Not subscribed? Sign up to get it in your inbox every week.

A quick note before we start. I'm trying something a little different this week. Most weeks I send you a short riff from inside the work; this one is a deep dive, a full feature on one tool and one company, because I want to show what it actually looks like when software and AI transform a business at the root, all the way down to how the work is organized. If it lands, I'll do more of them.

A few years ago at Copy.ai, I tried to get the whole company to post for a big launch.

I wanted every employee sharing it on the same day. So I did what everyone does. I dropped a message in Slack asking people to reshare, pasted the copy into a thread, and hoped. A few folks did it that morning, most saw it after lunch, and a handful never saw it at all. Our content calendar lived in a spreadsheet that two people opened. The company account and the exec accounts ran on a shared login and a password saved in a doc. When the launch was over, I could not tell you whether any of it had moved a single deal, because there was no way to connect a comment on LinkedIn to a name in our pipeline.

The energy was there. Forty people wanted to help, and most of them had a real audience. What we did not have was a way to point all of that in one direction, on one day, at one goal.

What follows is a deep dive on Ordinal, the tool one company used to turn exactly this mess into a 15x, and a preview of the move every other company is about to make with AI.

The idea underneath it is already consensus: buyers trust people more than they trust brand pages, so a company's employees are its most credible channel.

Here is Alex Lieberman making the case:

Think about what running that motion by hand actually looks like. The profiles are scattered across the company, each with its own login and its own posting habits, and someone keeps a shared password in a doc so the company account can go out at all. The social lead lives in a Slack channel begging for reshares and quietly noting who ignored them, while the founder maybe gets a ghostwriting shop and everyone else gets a prewritten corporate post rather than the perspectives of forty real people. None of it stays on message, none of it runs on a shared calendar, and none of it can tell you whether a post ever became pipeline.

A company already owns the distribution: it is the sum of every employee's following, and most of it sits there unused. What it lacks is the one layer that points that distribution at a single goal and proves it landed.

The company that got this right is Clay, the GTM data company, and the person who built it is Peter Kang, their head of storytelling. Ordinal is the command center he runs the whole motion from: one operator driving the CEO's profile, the exec team's profiles, and every employee who opts in, all as one coordinated channel. Think of it as a social media manager wired into every profile at once, plus the calendar, the approvals, and the analytics that hold it together.

When Peter started, he was a one-person social team stitched across five or more disconnected tools. He dealt with "the nightmare that is manual LinkedIn posting," juggling a handful of tools and doing rote, repetitive work just to get a post out the door. There was no shared calendar, no clean approval flow, and no way to run more than a couple of accounts without dropping one.

Then he put it all on one system. A single content calendar now covers the company account, the exec accounts, and partner accounts. Approvals run through Slack, engagement is handled for him, and contributors get scoped access instead of a shared password. On that base, Peter grew a single account into a 25-person employee advocacy program that he runs largely on his own.

The numbers are the reason this is worth your time. Clay's main LinkedIn account went from 8,000 followers to 120,000 in a year, a 15x jump. Peter published 1,080 posts across the accounts in that same stretch. The system took 10 to 20 hours of manual work off his plate every week and saved the company the one or two hires it would otherwise have needed to keep up.

Read those numbers carefully. What they describe is coordination. One operator pointed 25 people at one goal, the software took the friction off each profile, and the command center supplied the direction.

The piece that turns all of this from activity into a goal is measurement. Posting a lot is easy to feel good about and hard to defend. So Clay closed the loop. They combined Ordinal's analytics with their own internal tooling to see which specific profiles engaged with the content, then followed those people into the pipeline and matched them to deals that eventually closed. Engagement stopped being a vanity chart and became attribution.

That is the move that makes a people-channel accountable the way paid and outbound already are. A comment on a founder's post becomes a named account you can route, follow up on, and eventually book as revenue. Once you can measure it in pipeline, the channel earns a real budget and a real owner, and stops being the thing marketing does on the side.

Step back from the LinkedIn numbers for a second, because the real story is what Clay changed underneath them. The 15x is the output of a reorganization: 25 people, one calendar, one goal, one measurement, run by one operator. Buying the software was the easy part.

We have watched this exact pattern before, and the cleanest example is almost a century old. When factories first electrified, most of them simply pulled out the big steam engine and dropped one big electric motor into the same spot. They kept the same layout, the same overhead shafts and belts running the length of the floor, the same everything. Productivity barely moved for about two decades. The real gains showed up in the 1920s, when a new generation of factories reorganized around the motor itself: every machine got its own, and the floor was laid out around the flow of work instead of the location of the power source. Same technology, twenty years apart. The value arrived only once the work was rebuilt around it. (The economist Paul David told this story in 1990, and it has anchored how people think about technology shifts ever since.)

Social is running the same play. For years the constraint was that good content could only come from a few people, so it flowed through the center: the marketing team wrote it, the brand handle pushed it, and everyone else stayed passive, because creating and editing and posting on top of a full-time job was more friction than anyone could sustain.

AI is what removes that friction. Once drafting, scheduling, and engagement take minutes per person instead of hours, creating stops being the bottleneck, and the work flips: the forty people with the real relationships become the production line, each posting in their own voice, while one operator and one system handle the coordination and measurement that used to sit at the center. The motor moved to every desk. Clay rebuilt the floor around it.

Almost no one else works this way yet. Most companies are still dropping one big motor into the old layout: a chatbot bolted onto support, a copilot seat handed to every engineer, an assistant sitting next to a workflow that otherwise runs the way it always has. It produces demos, a little time saved, and a slide for the board. The 15x never shows up, because the org chart underneath it never changed.

The companies pulling ahead are asking a harder question: given what this technology can now do, how should the work itself be arranged, who should do what, and which goals are within reach now that were not before.

If you have read this newsletter for a while, this is the same shape I keep circling. The winning move is to build one place that coordinates the many, whether that is Ramp's internal skill library or Clay's social command center. Social is just the version you can see from the outside.

So here is the test I would run this week. Add up the LinkedIn following of everyone at your company, founder first, then the exec team, then everyone else. That number is distribution you already own and almost certainly are not steering. Name the one goal you would point it at, and the pipeline number that would prove it moved.

And if you want to run the test on the same system Clay uses, I talked to the Ordinal team while writing this and they agreed to a discount for Bottleneck readers.

Somewhere on your calendar there is another launch day coming, and forty people who would help if someone pointed them at the same thing. This time you can.