed next to it: that question has a written answer, and answering it well is half the value of sending a report at all.
