> ## Documentation Index
> Fetch the complete documentation index at: https://b2-b-playbook.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# LinkedIn organic
**Last reviewed:** 2026-09-08 · **Reading edit:** 2026-09-08
The founder's post has twelve likes. Most are from people the team already knows.
The marketer's first reaction is to rewrite the opening, try a different format, or move the next post to a supposedly better time. Then a prospect joins a call and says, “Your explanation of the handoff problem sounded exactly like our Monday meeting.”
That does not prove the post created the deal. It does suggest that the feed is doing something the reaction count cannot fully describe.
The opposite happens too. A broad career post attracts thousands of reactions, but almost nobody associates the author with the problem their business solves. The attention is real. Its business usefulness is still a question.
LinkedIn organic is the work of becoming useful and recognizable to relevant people through posts, conversation, and a credible professional presence. It can support discovery, relationships, hiring, customer education, and sales. Those jobs need different content and different expectations.
This guide covers how to choose a voice, find material in real work, write posts people can use, participate without manufacturing engagement, and review the results. It is not a recipe for a particular reach number.
*Reading guide: choose the reader and speaker → collect real material → write a useful post → join the conversation → review what changed.*
If you are starting from an empty calendar, begin with [the material bank](#build-a-material-bank-from-work-you-already-do). If your posts sound generic, use [the before-and-after edit](#turn-a-general-claim-into-a-useful-post). If people respond but the business gets little value, read [the conversation handoff](#move-a-conversation-to-private-only-when-it-helps). The [worked example](#worked-example-illustrative) follows a small team's first four weeks.
## Use this when
Relevant people actually use LinkedIn to learn, compare notes, follow practitioners, or keep up with their industry. You do not need proof that every buyer posts there. You do need something better than “all B2B companies should be on LinkedIn.”
Someone on the team has useful knowledge and is willing to stand behind what is published. That person may be a founder, specialist, customer-facing operator, researcher, or a company-page editor representing documented work.
You can connect a topic to a reader's real situation. “AI” is a category. “How a support lead reviews an AI-drafted reply before sending it” is a situation someone can recognize and discuss.
The team can maintain a modest publishing and response habit. Organic does not mean free: finding material, checking claims, editing, and answering people all take time.
Your expectations allow for both immediate and delayed value. Some readers will ask for help now. Others may simply learn to associate you with a problem they will face later.
## Do not use this when
You need to reach one specific decision-maker by Friday and are hoping they will happen to see your next post. That is a different task; use [LinkedIn outbound](../05-outbound-and-prospecting/linkedin-outbound.md) or another appropriate direct route.
You are planning to scrape everyone who reacts, add them to an automated sequence, or buy coordinated comments. A public interaction is not a request to enter your sales process.
The only material available is confidential, unverifiable, or something the supposed author does not actually believe. Fix the source problem before polishing the writing.
Nobody can review a draft or respond to a genuine question. Reduce the commitment before adding another channel to an already overloaded team.
You also do not need to force a reluctant founder into public performance. A willing specialist, an editorial company page, a different format, or another channel may fit better. The title on the business card is not the content strategy.
## A few useful terms
The **speaker** is the person whose experience and judgment the content represents. An editor can help shape the writing without inventing the speaker's experience.
A **content lane** is a recurring reader problem you can address from several useful angles. It is narrower than an industry and broader than one post.
**Recall** means a reader remembers you in a relevant situation. A recognizable name without an associated problem may not be enough.
**Organic distribution** is distribution you have not purchased as an ad placement. Production, tools, and people's time can still cost money.
An **impression** is a display of content, not proof of careful reading. Distinct reach is a different measure, and neither establishes buying intent.
**Dark social** is shorthand for sharing that your usual attribution tools do not fully observe: a screenshot in a team chat, a copied link, or a colleague mentioning your explanation.
A **company page** speaks for an organization. A personal profile speaks for a person. They can support each other without publishing identical words or pretending to be interchangeable.
## Keep this in mind
Make it easy for a relevant reader to answer three questions: **What does this person understand? Why should I trust this particular claim? What can I do with it?**
You can answer those questions in a short observation, a longer explanation, a demonstration, or a company announcement. You do not have to disguise every commercial topic as a personal story.
The other half of the work is listening. A useful correction, a practical question, or a reader explaining where your advice breaks can improve the next post and the underlying product.
That is a better foundation than trying to make every piece look popular. It also gives a small team something it can control when distribution varies.
## How to do it
Begin with a small editorial system, not a large posting quota. Decide whom you want to help, who can speak credibly, where the material will come from, and how the team will respond.
The system should make good work easier to share. It should not require everyone to become a full-time creator.
### Step 1: Choose the speaker and the topics
Start with the reader's working day.
A support manager, a VP of customer experience, and a software founder may all care about customer support. They do not ask the same questions or control the same decisions. A post that tries to address all three can end up saying little to any of them.
Choose one reader situation to begin with. For example: a support lead is introducing AI drafts but still has to explain who owns an incorrect reply. That situation gives you practical material about review, escalation, measurement, and the limits of automation.
Then choose a speaker who has access to that work. The founder may understand the market and the product decision. A support operator may have a much better example of what happens when a queue is busy. Both can contribute; they do not need the same voice.
