You publish a product update on LinkedIn. A few former colleagues like it, one person writes “Congrats!”, and by lunchtime it has disappeared under a sales-award selfie and somebody's B2B leadership lesson from missing a train.
The post asked a cold feed to care about a company it barely knew. LinkedIn gives you a better route because your work history and professional relationships sit beside whatever you say. If you leave a smart comment about a problem you know well, the person reading it can open your profile and see why you know it.
This is where LinkedIn is strongest for a solo founder: it can make your expertise recognizable before you need to ask for a call, trial or sale.
Check that your buyer uses LinkedIn
LinkedIn is often a common route for B2B products because the people involved have a professional identity there. Your buyer may post about their work, follow practitioners in the field or attend events that produce a public conversation. Even when they rarely publish, their colleagues and referrers may be active.
Having the right job titles isn't enough on its own. You need some sign that LinkedIn gives you a route into the subject these people care about.
| If this is what you find | What LinkedIn may help you do | What you still don't know |
|---|---|---|
| Buyers discuss the problem and answer their peers | Join the same conversations and become recognizable | Whether attention will turn into a serious conversation |
| Buyers are present but rarely publish | Research profiles and support later introductions | Whether public posts will reach enough of them |
| Practitioners and referrers are active | Build familiarity with people your buyer trusts | Whether those relationships can reach the buyer |
| The buyer and their professional circle are mostly absent | Very little | Which channel offers a shorter route |
LinkedIn is a weaker choice when the subject is too private for public discussion or the sale depends on a small group of senior people who are easier to reach through an introduction. The earlier channel chapter covers that in full.
Give the first six weeks one purpose
“Grow on LinkedIn” isn't a job. You can gain followers from people who'll never use the product, and a post can travel widely without producing one relevant conversation.
The first task should reflect what the business needs now:
| Your first task | What progress looks like |
|---|---|
| Become professionally recognizable | More profile interest from the roles or companies you named |
| Start qualified conversations | Comments or posts lead to exchanges worth continuing |
| Distribute proof | The right people save, send or ask about evidence from your work |
| Send people somewhere you own | Relevant people continue to a guide, tool, newsletter or product |
You may see more than one result, but choosing a primary job gives you something solid to review at the end. This is a six-week experiment because it'll take that long to see something meaningful, and anything longer to feed a LinkedIn that might be reaching the wrong people may destroy your soul. This six weeks is a recommendation, not a secret LinkedIn incubation period.
Your personal profile leads
For most solo founders, the personal profileThe LinkedIn page attached to an individual member. It contains the person's work history, About section, activity and connections. should do the early distribution work. Your experience gives somebody a reason to take the contribution seriously, and your existing relationships give them context that a company pageThe LinkedIn presence for an organization. In this chapter it supports company information, releases and proof while the founder's personal profile leads early distribution. can't manufacture.
The company page still has its place: keep the company description accurate and give product news or formal proof somewhere to live. I just wouldn't ask it to start the conversation when you can comment and post in your own voice.
A cold-profile read starts with somebody who has only seen one comment from you:
- Can they tell which problem you understand?
- Does the opening of your About sectionThe profile section where a member can explain their work, experience and the people they help in more detail. explain who gets help and what changes?
- Does your current role match the work you're doing now?
- Is there one piece of proof or explanation in FeaturedA profile area where a member can highlight selected posts, articles, links or media.?
- Can they find one sensible next step?
The positioning you wrote earlier in the playbook should supply most of this. LinkedIn doesn't need a bigger claim or a title containing six nouns separated by vertical bars. It needs a stranger to understand you without conducting an investigation.
Take a hypothetical founder building rota software for independent care homes. Founder | SaaS | AI | Future of Work tells the manager almost nothing. Building rota software for independent care homes still coordinating shifts in spreadsheets gives the work a subject and a buyer. The About section can then explain the problem in ordinary language, while Featured holds one clear product walkthrough. None of this claims a customer result the founder hasn't earned.
