← all playbooks

Get two or three talking points per account from what they post on LinkedIn

For each account, the two or three things from the company's recent LinkedIn posts and its leaders' posts that are worth raising in a conversation, each with the post link, the date and why it matters to your desk. Accounts that posted nothing useful are listed as such. Nothing is sent.

use when
you have a named account list and want to know what each company and its leaders have been saying on LinkedIn before you speak to them
starts from
A list

The prompt

Paste it into Claude Code or the Claude desktop app with Hyreflow connected. The first line tells your agent to use Hyreflow, so it reads the play, asks before it spends anything, and hands the work back to you.

paste this into Claude
Use hyreflow and tell me what my target accounts and their leaders have been
posting on LinkedIn lately, and what is worth raising with them.

The accounts are in <ACCOUNT_LIST_FILE_OR_PASTED_NAMES>, with LinkedIn
company pages where I have them. I recruit <FUNCTION> in <REGION>.

- Confirm each LinkedIn page and profile before you fetch it. Do not guess
  a handle, ask me.
- Leaders to read: <NAMED_LEADERS, OR THE TITLES TO LOOK FOR>. At most
  <LEADERS_PER_ACCOUNT> per account.
- Only posts from the last <DAYS> days, and only their own words. Leave
  out reshares of other people's posts.
- Give me two or three points per account that are worth raising, each
  with the post link, the date and one line on why it matters to my desk.
  If an account has posted nothing useful, say so.
- Do not write messages and do not quote anyone's post back at them.

Run three accounts first and tell me what the full list costs before you
do the rest. Nothing gets sent.

Replace every <PLACEHOLDER> with your own detail. Everything else can stay as written.

What you need first

  • A list of named accounts, ideally with each company's LinkedIn page
  • The leaders you want read at each account, or the titles to look for
  • A Hyreflow workspace with credits

Tools it can reach for

The agent picks per step from what your workspace has. Nothing here is required by name.

What happens when you run it

Free steps are marked free. Anything that spends credits is marked, and the agent asks before the first paid run of any size.

  1. 1

    Confirm the LinkedIn page for each account

    credits

    A post feed is fetched by the handle in the LinkedIn address, not by the company name. Where your list has no page, a web search finds it and you confirm it, because two similar handles are two different accounts.

  2. 2

    Settle which leaders to read

    credits

    You name them, or a company-scoped search finds the head of the function you recruit for and returns their LinkedIn profile. The number per account is capped before anything is fetched, because every feed is a separate charge.

  3. 3

    Read the company's recent posts

    credits

    One request returns the most recent page of the company's posts, newest first, each with its text, date, link and reaction and comment counts. It is charged per request, whatever the page carries, including nothing.

  4. 4

    Read each leader's recent posts

    credits

    The same read, per person. A reshare appears under the original author, so the author of every post is checked against the person asked for before anything is attributed to them.

  5. 5

    Drop what is not theirs or not recent

    free

    Reshares, posts outside your window and stock posts such as awards and anniversaries are removed. What is left is what the account actually said, lately, in its own words.

  6. 6

    Reduce each account to two or three points

    credits

    The remaining posts are read against your desk: a team being built, a site opening, a product launch, a problem a leader described. Each point keeps its post link and date, and an account that supports only one point gets one.

  7. 7

    Review the points beside the posts

    free

    You get the points per account with the source posts next to them. No message is drafted, nothing is loaded into a sequencer, and nobody is contacted.

What a run costs

Credits are spent per company the play actually works, and a lookup that finds nothing usually costs nothing. The two figures are the run where the first provider answers and the run where every lookup walks its full chain.

companiesif the first provider answersif every lookup walks the chain
25$220 credits$220 credits
100$880 credits$880 credits
500$40400 credits$40400 credits
1,000$80800 credits$80800 credits

Free before anything is charged

  • Drop what is not theirs or not recent
  • Review the points beside the posts

What moves the number

  • Coverage on people search. The chain stops at the first provider that answers, and only that provider bills.
  • How many companies survive the free filters. Everything dropped before the paid steps costs nothing.
  • The scoring and drafting steps run on the metered agent, charged on what they read and write rather than per company, so they sit outside this table.
  • Providers you connect with your own key. Those calls bill your account, not your credits.