Do not choose speakers only by seniority or follower count. Ask who can provide a concrete example, explain the trade-off, and answer a follow-up question without reading from a script.
A person does not need to be the world's leading expert. They can describe a limited observation honestly: what they tried, what they saw, and what they still do not know.
### Give the profile a clear professional promise
A useful post often sends a curious reader to the author's profile. That profile should confirm the context, not require the reader to decode a string of slogans.
Use plain language to explain the work, the audience, and the relationship to the company. “Building review tools for support teams using AI drafts” tells a different story from “AI visionary helping businesses unlock exponential value.”
The first version still needs to be true. Do not add customers, responsibilities, awards, or outcomes because they would make a stronger headline.
Keep the current role, company connection, and relevant work easy to find. Where your profile's available sections allow it, feature a useful guide, a clear product explanation, or a piece that represents the topics you want to discuss.
The goal is continuity: the post raises an interesting question, the profile explains why you work on it, and a relevant destination lets the reader go deeper.
A portrait, banner, and consistent visual treatment can help recognition. They do not compensate for unclear positioning. Fix the explanation before commissioning an elaborate personal-brand package.
### Choose a few lanes that can produce real examples
A small team might start with three lanes: diagnosing a recurring problem, explaining an operating decision, and showing what it learned while building or delivering something.
For the support example, those could become:
* How teams decide which AI drafts need review.
* What an escalation handoff needs to contain.
* What changes when the product meets a real support workflow.
These are starting choices, not mandatory percentages. A research organization may have a different mix from a founder-led software company.
Test each lane against available material. Can you name several different questions within it? Do you have evidence or experience to contribute? Would a relevant reader recognize the connection to their job?
Avoid turning a lane into a slogan that appears at the end of every post. Consistency comes from returning to a useful set of problems, not repeating an identical conclusion.
You can occasionally write outside those lanes. A genuine personal observation may help people understand how you work. The question is whether your overall presence still creates a coherent professional association.
### Build a material bank from work you already do
Most teams have more material than they notice. The difficulty is capturing it before it disappears into calls, chat threads, and completed tasks.
Look for questions that required a careful answer: a prospect misunderstanding a feature, an implementation trade-off, a customer's objection, a failed experiment, or a decision your team reversed.
Capture the question, the relevant detail, the lesson, and the evidence boundary. “Customers need trust” is too vague. “The team wanted to know whether a supervisor could reconstruct why a draft was approved” gives you something to explain.
Keep private source material separate from publishable notes. A customer call may inform a general lesson without giving you permission to publish the customer's identity, screenshot, or exact words.
Anonymization is not always enough. A combination of industry, timing, company size, and incident details can identify someone even when their name is removed. When needed, use an explicitly fictional teaching example or obtain permission for a specific disclosure.
Do not manufacture a first-person story from a composite. “Here is an illustrative situation” is honest. “A customer told me this yesterday” is not, unless that conversation actually happened and you can appropriately share it.
### Interview the speaker instead of asking for a finished post
An editor can make participation much easier by collecting the substance first.
Ask what happened, what was surprising, what the speaker would do differently, and which detail another practitioner would need to apply the lesson. Follow up when the answer becomes abstract.
The following conversation is fictional.
**Editor:** You said the team needed more confidence in AI. What did they actually ask to see?
**Founder:** They wanted to know who approved a reply and what that person checked. I had assumed a confidence score would be enough.
**Editor:** What did that change in the product discussion?
**Founder:** We stopped treating the score as the complete explanation and mapped the review decision. That is the part I can describe; I cannot claim we solved every trust problem.
That exchange has an observation, a mistaken assumption, a decision, and a limit. The editor now has a post to shape, rather than a vague theme to decorate.
Keep a short record of the speaker's approval. If a draft introduces a stronger claim than the source supports, return it for review. Good editing makes an idea clearer; it does not make the history more impressive.
### Step 2: Help readers before asking them to buy
A reader may be learning about a problem long before they have a budget or project. Give them something useful in that state.
LinkedIn's May 18, 2022 explanation of the 95-5 Rule makes the case for considering category buyers who are not currently shopping. It is a framing for long-term demand, not a measurement showing that exactly 95% of your followers are out of market. [LinkedIn's 95-5 explanation](https://www.linkedin.com/business/marketing/blog/research-and-insights/why-you-should-follow-the-95-5-rule?ref=b2b-playbook)
You do not need to translate that framing into a rigid ratio of educational and sales posts. A useful product demonstration can educate. A personal anecdote can be empty. Judge the piece by what it helps a reader understand or do.
If you are announcing an available feature, say what it changes, who it is for, and what its limitations are. There is no need to hide the product name until the final line.
If you are sharing a lesson with no commercial next step, let it stand. Not every post needs to push someone toward a form.
### Turn a general claim into a useful post
Consider this fictional first draft:
> AI is transforming customer support. The most successful teams will combine human empathy with powerful automation. The future belongs to companies that embrace this change.
It is difficult to disagree with, but difficult to use. It does not tell a support lead what to inspect on Monday.
A more specific version might be:
> Before you measure how many replies AI drafts, decide who can approve one.
>
> Take a refund request as a test. If the draft promises an exception to policy, can the reviewer see that exception before sending?
>
> A useful review record needs more than an “approved” label: it should show the proposed reply, the relevant policy, and who made the decision.