What the algorithm is trying to work out
LinkedIn builds a different feed for each member. It first retrieves posts that might be relevant from the person's network and from people they don't follow. A rankingThe stage where LinkedIn orders the retrieved posts for one member. The order can change between members because their profiles, networks and activity differ. model then decides which of those posts the member is most likely to read or interact with now. The same post can sit near the top of one person's feed and never appear in another's.
The 2026 system does this with a Generative RecommenderLinkedIn's name for its 2026 feed-ranking model. It reads a member's previous interactions as a sequence so the feed can reflect changing professional interests., LinkedIn's name for a model that reads a person's activity as a sequence. It can consider more than a thousand previous interactions. Earlier systems evaluated each impressionOne instance of a post being shown on LinkedIn. The same member may produce more than one impression. independently. Somebody's industry, experience and skills give the model professional context; what they read, return to or skip shows how their interests change. (LinkedIn Engineering, 12 March 2026)
How LinkedIn builds one person's feed
LinkedIn starts with possible posts, retrieves a smaller set that may fit one member's professional interests, ranks those posts for that member, and learns from what the member reads, skips, returns to and visibly interacts with. Activity informs later recommendations. There is no single post score; two people can receive a different order.
- Possible posts
Posts from the member's network and relevant people they don't follow.
- Retrieval
LinkedIn selects a smaller set that may fit the member's professional interests.
- Personalized ranking
The model considers the reader, the post, the author and previous activity.
- What the member does
Reading, skipping, returning and visible interactions become part of the history used for later recommendations.
| What the system can consider | What it is trying to estimate | What this changes for you |
|---|---|---|
| The reader's profile, network and past activity | Which professional subjects and people are relevant to them | Choose a buyer and subject instead of writing for “LinkedIn” as one audience |
| The meaning, language and age of the post | What the post is about and whether it belongs in the current conversation | Use the words your buyer uses and make the subject clear early |
| Your profile and the reader's relationship to you | Why this post may be relevant from this particular author | Keep your profile and recurring subjects understandable together |
| Whether people read, skip, react, comment, share or return | What happened after the post appeared | Record attention, then look for response from the people you meant to reach |
You may see this wider change called 360BrewA name used in Romy's 2026 LinkedIn guide and wider commentary for the new feed system. LinkedIn's own engineering material calls the deployed model a Generative Recommender and doesn't publish fixed signal weights under the 360Brew name., including in Romy's guide to growing on LinkedIn in 2026. LinkedIn's own engineering material calls the deployed system a Generative Recommender and doesn't publish fixed weights for its signals. The label doesn't change what you do. LinkedIn now compares the post's subject with information about the author and the individual reader's history.
Time spent with a post is one of those signals. LinkedIn calls it dwell timeThe time somebody spends viewing a post. LinkedIn uses short and longer viewing behavior as part of feed ranking, including for people who read without reacting or commenting. and uses a prediction that somebody will scroll past quickly as a negative ranking signal. Its 2026 model also predicts both a skip and a longer read alongside visible actions such as a comment or share. A post can therefore interest the right reader without collecting a public reaction. (LinkedIn Engineering, 1 October 2024; LinkedIn Engineering, 12 March 2026)
The Romy guide turns that into an editing rule: put the direct claim before the “see more” fold, usually within the first two or three lines. Its 140-character figure works as a mobile editing check. The visible amount changes with the screen and formatting, so it isn't a permanent platform law. Once the post is open, a real objection or a method somebody can use gives the reader a reason to stay. Short paragraphs make that material easier to read on a phone.
The guide also emphasises savesAn action a LinkedIn member takes to keep a post for later. and sendsAn action a member takes to share a post privately with somebody else through LinkedIn. more than likes; a save tells you someone wanted the post again, while a send tells you they thought another person should see it. Both give you more to learn from than a passing reaction.
Comments put you inside an existing conversation
Your own posts begin with whatever audience the account already has. A comment begins inside a conversation that already contains a subject and people who care enough to discuss it. For a new or neglected account, that makes commenting compulsory during the first test.
