Use LinkedIn saved posts to prepare for a better sales call

Why recent LinkedIn posts can beat the About page

An About page is designed to stay broad and polished. It tells you how the company wants to be described. Recent posts can show what its people are choosing to discuss now: a product launch, implementation problem, hiring push, operating metric, customer objection, or change in strategy.

One post is not a reliable signal. A repeated theme across the CEO, a department leader, your contact, and people doing the work is more useful. The purpose of this workflow is to find those repetitions and turn them into questions—not to pretend public activity reveals private budget or buying intent.

Save posts from four perspectives

Open LinkedIn profiles and save two or three recent, relevant posts from each of these groups:

• The CEO or founder: company-level priorities, positioning, market changes, and major bets.
• The relevant department leader: functional goals, trade-offs, metrics, and initiatives connected to the problem you solve.
• The person you are meeting: the language they personally use, the work they highlight, and questions they appear to care about.
• Employees close to the work: implementation details, recurring friction, customer feedback, and operational reality.

Aim for roughly 8–15 posts published within the last three to six months. Skip generic motivational posts, hiring announcements unrelated to your offer, and reshares with no added commentary. You want a compact body of evidence, not a dump of everything anyone at the company has posted.

Export and isolate the posts you have bookmarked

Open Saved Items → Saved posts and articles and use LinkedIn Bookmark Exporter to download the full text, authors, dates, and source URLs.

1. Choose CSV if your saved-post history contains lots of unrelated material. Filter the file by the target authors, then copy the matching rows into a new company-specific sheet.
2. Choose Markdown if you have a small, clean set and want one readable document for Claude.
3. Name the working file clearly—for example, acme-pre-call-research-2026-08-28.md.
4. Remove unrelated saved posts and any private notes that Claude does not need.
5. Keep the author, date, full post text, and original URL attached to every record.

The filtering step matters. LinkedIn Bookmark Exporter can capture your saved-post history, but Claude should receive only the records relevant to this company and call. A focused file produces a better answer and avoids uploading the rest of your professional reading history unnecessarily.

Ask Claude for evidence, not personalization

Attach the focused file to Claude and use this prompt:

I am preparing for a sales call with [company] about [the problem our product solves]. Analyze only the LinkedIn posts in the attached file.

Answer: “What is this company repeatedly telling the market it cares about?”

Separate the findings into three buckets:

1. Directly stated priorities—explicit initiatives, goals, investments, or problems. Cite the author, date, and source URL.
2. Problems strongly implied by repeated discussion—patterns supported by more than one post or perspective. Explain the inference and cite the supporting posts.
3. Speculation and unknowns—things the posts do not establish, including budget, urgency, ownership, vendor status, and purchase intent.

Then list the company’s exact recurring terms for the problem and outcome. Do not invent facts, treat engagement as evidence, or infer sensitive personal information.

Read the citations, not just the summary. Open the source links for the two or three findings you might use on the call. If Claude combines unrelated ideas or overstates a weak pattern, move that finding into the speculation bucket.

Turn the findings into five discovery questions

Use a second prompt after you have checked the evidence:

Using only the verified findings above, write five open-ended discovery questions for my call.

Each question should:
• use the company’s natural terminology;
• test one hypothesis without assuming it is true;
• ask about a process, constraint, trade-off, or desired outcome;
• sound natural when spoken aloud;
• avoid mentioning that I analyzed employees’ LinkedIn activity.

For each question, show the evidence behind it and what I should listen for in the answer.

The “what to listen for” field is important. It turns a clever-sounding question into an actual discovery tool. Listen for ownership, current process, consequence, frequency, workaround, and what happens if nothing changes.

Worked example: from three posts to one useful question

Imagine the exported file contains these signals:

• The CEO says the company wants enterprise customers to reach value faster.
• The customer-success leader writes about making implementations more consistent.
• The person you are meeting describes handoffs between sales, solutions, and onboarding.

Directly stated priority: faster time to value and more consistent implementation.

Reasonable inference: the implementation process may contain delays or variation across handoffs.

Still unknown: where the delay occurs, whether leadership considers it urgent, who owns it, and whether software is part of the answer.

A natural discovery question would be:

“You have talked recently about helping enterprise customers reach value faster. Where does implementation tend to lose the most time today—handoffs, approvals, technical setup, or somewhere else?”

That opening demonstrates preparation without pretending you already know the diagnosis. It also gives the buyer room to reject the premise.

Keep it useful—not creepy or overconfident

This research can improve relevance, but it cannot prove budget, urgency, purchase intent, the current vendor, or the real root cause. Those belong in discovery. If the posts do not support a meaningful hypothesis, use a straightforward call plan rather than forcing a personalized angle.

For a reusable source-grounded prompt and verification process, see the LinkedIn saved-post AI knowledge-base guide. If you prefer filtering the exported evidence in a spreadsheet first, follow the Excel and Google Sheets workflow.