>
> That will not solve every support risk. It does make one important handoff easier to inspect.
This is still an illustrative draft, not a report of a real customer outcome. It gives the reader a test and makes a limited claim.
Notice what changed. The topic became a situation. The situation exposed a decision. The decision led to a practical check. The final sentence states a boundary instead of a sweeping promise.
The post can be improved further with a real, publishable example. But it no longer depends on a dramatic hook to conceal an absence of substance.
### Write an opening that earns the next sentence
The opening should help the right reader recognize the subject quickly. It can begin with a question, a surprising observation, a result, or a familiar scene.
“Your AI review queue can be fast and still be hard to audit” is more informative than “Most leaders are getting AI wrong.” The first line tells the reader what tension the post will examine.
Do not imply a result the body cannot support. If the post describes a design hypothesis, do not open with a claim that it doubled retention. If an example involves three interviews, do not present it as the view of an entire industry.
Preview the post on a phone. The visible opening and expansion behavior can vary, so avoid building a strategy around a supposedly permanent character threshold.
Short lines can improve scanning. Breaking every sentence into isolated fragments can also make an ordinary observation feel theatrical. Use spacing to clarify the reasoning, not to create artificial suspense.
Read the opening and ending together. If the opening promises a concrete method and the ending reveals only a product link, the post has not kept its promise.
### Use a structure that fits the thought
A practical observation can follow a simple sequence: situation, tension, decision, example, limit. You do not need to label those parts in the published post.
An explanation might start with a mistaken assumption and replace it with a more useful distinction. For instance, “a reviewed reply” and “a reply someone can later explain” are not necessarily the same thing.
A short teardown can examine one permitted example: what it does well, where a reader might struggle, and how to improve it. Critique the work accurately without turning the author into a target.
A product post can show a task before and after a change, while naming what remains manual. A research post can state the finding, sample, timeframe, and uncertainty without pretending one chart explains the whole market.
Choose one main job for a short post. If you need six unrelated lessons to make the piece feel substantial, you may have six drafts or a longer article.
### Make evidence easy to inspect
Separate what you observed from what you infer.
“We heard the same question in four onboarding calls” is an observation if the record supports it. “Every support team wants this feature” is a much broader claim.
When using numbers, state the denominator and period that make them meaningful. A 50% increase from two to three requests is not the same business situation as a large, repeated improvement.
If you share a screenshot, review it for customer information, internal URLs, personal data, and details that should not be public. A small unreadable image can still expose information when expanded or downloaded.
Credit the source of a framework, chart, or quotation. A link can support a claim; it does not make copying someone else's whole post into your own voice acceptable.
Do not force readers to leave the post for every caveat. Put the most important limitation beside the claim, especially when a screenshot or excerpt may travel without the surrounding context.
### Choose a next step that matches the post
The next step might be a useful question, an open worksheet, a longer explanation, an event invitation, or a product conversation. It should follow naturally from the content.
If the post explains a method and the reader needs the complete worksheet, link to the worksheet. If the post explains who a new paid service is for, a clear inquiry route may be appropriate.
Do not hide a necessary link solely because someone claims all external links are punished. The sources checked for this chapter do not establish a universal best placement for your links. Test reader experience and your own results without turning a small observation into an algorithm law.
Likewise, “comment a keyword and I will send the resource” is not a necessary step in content distribution. An accessible direct link is often simpler. If you genuinely need to provide something individually, describe the arrangement honestly and do not use it to manufacture comments or enroll people in unrelated follow-up.
A question should be one you are prepared to discuss. “Which exception is hardest for your team to review?” can produce useful context. A routine “Agree?” is usually asking for a signal, not an answer.
### Step 3: Make the post useful enough to share
A reader should be able to forward the piece to a colleague with a sentence such as “This is the handoff problem we discussed.”
That requires enough context to survive outside your feed. Define the situation, make the example readable, and keep the central limitation attached.
Being shareable does not require a branded carousel. A short text explanation can travel well. A diagram is useful when the relationship is difficult to describe linearly. A video is useful when the reader needs to see a task happen.
Choose the format after the idea. Otherwise the team can spend an afternoon designing slides around a sentence that still says very little.
| Format | Use it when | Check before publishing |
| ---------------------------- | --------------------------------------------------------------- | --------------------------------------------------------------------------- |
| Text post | A situation, distinction, or short explanation carries the idea | Does it contain an example rather than only conclusions? |
| Image or diagram | A relationship, comparison, or workflow is easier to see | Can a phone reader understand it without tiny text or color alone? |
| Document-style post | A sequence needs several readable steps | Does each page add value, with a coherent start and ending? |
| Native video | Motion, a demonstration, or the speaker's explanation matters | Are captions accurate, the opening clear, and the demonstration honest? |
| Longer article or newsletter | The topic needs sustained explanation or a recurring edition | Can you maintain the promise and distinguish it from owned-site publishing? |
These are editorial choices, not a ranking of formats by guaranteed reach. Verify current feature availability and upload requirements in the product before preparing a large batch.
For a diagram, remove decorative detail until the main relationship is obvious. For a document, use readable type and descriptive page headings. For a video, do not make sound the only way to understand the central point.
If an asset contains a lot of reference material, offer a maintained destination as well. A feed post is a poor place to keep correcting a complex procedure over several months.