An estimate, not a quote, priced at the volume credit rate. Your agent sizes the run against your own workspace and tells you what it will cost before it spends anything.

Most of a feed is noise, and you need the two things that are not

A target account says a good deal about itself in public. The company page announces the site it opened and the product it shipped. The head of engineering writes about the migration that is eating the team's year. The finance director mentions closing the books two people short. Any of these changes what you would say on a call.

Nobody reads it. Forty accounts with three leaders each is a lot of feeds, and most of every feed is awards, anniversaries and reshared articles. So the call opens with "how are things", and the one post that would have made it specific goes unused.

This play reads the feeds for a list you have already chosen and hands you what is left once the noise is gone. It is preparation, not copy. For a first line per person on a campaign list, see write one sourced opening line per person.

What you get back

One short block per account: two or three points worth raising, each with the post it came from, the date, who posted it, and a line on why it matters to the roles you recruit. The posts sit underneath, so you can read the original before you rely on it. Accounts and leaders with nothing useful are listed as read, so silence is a result and not a gap. No messages are drafted and nothing is sent.

Variations worth knowing

Company pages only. Skip the leaders and read the company feeds. It is the smallest run, and a reasonable first pass over a long list.

One account, in depth. Before an important meeting, a feed can be walked further back, a page at a time. Every page is another charge, so say how far back to go.

Alongside the rest of the research. Posts are one input. For funding, news, open roles and who leads the function, see walk into a business development call with a one-page account brief.

Where this goes wrong

A guessed handle. The handle for acme and the handle for a similarly named company are different accounts, and the wrong one is charged like the right one. Confirm the page when the agent asks.

A reshare read as their view. A leader who shares an article did not write it. Attributing someone else's words to them on a call is worse than saying nothing.

Quiet accounts. Many leaders never post at all. A thin feed costs the same as a full one. It is a true answer about that account, and a reason to cap the list.

Marketing copy mistaken for intent. The company page is written by a marketing team, and a post about culture is not a hiring plan. What a leader writes about their own team carries more weight.

Reciting the post. Telling someone what they posted is awkward. Use it to choose the question and let them tell you.

Questions

Why not quote the post back in my message?

Because it reads as surveillance, and as a template. Everyone has received the message that opens with 'loved your post about'. The value of knowing what someone said is that your conversation starts in the right place: you ask about the team they are building because you know they are building one. You do not need to show how you know.

What am I charged for?

Each feed read is one request, charged whatever comes back. A wrong handle is billed like a real answer, and so is a page or a person that has never posted. That is why pages and profiles are confirmed first and the number of leaders per account is capped. Finding a leader you did not name is a paid search, and the reduction to talking points is metered on usage. Three accounts run first and the rest is quoted.

What can it not see?

Comments and reactions come back as counts, so you cannot see who commented or what they said. Only public posts are read. A person who never posts returns an empty feed, and a reshare belongs to its original author, not to the person who shared it. It is a public posting history and not a record of everything someone does on LinkedIn.

How is this different from writing a first line per person?

That play produces copy for a sequence, one sentence per person, from several kinds of signal. This one produces preparation for a conversation, per account, from posts alone, and writes no copy. If you are about to ring ten accounts, use this. If you are about to load two hundred people into a campaign, use that.

Can I run this every week?

Yes, once you have run it manually and trust the shape of the output. It converts naturally into a scheduled workflow that re-reads the same pages and profiles and leaves the points waiting in your workspace. Prove it by hand first, because a scheduled run spends within the limits you set without asking each time, and every feed is charged on every run. A scheduled run never sends anything.