A LinkedIn analytics tool that measures the deal, not the impression
LinkedIn already tells you how many people saw a post. It cannot tell you which post started the conversation that became a client. PublishFlow tracks that, and then writes more of what worked.
11,240 impressions this month
You have looked at the analytics tab. Eleven thousand impressions on the post you dashed off, four hundred on the one you spent an hour on. You stared at it, learned nothing you could act on, and went back to guessing. The number went up. Nothing happened.
Impressions are a scoreboard, not a signal
Every analytics tool in this category counts the same four things, because those are the four things the platform reports. None of them can see the DM that arrived on Tuesday, or the call it turned into, or whether the call closed.
Those are the only numbers that decided whether the posting was worth it. So PublishFlow asks you for them, keeps them next to the post that caused them, and then does the thing a dashboard never does: it changes what gets written next week.
How it works
It asks you for the numbers
Published posts show up waiting for their result, so logging is prompted rather than remembered. About fifteen seconds each.
“Your churn problem probably isn’t a churn problem”
Log the things a dashboard cannot see
A DM. A call booked. A deal. A mention. Plus the weekly signals that move before revenue does.
What worked comes back in next week’s drafts
Not a chart you have to interpret. A change in the output. The writer is told which of your formats earns its place.
Next Monday’s drafts lean toward story, because yours land.
The bar for calling something a pattern
It will not tell you a format works off one good post. A format needs at least two measured posts before it counts at all, and the best has to beat the weakest by two times the median before the writer is told anything. Below that bar it stays quiet, because a confident conclusion from four data points is worse than no conclusion.
Story: 4,200 median across 3 posts. Proof: 1,600 across 2. That is a 2.6x gap, so the writer hears about it.
Story: 2,100 across 2. Proof: 1,800 across 2. Too close to mean anything, so it says nothing and keeps counting.
What it will not do
The limits, stated plainly, so you can decide before you sign up.
It will not pull your numbers from LinkedIn yet.
You enter them, which takes about fifteen seconds a post. Automatic collection is on the list.
It will not tell you a format works off one post.
Two measured posts minimum, and a 2x gap, before it says anything at all.
It will not chart your follower count over time.
LinkedIn does that already, for free.
It will not attribute a deal for you.
You decide which win came from which post. It keeps the record and does the maths.
Works with
Post generator
Three posts from one thing that actually happened.
A shaped week
Five formats across five days, so there's something to compare.
The Monday digest
Monday morning, the week's posts are already drafted.
Library
Everything you've written, in one place.
Questions
Does it connect to LinkedIn analytics automatically?+
Not yet. You enter the numbers yourself, which takes about fifteen seconds a post. Automatic collection is something we want to add.
Then why not just use LinkedIn's own analytics?+
Use them. They are free and they are the source of the impression count. This tracks the part LinkedIn cannot see: which post produced the DM, the call, or the deal, and what to write more of because of it.
How long before it tells me anything useful?+
It needs at least two measured posts in a format, and one format has to clearly outperform another, before it changes anything. Realistically a few weeks of posting.
What counts as a win?+
A DM worth having, a call booked, a deal, or a mention. Anything you would have been pleased about on the day.
Does logging this change the posts I get?+
Yes. That is the point. Your best-performing formats are fed into the writer, so the drafts lean toward what has actually worked for you.
Is this a replacement for Shield?+
No. Shield collected your numbers for you. This asks for them, and then does something with them that Shield did not.