### Participate where the topic already matters
Publishing is only one way to contribute. Read relevant practitioners, customers who choose to speak publicly, researchers, and industry publications.
Start with a small set you can actually follow. Look for recurring questions and differences in how people describe the same problem. That helps you learn the language of the work.
A useful comment adds an example, a distinction, a source, or a genuine question. It should make sense even if the author never visits your profile.
Do not turn every comment into a miniature sales pitch. If your product is directly relevant and you mention it, make your relationship clear and explain the relevance without hijacking the discussion.
Avoid quotas that reward the appearance of participation. Twenty generic replies are not automatically more valuable than one careful answer to a question you understand.
### Treat disagreement as useful information
The following exchange is fictional.
**Reader:** Your review checklist looks too heavy for a queue handling simple order-status questions.
**Founder:** That is fair. I meant replies that can change an entitlement or promise an exception. I should have made that boundary clearer.
**Reader:** Then the hard part is sorting the queue before review, not adding the same checks everywhere.
**Founder:** Yes. I will separate the routing decision from the review checklist in the next version.
That is a better outcome than defending an overbroad claim. The reader supplies a missing distinction, and the author improves the work.
You do not have to agree with every objection. If you disagree, address the actual point and explain the evidence or assumption behind your position.
Do not expose a customer's private situation to win an argument. If a response would require confidential context, say that you cannot discuss the details and offer a general explanation where useful.
Correct material mistakes visibly. If an error affects the central claim, a quiet edit may not give earlier readers enough context. Choose an appropriate correction and update the maintained source too.
### Move a conversation to private only when it helps
A reaction, profile visit, or follow is weak evidence about what someone wants from you. Do not treat it as a standing invitation to pitch.
A private conversation becomes more natural when the person requests a resource, wants to discuss their own situation, or agrees that the question would be easier to handle privately.
Keep the purpose explicit. Moving to private should help the reader, not conceal the fact that you are changing a content discussion into qualification.
The following exchange is fictional.
**Reader:** Do you have the review checklist in a format I can adapt for our team?
**Founder:** Yes, the public version is linked in the post. If you want to discuss your routing problem, we can continue here or privately—whichever you prefer.
**Reader:** Privately would be easier. We have a few constraints I do not want to put in the comments.
**Founder:** Understood. We can start with the workflow. There is no need to share customer records or sensitive examples.
If the discussion reveals a relevant product need, ask whether the person wants to explore it. Do not pretend a requested worksheet was a request for a sales call.
If it becomes a business opportunity, record the source and the person's actual question. Avoid storing unnecessary personal detail or describing an inferred buying timeline as something they told you.
### Do not manufacture the conversation
LinkedIn's March 13, 2026 update explicitly says engagement pods and automated comments are not allowed, and describes reach limitations and possible account restrictions. A coordinated group designed to inflate engagement is different from colleagues independently contributing useful, genuine perspectives. [LinkedIn's authenticity update](https://news.linkedin.com/2026/authentic-content-and-conversations?ref=b2b-playbook)
The platform's March 12 feed update also describes efforts to reduce repetitive, low-substance content and engagement bait. It explains a direction of product development, not a formula that guarantees distribution for any particular post. [LinkedIn's feed update](https://news.linkedin.com/2026/ImprovingTheFeed?ref=b2b-playbook)
Do not buy a service that promises to solve this through human-looking automated replies. Nor should you rotate accounts, copy cookies, or disguise coordinated activity to avoid detection.
The practical alternative is less glamorous: publish work you can defend, answer real questions, and build relationships with people who have a reason to care.
Check the basics before changing the whole strategy. Is the intended reader clear? Is the opening understandable? Does the post contain a useful example? Was the content shared with the intended audience setting?
Then inspect the response you did receive. A small number of relevant questions can identify a promising topic, while a large volume of generic reactions may reveal little.
Look across several comparable posts rather than treating one result as a stable rule. Timing, audience, topic, format, and prior relationships can change together.
Do not assume that low reach proves an account restriction or that a specific formatting trick will fix it. If the product shows an actual account or policy notice, address that notice through the platform's documented process.
### Step 4: Give the company page a supporting role
For a small founder-led team, the company page can provide continuity: what the business does, what exists now, evidence it can publish, open roles, and useful resources.
That is a starting arrangement, not a rule that pages must be passive libraries. A company page can have its own editorial purpose and build an audience around research, customer education, industry explanation, or product use.
Choose a job that justifies the work. If the page has a strong specialist readership, it may be a primary publishing surface. If the founder's audience is more relevant, the page may support that activity with dependable product information.
Avoid making the page impersonate a person. “We changed the escalation interface after reviewing these usability problems” is a legitimate company statement. A fabricated first-person founder anecdote is not.
The personal post and page post can address different parts of the same work. The founder explains the trade-off; the page publishes the release details and current limitations. Readers who encounter both should gain context rather than see duplicate slogans.
### Make employee participation voluntary and distinct
Provide source material, context, and permitted claims. Let employees decide whether they have a perspective worth sharing through their personal accounts.
A product manager might explain a decision. A support specialist might describe a workflow pattern they are allowed to discuss. A recruiter may be the right person to explain a role. Their contribution does not need to begin with the same approved sentence.
Do not require people to endorse a personal opinion or disclose experiences they prefer to keep private. Employment is not a reason to take ownership of someone's voice or relationships.