You don't need a daily quota. Three short sessions each week give you enough room to follow current conversations without turning LinkedIn into a second job. Build a small list of buyers, practitioners and referrers, then look for places where your experience adds something the thread is missing.
Your activity also helps LinkedIn personalize the feed you see, so following the right people improves the conversations available for those sessions. LinkedIn doesn't promise that commenting on one buyer's post will increase distribution of your next post. A relevant comment still puts your experience inside a conversation where other people can respond to it.
A comment may add:
- direct experience;
- an implementation detail;
- a supported correction;
- a counterexample;
- a relevant question; or
- a joke that works inside the conversation.
“Great post!” adds none of these. Neither does a paragraph generated from the post's first sentence and sprayed under 40 accounts before breakfast. LinkedIn announced in March 2026 that it was taking action against automated comments, engagement podsA coordinated group whose members agree to react to or comment on one another's posts to inflate engagement. and third-party tools that create fake conversations. (LinkedIn Product News, 12 March 2026)
The first-week pack includes three prepared comments. Each one should be attached to a real conversation and something you can support. A current thread where you have nothing to add can stay unanswered.
A reaction gives you context, not permission
A good public exchange may give you a sensible reason to connect. Continue the subject that brought you together: the event you both attended, the point they raised or the question they asked. A like on your post doesn't automatically authorize a demo request, and a profile view tells you somebody looked without explaining why.
Start with something you know
An early LinkedIn post can do one of three things:
- Explain a specific problem your buyer recognizes.
- Teach a method you have used.
- Show an observation or piece of proof from the work.
The material may come from a customer question, a product decision or a mistake you corrected. The writing format comes after you know what the reader should get from the post. The LinkedIn Playbook explains seven formats; the table below shows where each one fits.
| What you have | A suitable format |
|---|---|
| A repeated customer question | Observation or question |
| A process used in real work | Method |
| A result with a known limit | Number or before-and-after |
| A defensible disagreement | Polarizing statement |
| An experience people avoid discussing | Unspoken truth |
Each format helps organize an idea; none promises distribution. A clear subject and evidence from real work help the model understand the post.
Professional doesn't mean bloodless
Your work gives the account a recognizable territory; your humor and experience give people a reason to remember the person inside it. Career stories, values and parts of ordinary life can all belong on LinkedIn when they make sense coming from the same human. They don't each need a final paragraph explaining what the dentist taught you about leadership.
You still choose the boundaries. A personal profile doesn't require public access to your family, health or finances. Customer details stay private unless you have permission, and an unfinished piece of work doesn't become content simply because the calendar wants a post.
Two posts, three comment sessions
For the first six weeks, I recommend two posts and three short commenting sessions each week. That produces enough attempts to see patterns while leaving time to answer anybody who responds.
| Session | What you're doing |
|---|---|
| Comment session 1 | Respond to current conversations in your named-person list |
| Post 1 | Explain one buyer problem or observation |
| Comment session 2 | Continue replies and join another relevant discussion |
| Post 2 | Teach a method or show proof from real work |
| Comment session 3 | Answer responses and save questions for later material |
The days can move wherever they fit. The routine depends on having something to say and enough time to stay for the response; LinkedIn hasn't published an algorithmic minimum of two or three posts a week.
Give the post an action
LinkedIn says it is reducing engagement baitA prompt designed mainly to manufacture reactions or comments, such as asking people to “comment to agree.” LinkedIn says it is reducing distribution of this kind of material. such as “comment to agree” prompts and recycled posts written to game distribution. Asking a real question remains perfectly normal when you want an answer. The empty request for engagement can go. (LinkedIn Product News, 12 March 2026)
A post may include a link when the destination finishes the job. A checklist can lead to the complete template; a product explanation can lead to the product. Some posts need no link because the idea is complete where it stands.
The same applies to hashtags. A stack of broad tags doesn't give a vague post a subject. One relevant event, campaign or community tag may help somebody follow that conversation, so I wouldn't turn “no hashtags” into a religion either.
