Finding LinkedIn posts by keyword is the cheapest buying signal available to a B2B team, and almost nobody does it systematically. Someone writing a post that says "we're finally replacing our old CRM" has told you more in one sentence than a $40,000 intent-data contract will tell you in a quarter.
Short answer: use LinkedIn's Content filter with tight, buyer-language keywords, check it daily rather than weekly, and message within 24 hours while the post is still live in the author's head.
Key Takeaways
| Keyword type | Example | Intent |
|---|---|---|
| Switching language | "moving off", "replacing", "migrating from" | Highest |
| Recommendation requests | "anyone recommend", "looking for a tool" | Highest |
| Frustration | "frustrated with", "broke again", "hate that" | High |
| Category mentions | your product category by name | Medium |
| Competitor names | brand names, product names | Medium-high |
Using LinkedIn's Native Search Properly
LinkedIn's post search exists, but it is deliberately shallow. To use it:
- Search your keyword from the main search bar.
- Click Posts in the filter row — this is the step most people miss, since LinkedIn defaults to People.
- Set Date posted to Past 24 hours or Past week. Anything older is dead weight.
- Optionally filter by Posts from → People you follow or 1st connections if you want warmer results only.
Two limits you will hit fast. First, there is no boolean support in post search the way there is in people search — no AND, no OR, no quotes that behave reliably. You search one phrase at a time. Second, results are ranked by engagement and recency, not completeness, so you are seeing a sample rather than everything.
That means manual keyword monitoring works for five keywords, not fifty.
Choosing Keywords That Actually Convert
The mistake is searching for your product category. Buyers rarely name the category — they describe the pain in their own words.
Tier 1 — switching language. "Moving off", "replacing", "migrating from", "finally ditching", "our contract is up". These people have already decided to buy something. There is no earlier moment worth catching.
Tier 2 — recommendation requests. "Anyone recommend", "what are people using for", "looking for a tool that". A public request for vendor suggestions is an open invitation, and responding is not cold outreach at all.
Tier 3 — frustration. Competitor name plus a negative word. "Zapier keeps timing out", "our HRIS is a nightmare". These posts are gold because the poster has already articulated the gap you fill.
Tier 4 — category and competitor names. Broadest volume, lowest average intent. Useful for market awareness and for catching mentions you would otherwise miss.
Write your keywords the way buyers write, not the way your marketing site writes. If your site says "revenue orchestration platform" and your buyers say "we need to stop tracking deals in a spreadsheet", search the spreadsheet phrase.
Filtering the Noise
Most keyword searches return more junk than signal. Three filters, applied in this order, kill about 80 percent of it:
- Is the author a buyer or a seller? Half of any category keyword search returns vendors marketing at each other. If the author's headline includes your category, they are a competitor, not a lead.
- Is it a real post or a repost with no comment? Reposts without commentary carry almost no intent.
- Does the poster fit ICP? Role, company size, geography. Check this last — the post content tells you intent, and intent is what you are sorting on. The same logic applies when you qualify commenters.
The 24-Hour Rule
A LinkedIn post has a working life of roughly 48 hours. The author is checking notifications, replying to comments, and mentally engaged with the topic. After a week they have moved on, and "saw your post about X" starts to sound like surveillance rather than attentiveness.
Practically: check your keyword list once daily, act same-day on Tier 1 and Tier 2 hits, and let Tier 3 and Tier 4 accumulate into a weekly review.
What to Send
Comment first, message second. A substantive public reply to the post costs nothing, is visible to everyone else reading the thread, and makes your name familiar before a connection request lands. Then:
Hi [Name] — saw your post about [specific thing they said]. We ran into the same wall at [context]. The thing that actually fixed it for us was [genuine one-line answer]. Happy to share how if useful.
Note what is missing: no product name, no calendar link, no "quick 15 minutes". The post gave you a reason to appear; the reason is not "I sell something". More patterns by trigger type are in our signal-based outreach templates.
Why This Breaks Down Manually
Keyword monitoring is a discipline problem, not a difficulty problem. Any one search takes 30 seconds. Twenty keywords, checked daily, with results read, filtered for vendors, checked against ICP, and drafted individually, is 60 to 90 minutes a day. Teams sustain that for about two weeks.
The failure mode is worse than stopping, because keyword coverage decays silently. You do not notice the posts you did not search for.
Updately runs keyword groups continuously across LinkedIn, X, and Reddit, drops vendor and low-intent noise automatically, enriches the poster into a full profile with ICP fit, and drafts a reply grounded in what the post actually said. The broader category is covered in best signal-based selling tools 2026.
Frequently Asked Questions
Basics
Can I search LinkedIn posts by keyword without Sales Navigator? Yes. Post search with the Content filter and a date range is available on a free account. Sales Navigator improves people search, not post search.
Does LinkedIn support boolean operators in post search? Not reliably. Boolean works in people search. For posts, run one phrase per search or use a monitoring tool that queries continuously.
Execution
How many keywords should I track? Start with 10 to 15: three or four switching phrases, three recommendation phrases, your top competitors by name, and two or three pain phrases in buyer language.
Should I comment publicly or go straight to DM? Comment first when the post has an active thread — it builds recognition and other readers see you. DM directly when the post is a private frustration with few comments.
Risk
Is monitoring keywords and messaging posters against LinkedIn's rules? Reading public posts is normal usage. Restrictions come from volume and bot-like patterns in connection requests and messages — stay near 20 to 25 requests a day on a warmed account, as covered in is LinkedIn automation safe in 2026.
Why do my keyword searches return mostly competitors? Because you are searching category language. Switch to the words your buyers use to describe the problem and the vendor noise mostly disappears.