Keep company-page access separate from personal-account control. LinkedIn provides Page admin roles with different permissions; use the role appropriate to the work rather than sharing one person's login. [LinkedIn Page admin permissions](https://www.linkedin.com/help/linkedin/answer/a550647/linkedin-page-admin-roles-permissions?lang=en\&ref=b2b-playbook)
LinkedIn's User Agreement also prohibits account sharing and unauthorized automated activity, including scraping and certain messaging or engagement actions. Review a tool's authorized access and actual behavior rather than assuming every extension is permitted. [LinkedIn User Agreement](https://www.linkedin.com/legal/user-agreement?ref=b2b-playbook)
An editor can prepare a personal draft outside the platform. The speaker still needs to review and approve what represents them. Do not solve a publishing bottleneck by asking for their password or session cookies.
### Use AI as an editor, not as a source of experience
AI can help turn rough notes into alternatives, shorten a repetitive paragraph, identify an unsupported claim, or suggest a clearer structure.
It cannot supply a real customer conversation that never happened. It should not invent a founder's emotions, personal hardship, operating results, or expert opinion to make a post more compelling.
Start with actual notes and a clear boundary: what is known, what is inferred, and what must not be disclosed. Ask for edits to that material, then verify the output.
A useful review question is: “Could this sentence have been written for almost any company?” If yes, replace it with a relevant detail or remove it. Another is: “Would the named speaker say this in a working conversation?” If not, ask why the draft sounds different.
LinkedIn's June 4, 2026 statement, originally published May 20, says AI assistance can help with writing but that posts and comments should represent the author's perspective. It also describes reducing distribution of generic AI-like content and acting on low-human-involvement comment automation. It does not say all AI-assisted writing is prohibited. [LinkedIn on keeping conversations real](https://news.linkedin.com/2026/keeping-conversations-real-on-linkedin?ref=b2b-playbook)
Do not treat “make it sound human” as a request to add fake mistakes or invented vulnerability. Better writing usually comes from better source material and a more specific point.
### Run a small weekly production loop
Set aside one short source conversation or notes review. Choose a few promising ideas, draft only the ones with enough substance, and get the speaker's approval before publication.
Schedule production separately from participation. A post can be prepared in advance, but an answer to a reader should respond to what they actually said.
LinkedIn offers native scheduling for supported posts. Check the current composer and scheduled queue for the account and format you are using; do not assume every content type has identical support. [LinkedIn scheduling help](https://www.linkedin.com/help/linkedin/answer/a1347212/schedule-posts?lang=en\&ref=b2b-playbook)
Before a scheduled post goes live, check whether its context is still accurate. A release may be delayed. A customer may withdraw permission. An unrelated event may make the planned wording insensitive or confusing.
Give someone responsibility for genuine questions and corrections. You do not need to stay online continuously or perform a first-hour engagement ritual. You do need a response arrangement that matches the expectations you create.
Keep a small reserve of evergreen material so a busy week does not force the team to publish an unfinished claim. If there is nothing useful ready, a missed slot is better than a false story.
### Make repetition earn its place
A useful topic can return with a new example, a different audience question, or a correction to an earlier explanation.
Do not confuse repetition of a relevant idea with reposting the same text indefinitely. Ask what a returning reader gains this time.
A practical sequence might introduce a problem, show a checklist, examine an exception, and demonstrate a product capability. Those pieces reinforce an association while each doing its own job.
When adapting an older post, check whether the facts, permissions, and product state remain current. Preserve attribution if the idea or material came from someone else.
A maintained article or guide can hold the full explanation. LinkedIn posts can open different doors into it without pretending each excerpt is a newly discovered insight.
### Step 5: Review buyer feedback alongside platform metrics
Use platform metrics to understand distribution and interaction. Use conversations and business records to understand what relevant people did next. Neither view is complete on its own.
LinkedIn's current post-analytics help distinguishes impressions from estimated distinct members and Pages reached. It also describes saves, sends, external-link clicks including repeats, and demographic breakdowns subject to availability and privacy protections. These are not interchangeable counts of potential customers. [LinkedIn post analytics](https://www.linkedin.com/help/linkedin/answer/a516971/post-analytics-for-your-content?intendedLocale=en\&lang=en-us\&ref=b2b-playbook)
Keep a dated snapshot for the review window you chose. Label the metric, account, post, and observation period. A number copied from a dashboard without that context becomes hard to interpret later.
Do not assume a reach estimate provides a list of everyone who read a post. Visible reactions, profile information, and private messages each reveal only part of the interaction.
### Ask what the content is teaching people to associate with you
Read a sample of genuine replies and questions. Are people discussing the problem you wanted to explain, or responding mainly to a dramatic opening?
If a broad opinion attracts the wrong kind of attention, you do not need to declare it a failure as a piece of writing. You do need to decide whether producing more of it serves your intended professional presence.
Compare themes over several posts. Does a recurring practical topic attract relevant questions? Do people use the same language when they ask about the product? Are misunderstandings becoming clearer?
Do not infer everyone's role or need from their job title. A student, consultant, or adjacent practitioner may be a useful participant without being a buyer. Unknown fit should remain unknown.
Audience learning can be valuable, but name it honestly. A lively discussion that changes your explanation is not automatically pipeline.