Keep the participation human
AI can organize your source material, suggest a format and prepare a draft. You remain responsible for what's behind it. LinkedIn currently recommends disclosing heavy AI use when it wouldn't otherwise be obvious, and it asks members to review and approve anything AI helped create. (LinkedIn Help, updated August 2026)
AI shouldn't invent the story you wish had happened or mass-produce comments that imitate attention. LinkedIn also prohibits third-party software and browser extensions that scrape the site or automate activity, and it may restrict an account when it detects that behavior. Any tool that touches the account needs a fresh check against the current rules. (LinkedIn Help: Automated activity, accessed 17 September 2026)
For the first test, the boundary is straightforward:
- no engagement pods or bought reactions;
- no generic automated comments;
- no mass connectionA two-way LinkedIn relationship created when one member accepts another member's invitation. or message automation;
- no copied posts or invented experience; and
- no customer material you don't have permission to share.
The same rule applies when somebody disagrees with you. Correct an error plainly when one exists, but don't turn every argument into a sales approach. LinkedIn's Professional Community Policies require authentic information and prohibit artificial engagement and unwanted commercial messages. (LinkedIn Professional Community Policies, accessed 17 September 2026)
Measure whether the attention goes anywhere
LinkedIn's current post analytics show how far a post traveled and what people did next. LinkedIn describes these numbers as estimates, and some demographic detail stays hidden when the audience is too small. (LinkedIn Help: Post analytics, updated August 2026)
The job you chose at the start decides what deserves attention:
| What happened | What it may tell you | What it can't prove alone |
|---|---|---|
| A relevant person viewed your profile | Your name or contribution prompted professional interest | Why they looked or whether they'll return |
| Somebody saved or sent the post | They wanted to keep it or show it to somebody else | Buyer fit or purchase intent |
| A qualified comment, connection or DM followed | The subject produced a relevant exchange | A sale |
| Somebody visited the destination | They continued beyond LinkedIn | Whether the visitor was the buyer |
| A call, referral, trial or purchase followed | LinkedIn contributed to a customer action | That every post or impression helped |
Profile search appearancesAn instance of a profile appearing in LinkedIn search results. It is different from a profile view because the searcher may never open the profile. can add another clue about how people find you. (LinkedIn Help: Member analytics, accessed 17 September 2026)
How LinkedIn turns expertise into a route
A readable founder profile supports a relevant comment or post, which may produce recognition, profile interest or conversation, one sensible next step and a customer action or learning. After the contribution there may be no relevant response; recognition may go no further; or conversation may reveal poor fit. Each exit means changing the people, material or channel. After the final stage, continue, change or stop.
- Readable founder profile
- Relevant comment or post
- Recognition
- Profile interest or conversation
- One sensible next step
- Customer action or learning
- Readable founder profile
- The intended person can understand the problem territory and find one next step.
- Relevant comment or post
- The founder adds experience, explanation or proof to a professional subject.
- Recognition
- The right person recognizes the name or the subject attached to it.
- Profile interest or conversation
- They view the profile, reply, connect or continue the discussion.
- One sensible next step
- They visit the selected guide, tool, newsletter or product, or continue in a relevant private conversation.
- Customer action or learning
- A call, referral, trial or purchase follows, or the founder learns why the route stopped.
Your LinkedIn operating brief and first-week pack
The LinkedIn operating brief and first-week pack starts with one intended reader and one job for the first six weeks. It also holds your profile read and the subjects you can support from your own work. The people and conversations you want to follow have their own section.
The completed pack contains one post and three comments, plus your boundaries for connections and messages. Its outcome fields stay blank until something happens. Publishing remains your decision.
At the end of six weeks, the record should make the next choice much easier: continue with the same reader and subjects, change the route or stop giving LinkedIn the time.
Prompt: prepare your first LinkedIn week
The prompt below uses your positioning, current profile and source material to prepare one post and three comments for approval. It can't publish or contact anybody.
