Here's the problem: a post you spent 40 minutes on produces zero discovery calls. A throwaway one-liner you fired off between meetings produces three. You have no idea why, and no feedback loop to learn from.
Most fractional operators run their entire content function on superstition. Every Monday, they pick a topic that feels right, write a post that takes between 20 and 90 minutes, ship it, and then never tie the outcome back to the input. The signal is too noisy, the loop is too slow, and they're a one-person team.
Why naive attribution fails
The first instinct is to look at LinkedIn analytics — impressions, reactions, comments. These metrics will mislead you. Impressions tell you about the algorithm, not about your pipeline. The post that gets 50k impressions because a recruiter reshared it is irrelevant to a fractional business model. You don't need impressions. You need the right ICP booking a call.
The second instinct is to ask new leads "where did you find me?" This works ~30% of the time. The other 70% give you "LinkedIn" with no post attribution, because they read seven of your posts over two months and can't remember which one finally moved them.
The loop we'd build, if we were starting over
Three components:
- A scheduling endpoint that captures referrer. Whether you're on Cal.com, Calendly, or rolling your own, expose the booking link with a UTM-style parameter per post. The post links to yourcalendar.com/intro?src=2026-05-12-hedge-tax. Now every booking carries the exact post that produced it.
- A 30-day attribution window. Most discovery calls happen within 30 days of the post that produced them, but it's rarely the same week. A weekly view will mislead you. A 30-day rolling view tells the truth.
- Manual tagging for the calls that arrive without UTM. In the booking confirmation, ask one optional question: "Anything specific that prompted the call?" About 40% of people will name a post.
Run this for one quarter. You will be surprised — and probably annoyed — by what you find. Most fractional operators discover that 80% of their posts produce zero pipeline, and a handful of recurring angles produce almost everything. That's the most valuable information you can have.
What to do with the data
Two moves, in this order.
First, kill the angles that don't pay. Not because they're bad posts — they may be — but because your time is the constraint and you can't afford to keep writing into a void.
Second, do not over-repeat the winners. The intuitive move when you find a winning angle is to do it three more times. The actual right move is to write adjacent posts that hit the same ICP from a slightly different angle. The audience is the same. The novelty keeps the algorithm and the human reader engaged.
The honest answer
Almost no posting tool actually does this attribution loop. We're building one inside Quarz because we couldn't find anything that worked. If you're rolling your own — Cal.com + UTMs + a Notion log — that's about 80% of the benefit. The remaining 20% is mostly the discipline to actually look at the data each month.