### Connect the activity to the business without forcing attribution
Use clear destination links when a click is part of the reader's next step. Retain normal campaign parameters where appropriate, but do not put personal information in URLs.
An optional “How did you hear about us?” question can capture a remembered source. Preserve the answer in the person's own terms before classifying it.
“I saw your posts” may describe an ongoing exposure, not the last post. “A colleague sent your checklist” may include a sharing step your web analytics never recorded.
Keep self-reported information alongside observed visits and existing account history. Do not overwrite one with the other to make a dashboard look tidy.
If an opportunity already existed before the post, a later content interaction may be helpful without making the opportunity newly sourced. A consistent attribution rule is a reporting choice, not proof that the post caused the purchase.
### Review effort as well as output
Count the time spent gathering material, drafting, reviewing, designing, responding, and measuring. Otherwise an apparently free channel can quietly consume the founder's week.
You can keep hours separate from cash. If you assign a planning value to internal time, state the rate and treat it as an assumption.
A useful review can end with “keep the topic, simplify the format.” Perhaps the written explanation works well, while the elaborate document version takes too long to produce.
It can also end with “reduce the cadence.” Consistency means a sustainable pattern, not the maximum number of posts the team can briefly force through approval.
A few weeks can reveal production problems and promising questions. They rarely justify a confident long-term revenue forecast for a small B2B audience.
First, separate the job you chose from the result you hoped for later. A customer-education series should not be judged only by new-business inquiries. A series intended to attract suitable prospects should eventually be examined against that objective.
Then inspect the path. Can a curious reader understand what you do, find relevant evidence, and take a reasonable next step? A useful post can lead to a profile or page that leaves the commercial context unclear.
Review the audience and topic together. Broad inspiration can attract people who like the sentiment without having the problem you solve.
Do not turn every post into a pitch to compensate. Try a more specific use case, a clearer product explanation, or a better destination, then observe the result within a defined effort budget.
## Worked example: illustrative
The company, people, posts, costs, and results below are fictional. They show an operating method, not a LinkedIn benchmark or a Lensmor campaign.
A small company is building a review workspace for support teams using AI-drafted replies. The product helps a reviewer see the proposed response, the relevant policy note, and the approval record. It does not automatically decide every exception or replace the team's judgment.
The founder has useful product and customer-research context but does not want to become a daily motivational writer. An editor can help for a limited number of hours each week.
They choose one reader: a support lead introducing AI drafts while still owning quality and escalation. Their initial job is to make the review problem understandable and start relevant conversations, not to maximize total followers.
### Week one: collect material and clarify the profile
The editor asks the founder to explain three recurring questions from work the team can discuss. They remove identifying details and distinguish actual observations from illustrative examples.
The profile is updated to explain the work plainly. The main destination is a guide to review decisions, with a separate product explanation for readers who want it.
They draft two personal posts. One explains why a confidence score and an approval record answer different questions. The other offers a simple test using a fictional refund exception.
The company page does not copy both posts. It carries a clear description of the current product and the limits of the early workspace.
The team agrees that nobody will scrape reactors, send automatic follow-up, or ask employees to post identical comments. The editor supplies drafts; the founder owns their approval and personal publishing.
### The four-week plan
The plan contains six personal posts and two company-page posts. It is a workload choice for this fictional team, not a recommended universal frequency.
| Week | Publishing plan | Question the team wants to learn from |
| ----- | --------------------------------------------------------------- | -------------------------------------------------------------------------- |
| One | Two personal explanations of review decisions | Do support leads recognize the distinction and ask practical questions? |
| Two | One personal checklist and one company-page product walkthrough | Can readers understand both the method and what the product actually does? |
| Three | Two personal posts about exceptions and a corrected assumption | Does a more precise boundary improve the discussion? |
| Four | One personal synthesis and one company-page guide update | Which material deserves a maintained version and another round of work? |
Each post has a source note, a permitted example, an intended reader, and one useful next step or an explicit decision not to include one.
The team also reserves time to read and respond. That time is not a requirement to visit a fixed number of posts or deliver a fixed number of comments.
### A publishable teaching draft
The week-two personal post uses an illustrative scenario, clearly identified:
> Illustrative example: a support lead has ten AI-drafted replies waiting for review.
>
> Nine explain a routine delivery update. One promises a refund outside the normal policy.
>
> Giving all ten the same review checklist does not solve the first problem: deciding which reply needs a different kind of judgment.
>
> Before adding more review steps, write down what should trigger escalation. A policy exception, a missing fact, and a routine answer may need different routes.
>
> The checklist is useful after that decision. It should not hide the decision.
The post can help a reader without requiring a click. A linked guide provides a fuller example for someone who wants it.
The company-page walkthrough explains which part the product supports. It does not claim that the software automatically recognizes every policy exception if that capability does not exist.
This division makes the commercial context available without forcing the personal post to pretend it is independent of the company.
### A reader changes the next draft
A practitioner points out that the checklist assumes the policy note is current. Their real problem is that reviewers sometimes reference different versions.
The founder does not respond with a product pitch. They ask a clarifying question about the general workflow, without requesting private customer records.
The team realizes that the next explanation should separate policy-version control from approval ownership. It updates the guide and makes that distinction the subject of a later post.
This is a useful outcome even before an inquiry appears. It is recorded as a content and product-learning result, not assigned a fictional monetary value.