Prepare your first LinkedIn weekShow or hide prompt
I want to prepare my first week of a six-week LinkedIn test.
Reader and route
- Intended buyer, user, referrer or credible peer: [specific person or group]
- Professional problem or subject they care about: [details]
- Evidence they use LinkedIn: [profiles, posts, events or conversations]
- Route from LinkedIn to a customer action: [details / unknown]
Primary job
- Choose one: professional recognition / qualified conversation / proof
distribution / owned destination
- Progress I would be pleased to see after six weeks: [details]
Profile
- Current headline: [paste]
- Opening of About section: [paste]
- Current role: [paste]
- Featured material and destination: [paste links / none]
- Positioning from my earlier work: [paste]
People and conversations
- Buyers or users: [profiles]
- Practitioners, referrers or credible peers: [profiles]
- Current conversations I may join: [links and text]
- Why I have something to add: [details]
Post source
- Buyer problem, method or observation I want to explain: [details]
- Direct experience and facts: [paste]
- Evidence and its limits: [paste]
- My opinion: [paste]
- Personal or humorous material, if relevant: [paste / none]
- Private, customer or confidential material that must stay out: [details]
- Possible destination or product link: [link / none]
Voice and boundaries
- My tone of voice and approved examples: [paste]
- Topics I want to be known for: [two or three]
- Topics or details I won't discuss: [details]
- When I would accept a connection or continue in a DM: [details]
- Time I can give responses: [details]
Your task
1. Separate the facts, direct experience, opinion and missing information I
supplied.
2. Assess whether the intended reader appears in the supplied LinkedIn
material. Mark an uncertain customer route as uncertain.
3. Review my profile as a stranger. Suggest only the changes needed to explain
the problem I understand and the next sensible step. Reuse my positioning.
4. Choose one post job: explain the buyer problem, teach the supplied method or
show the supplied observation or proof.
5. Choose one format from observation, number, before-and-after, question,
polarizing statement, unspoken truth or method. Explain why it fits.
6. Draft one post in my voice. Put the direct claim within the first two or
three lines, then format the rest for a phone screen. Use one relevant link
or next step only when it completes the post's job. Don't add engagement
bait or a decorative hashtag stack.
7. Choose three real conversations from the material I supplied and draft one
comment for each. Each comment must add experience, detail, a supported
correction, a counterexample, a relevant question or humor.
8. Keep every comment inside the conversation. Don't add a pitch or turn a
reaction into permission for an unsolicited demo request.
9. Check every claim against my source material. Flag privacy, permission and
current LinkedIn-rule concerns.
10. State whether the amount of AI assistance should be disclosed under current
LinkedIn guidance. Keep publishing and every response under my approval.
11. Leave impressions, reach, profile activity, saves, sends, link visits,
conversations, calls, referrals, trials and purchases blank.
Rules
- Don't invent experience, facts, customer words, results, metrics, profile
history or LinkedIn rules.
- Don't post, comment, connect, message, react, follow or take any external
action.
- Don't scrape LinkedIn, coordinate engagement or create generic comment
variants for several people.
- Don't copy another person's post or imitate their voice.
- Don't force personal material, a product mention or a call to action into the
post.
- Use American English, connected sentences and my supplied tone of voice.
Return exactly these headings
## Reader and route check
## Facts, experience and opinion supplied
## Missing information and evidence limits
## Cold-profile read
## Post job and format
## Post for approval
## Three comments for approval
## Link, hashtag and next-step decision
## Connection and message boundary
## AI disclosure decision
## Blank six-week result record
## Continue, change or stopThe post and comments still need your review before they leave your notes. The model can prepare the words; it can't know that your experience happened or that the other person wants a private message.
Glossary
Glossary21 terms
The meanings carried by the highlighted terms in this chapter.
- Personal profile
- The LinkedIn page attached to an individual member. It contains the person's work history, About section, activity and connections. Chapter definition
- Company Page
- The LinkedIn presence for an organization. In this chapter it supports company information, releases and proof while the founder's personal profile leads early distribution. Chapter definition
- Professional familiarity
- Recognition built when the same person repeatedly sees your name attached to a subject they care about and can understand why you know it. Chapter definition
- Retrieval
- The first stage of feed selection. LinkedIn gathers a set of posts that could be relevant to one member from their network and from people they don't follow. Chapter definition
- Ranking
- The stage where LinkedIn orders the retrieved posts for one member. The order can change between members because their profiles, networks and activity differ. Chapter definition
- Generative Recommender
- LinkedIn's name for its 2026 feed-ranking model. It reads a member's previous interactions as a sequence so the feed can reflect changing professional interests. Chapter definition
- 360Brew
- A name used in Romy's 2026 LinkedIn guide and wider commentary for the new feed system. LinkedIn's own engineering material calls the deployed model a Generative Recommender and doesn't publish fixed signal weights under the 360Brew name. Chapter definition
- Dwell time
- The time somebody spends viewing a post. LinkedIn uses short and longer viewing behavior as part of feed ranking, including for people who read without reacting or commenting. Chapter definition
- Headline
- The short description shown beside a member's name in several parts of LinkedIn. It should help the intended reader understand the founder's current work or territory. Chapter definition
- About section
- The profile section where a member can explain their work, experience and the people they help in more detail. Chapter definition
- Featured section
- A profile area where a member can highlight selected posts, articles, links or media. Chapter definition
- Connection
- A two-way LinkedIn relationship created when one member accepts another member's invitation. Chapter definition
- Impression
- One instance of a post being shown on LinkedIn. The same member may produce more than one impression. Chapter definition
- Members reached
- LinkedIn's estimate of the distinct members and Pages that saw a post. Repeat views are excluded. Chapter definition
- Profile viewers from a post
- LinkedIn's estimate of the distinct members and Page administrators who opened the author's profile from that post. Chapter definition
- Save
- An action a LinkedIn member takes to keep a post for later. Chapter definition
- Send on LinkedIn
- An action a member takes to share a post privately with somebody else through LinkedIn. Chapter definition
- Search appearance
- An instance of a profile appearing in LinkedIn search results. It is different from a profile view because the searcher may never open the profile. Chapter definition
- Engagement bait
- A prompt designed mainly to manufacture reactions or comments, such as asking people to “comment to agree.” LinkedIn says it is reducing distribution of this kind of material. Chapter definition
- Engagement pod
- A coordinated group whose members agree to react to or comment on one another's posts to inflate engagement. Chapter definition
- Owned destination
- A website, product, newsletter or resource the founder controls and can direct an interested LinkedIn visitor toward. Chapter definition
Sources
- LinkedIn Corporate Communications Team. “How LinkedIn Is Improving the Feed to Show More Relevant, Authentic Professional Content.” 12 March 2026.
- Danchev, Hristo. “Engineering the Next Generation of LinkedIn's Feed.” LinkedIn Engineering, 12 March 2026.
- Zhang, Fengyu, Mohit Kothari and Birjodh Singh Tiwana. “Leveraging Dwell Time to Improve Member Experiences on the LinkedIn Feed.” LinkedIn Engineering, 1 October 2024.
- LinkedIn Help. “LinkedIn Relevance: Optimizing the Member Experience.” Accessed 17 September 2026.
- LinkedIn Help. “How the Feed Ranks Content.” Accessed 17 September 2026.
- LinkedIn Help. “Automated Activity on LinkedIn.” Accessed 17 September 2026.
- LinkedIn Help. “Best Practices for Content Created With the Help of AI.” Updated August 2026.
- LinkedIn Help. “Post Analytics for Your Content.” Updated August 2026.
- LinkedIn Help. “Analytics for LinkedIn Members.” Accessed 17 September 2026.
- LinkedIn. “Professional Community Policies.” Accessed 17 September 2026.
- Tincture. “LinkedIn Playbook.” Accessed 17 September 2026.
- Bull, Alice. “How to Grow on LinkedIn in 2026.” Romy, updated 27 July 2026.