The editor also removes a planned post that repeats the first week's message without adding an example. There is enough material to keep publishing, but no reason to publish every draft.
### What the first review shows
Across the eight posts, the team records 18,000 total impressions for its chosen reporting window. This is a sum of displays across posts and accounts, not 18,000 distinct people.
The two posts that include the guide link show eighty external-link clicks in the relevant platform reports. Website analytics records sixty-five attributed sessions. The counts differ in unit and collection method; the team does not force them to match.
Eight people make substantive inquiries during the period, representing six companies. Two companies have more than one person involved. The team deduplicates at both levels.
Four of those companies have a suitable problem and take part in a substantive conversation. Two are existing relationships; two are newly recorded. One of the two new companies agrees to an evaluation next step and meets the team's opportunity definition. No customer has closed by the review date.
| Observed result | Count | What it does not establish |
| --------------------------------------------- | -------------------- | ---------------------------------------------------- |
| Impressions across eight posts | 18,000 | Distinct people, careful reading, or buyer fit |
| Reported external-link clicks on two posts | 80 | Eighty unique visitors or customers |
| Attributed website sessions | 65 | Complete visibility into private sharing |
| People making substantive inquiries | 8 across 6 companies | Eight independent buying opportunities |
| Suitable company conversations | 4 | Four new accounts; two relationships already existed |
| Newly recorded suitable company conversations | 2 | That LinkedIn alone caused them |
| New opportunities under the stated definition | 1 | Closed revenue or a stable conversion rate |
Some inquiries mention a particular post. Others say they had seen the founder's explanations over time. The team keeps those answers rather than assigning all eight inquiries to the most recent post.
It also checks whether any content was boosted. In this scenario none was, so the report is about unpaid distribution. If paid activity had been added, that would need separate labeling.
### The cost review
The founder spent eight hours gathering material, reviewing, publishing, and responding. The editor spent twelve hours interviewing, drafting, coordinating, and reviewing results. External design help cost \$300.
For planning, the team values founder time at \$100 per hour and editor time at \$50 per hour. That gives \$800 plus \$600 plus \$300, or \$1,700 of assumed all-in effort and cash cost.
The actual incremental external cash payment is \$300 in this example. The internal-hour values are assumptions, not additional vendor invoices.
Dividing \$1,700 by four suitable company conversations gives \$425 per conversation. Dividing by the two newly recorded suitable company conversations gives \$850. The denominators answer different questions.
Neither figure is customer acquisition cost. There are no closed customers in the observed cohort, and the source information does not prove incremental causation.
The company decides that the work is affordable enough for another bounded period. It does not conclude that every additional \$1,700 will produce the same results.
### What changes in the next period
The team keeps the practical review lane because it produced specific questions and improved the guide. It reduces design work because a clear text explanation performed its reader job without a complex asset.
It adds one product-focused post to clarify what exists now, while continuing to explain the workflow without a mandatory sales CTA.
The company page retains the product walkthrough and maintained-guide updates. It is not asked to imitate the founder's personal commentary.
The next review asks whether new readers can correctly explain the problem and whether suitable conversations remain manageable. The team continues observing the existing opportunity separately.
If the effort grows without clearer audience fit or useful outcomes, it will reduce the program or change the topic. That is a channel decision, not a verdict on the founder's personal worth or writing talent.
## Copy: LinkedIn organic card (fill)
Use this before drafting a small batch. It should clarify the work, not become a lengthy approval form for every sentence.
```text theme={null}
LINKEDIN EDITORIAL CARD
Reader and working situation:
Speaker or company-page editor:
Connection between this topic and our actual work:
Two or three recurring content lanes:
For each draft:
Reader question:
Specific observation, example, or demonstration:
What is factual / inferred / explicitly illustrative:
Source and permission checks:
Main point in one sentence:
Useful next step, or why none is needed:
Speaker approval and response owner:
Profile or Page destination to keep current:
Publishing and response time available:
Distribution and interaction metrics to observe:
Useful business or learning outcome:
Review date and effort cap:
What we will not automate or manufacture:
```
The [existing working file](../../templates/linkedin-organic.md) is a shorter, personal-voice-first card. Its “95” framing and supporting-page assumptions are starting prompts, not a measured audience split or a prohibition on company-page publishing, useful product posts, or platform metrics.
### Copy: weekly reading and response review
Use a short record to connect the published work with what you learned. Keep it free of unnecessary personal information.
```text theme={null}
LINKEDIN WEEKLY REVIEW
Account / posts / observation window:
Topics and formats published:
Actual production and response hours:
Cash cost and any internal-time valuation:
Platform metrics with their exact names and units:
Relevant questions or corrections:
What readers understood or misunderstood:
Source mentions in inquiries, preserving original wording:
Distinct people and distinct companies:
Existing relationships versus newly recorded companies:
Suitable conversations / opportunities / closed outcomes:
Attribution limits and unobserved sharing:
Content or product explanation to improve:
Claims or links that need correction:
Keep / change / stop for the next period:
Specific next test and owner:
```
## Before you start
Read the profile and one draft together. Would a stranger understand the connection between the person's work, the post, and the company?
Check the source behind the most impressive sentence. Confirm that the named speaker can stand behind it and that any customer material is permitted.
Preview the post and asset at phone size. Important text should be readable, and the central point should not depend on color or sound alone.
Test the destination a reader is invited to use. Make sure it delivers the promised resource or explanation without an unexpected sales gate.
Confirm who will answer a practical question, handle a correction, and pause a scheduled post if circumstances change.
Finally, agree on a modest effort budget and review date. You are building a repeatable way to contribute, not promising a particular impression count.
## Metrics
### Distribution is worth measuring
Impressions, reach estimates, profile activity, and follower changes can help diagnose whether the work is being encountered. They are not automatically meaningless vanity metrics.
They become misleading when they are used as a substitute for the outcome being claimed. A reach chart cannot, by itself, establish that buyers trust the company or that pipeline increased because of a post.
Compare like with like. Keep account, format, observation window, and whether paid distribution was involved visible. A month-old post and yesterday's post have had different opportunities to accumulate activity.
Do not add per-post reach estimates and label the sum a deduplicated audience. The same person may encounter several posts.
### Interaction needs interpretation
A save may suggest someone wants to return to the material. A substantive reply may reveal a relevant problem. A send may indicate sharing. None gives you the full story of what happened afterward.
A comment from a colleague can be genuine and useful. It should not be counted as an independent customer endorsement if it is not one.
Look at the substance of questions as well as the volume. A misunderstanding repeated across several readers may be more actionable than another small change in an engagement ratio.
If you calculate a rate, state the numerator and denominator. Reactions divided by impressions is not the percentage of readers who want a sales conversation.
### Business outcomes require separate records
Track actual inquiries, suitable conversations, opportunities, and closed outcomes using definitions your team already understands.
Keep people and companies distinct. Multiple members of a buying team may interact with the content, while one person may appear in several source records.
Preserve uncertainty. A self-reported source can be useful without identifying the causal effect of a particular post. A prospect may have seen content after an introduction or while already evaluating you.
Use the [measurement model](../09-operations-pipeline-and-measurement/measurement-model.md) to connect these views. Do not make the channel look measurable by inventing a complete path that nobody observed.
## Common mistakes
**Writing for everyone in B2B.** Choose a recognizable working situation and a reader who would care.
**Editing away the useful detail.** A polished paragraph about transformation is weaker than a clear explanation of one actual decision.
**Inventing personal experience.** AI and ghostwriting support do not authorize fabricated customers, conversations, or outcomes.
**Treating the company page as inherently ineffective.** Give it a defined job and review its audience and output rather than assuming personal accounts always win.
**Making employees a distribution mechanism.** Provide context and let genuine participation remain voluntary.
**Turning every reaction into outbound.** Respond to actual requests and keep the transition to a business conversation explicit.
**Chasing a universal algorithm formula.** A platform update is not proof of a perfect posting hour, format, or link placement for your audience.
**Calling organic free.** Count the production and response work before increasing the cadence.
**Ignoring a commercial next step on principle.** A useful product post can include a clear inquiry route. The problem is a mismatch between the reader's expectation and the ask.
## What to read next
[Content strategy](../03-brand-story-and-content/content-strategy.md) covers the broader editorial system. [Founder story](../03-brand-story-and-content/founder-story.md) helps you use real experience without inventing a personal narrative.
Use [Messaging](../02-product-marketing/messaging.md) when people understand the post but cannot explain what the company does. [Community](https://b2-b-playbook.mintlify.app/playbooks/04-channels-and-distribution/community) covers ongoing participation beyond a feed.
For direct, person-specific outreach, read [LinkedIn outbound](../05-outbound-and-prospecting/linkedin-outbound.md). For purchased distribution, read [Paid media](https://b2-b-playbook.mintlify.app/playbooks/04-channels-and-distribution/paid-media); separate those costs and results from an organic review.
[Creator partnership](https://b2-b-playbook.mintlify.app/playbooks/04-channels-and-distribution/creator-partnership) covers collaborating with another person's audience and judgment. [Measurement model](../09-operations-pipeline-and-measurement/measurement-model.md) explains how to combine platform, website, and buyer-reported evidence.
## Sources and evidence boundary
LinkedIn's primary materials were checked on September 8, 2026. Links beside the relevant statements identify the source.
The March 12 feed update and March 13 authenticity update describe the platform's stated direction and restrictions. The June 4 AI-content statement notes its original May 20 publication. None provides a controlled comparison proving that a particular organic format or cadence will work for this business.
The post-analytics help defines platform measurements and their limits. It does not reveal every reader's identity or establish a complete attribution path. Product features and available reporting can change; check the current interface before building a workflow around them.
The 95-5 source is a long-term marketing framework, not a measured split of this repository's audience or your followers. Page-role, scheduling, and account-policy sources support operational boundaries, not a recommendation to buy a tool or subscription.
The editorial methods, drafts, conversations, templates, and numerical example are original teaching material. All example results and costs are fictional. No LinkedIn account analytics, private messages, customer data, or actual Lensmor campaign were inspected for this chapter, and no LinkedIn post, comment, connection request, or message was sent.
***
Copyright © 2026 Ivan Xu. All rights reserved. See the [copyright and reuse terms](https://github.com/weilun88313/B2B-Playbook/blob/main/LICENSE).
Canonical source: [github.com/weilun88313/B2B-Playbook](https://github.com/weilun88313/B2B-Playbook)