Questions
Every question this site answers, in one list. Product questions first; the rest come from the articles, each linked back to its source.
About the product
Do the posts really sound like me?
The voice profile is derived from your own posts and handed to the writer on every run. Edit the output during the first week — your edits feed the learning loop.
Is my LinkedIn account at risk?
Publishing happens through LinkedIn’s official API with the OAuth permission you granted. We never see your password, and you can revoke access from LinkedIn’s settings at any moment. Reading your profile and comments uses a third-party service, only for your own profile, and runs infrequently.
Why don’t you post the replies automatically?
LinkedIn does not let applications comment on a person’s behalf; that permission is open only to selected partners. Rather than hide this behind a button we say it plainly: we write the reply, you send it. The part that costs time is thinking of the reply, not pasting it.
Can I make sure nothing goes out without my approval?
Yes — that is the default. With auto-publish off, every post waits as a draft.
How often does it post?
Three times a week by default: Tuesday, Wednesday and Thursday at 10:00 — the window where LinkedIn engagement is highest. You can change it in settings, and we will warn you if you pick a weekend.
What does profile optimisation actually do?
You name a target title. LinkedIn Recruiter filters on standardised fields before it ranks anything, so we weight the title field in your experience section heaviest — not the headline. We list what is missing in order of impact, walk you through each fix, hand you text to paste, and score the profile out of 100. Make the change, re-run the analysis, and the profile is scraped again so the score moves.
What does it cost?
1,990 TRY a month. There is no free tier and no ladder. The platform is invite-only; once your invite is approved you request a subscription from the dashboard, a payment link arrives by email, and your account opens once the payment is confirmed.
Where does my data live?
In your own Railway project, in your own PostgreSQL database. LinkedIn access tokens are stored encrypted with AES-256-GCM. You can permanently delete your account and all of your data from the settings page.
What is LinkedIn’s feed actually ranking for?
What does the LinkedIn algorithm optimise for in 2026?
According to LinkedIn’s February 2026 paper, the production ranker predicts two tasks: long dwell — time on a post above a post-type-dependent threshold — and contribution, meaning a like, comment or reshare.
Does dwell time really affect ranking?
Yes. LinkedIn announced it in a 2020 engineering post and it appears as a model objective in both the 2024 and 2026 papers. The threshold in seconds has never been disclosed.
Do I need engagement in the first 60 minutes?
There is no LinkedIn source for this. No published paper or company post describes a golden-hour window; the advice circulates entirely within marketing content citing itself.
How many hours does a LinkedIn post live?
LinkedIn has never given a number. The only on-record comment says content moves quickly through the newsfeed but can resurface later via suggested posts. The 24- or 48-hour lifespan claim has no source.
Are engagement-bait posts penalised?
LinkedIn stated in 2022 that it would not promote content asking for reactions purely to boost reach, and in 2023 described classifiers that monitor engagement velocity.
Does putting a link in a LinkedIn post really cut your reach?
Does LinkedIn penalise posts with links?
LinkedIn has never said so. Independent measurements conflict: a 900,000-post study found a 26.5% reach deficit, while a 224,000-post study found posts with three or more links reached 140% more people.
Is the “links cost 60% of reach” figure real?
It has no traceable source. It appears uncited across dozens of articles, and an independent audit tracing it concluded it leads back to no identifiable origin.
Does putting the link in the first comment help?
Unknown. One of the two datasets large enough to answer states it did not measure this; the other does not disclose its subset size. LinkedIn has never commented on the practice.
Is there a difference between a preview card and a plain URL?
In a 567,000-post measurement, yes: posts with an attached preview card had a median of 414 impressions, while posts with a plain in-body URL slightly outperformed posts with no link. The effect may come from the article format rather than the URL itself.
So what should I do?
Write the post so it stands without the link, try a plain in-body URL instead of an attached card, and read the result from your own account data rather than published averages.
Where does ChatGPT actually read from when it answers?
Does ChatGPT still use Bing?
Measurements indicate the dependency is effectively gone. A 2026 study found ChatGPT using its own search index and scraped Google results, with roughly 1.5% overlap with Bing’s top 20. OpenAI has made no official statement about a separation.
Which bot do I need to allow to appear in ChatGPT answers?
OAI-SearchBot. OpenAI recommends allowing it explicitly. GPTBot is for training; blocking it does not directly remove you from search answers.
I allowed it in robots.txt but still don’t appear. Why?
Most often the CDN or firewall. OpenAI separately requires that your host accepts traffic from its published crawler IP addresses, and AI-crawler blocking ships on by default in some CDN plans.
Should I add an llms.txt file?
There is no measured evidence it helps. Google’s John Mueller stated no AI system uses it, and no vendor mentions it in crawler documentation. It has not been shown harmful either — its benefit is simply unmeasured.
Does schema.org markup increase AI citations?
In a controlled study of 1,885 pages that already received citations, the increase was indistinguishable from zero. The study did not test whether markup helps an unseen page get discovered. Treat it as hygiene, not leverage.
How often should you actually post on LinkedIn?
How many times a week should I post on LinkedIn?
There is no reliable single number. The best-designed study, covering two million posts, finds no per-post reach decline up to 11 posts a week; a 1.3-million-post analysis says the optimal range has dropped to two to four. Both cannot be right.
Will posting more reduce my reach?
No study with a disclosed sample and method shows this. An analysis comparing accounts against their own averages across two million posts found per-post impressions rising with frequency.
Is consistency more important than frequency?
The largest measured effect is in consistency. Across 100,000 users over 26 weeks, those posting in 20 or more weeks saw more than five times the engagement per post of sporadic posters — though that study is cross-platform, not LinkedIn-only.
How long should I wait between posts?
There is no measured rule. Neither a study nor a LinkedIn statement supports the commonly repeated 18- or 24-hour gap.
How many people does a typical LinkedIn post reach?
Across 476,000 personal-profile posts, the median was about 840 impressions. Half of all posts fall below that; 15,000 impressions marks the 94th percentile.
Is collecting LinkedIn profile data lawful under Turkish data protection law?
LinkedIn profiles are public — can they be collected freely in Turkey?
No. The Board’s settled reading is that data made public may only be processed in a way connected to and limited by the purpose of disclosure. A company that collected work e-mail addresses from search engines for marketing was fined 150,000 TRY on that basis.
I only collected the data and haven’t used it. Is that processing?
Yes. Article 3 of Law 6698 explicitly lists obtaining, recording and storing among processing operations. Collection alone is processing.
Is there a Turkish decision specifically on web scraping?
No. The published principle decisions contain nothing on scraping or bulk social-media collection, and no published summary names LinkedIn. The applicable rules are assembled by analogy from the statute, the 2020 announcement on public disclosure, and the 2019 and 2022 decisions.
Didn’t the hiQ case establish that scraping is legal?
No. The appellate ruling concerned computer-misuse law only. The case ended with a 500,000 dollar consent judgment against hiQ for breaching LinkedIn’s User Agreement and destruction of the data, and it did not address data protection law at all.
Does recruiting fall in the same category as sales prospecting?
That is unresolved in published decisions. A LinkedIn profile is created for career visibility, and whether recruiting outreach falls inside that purpose is arguable. Sales prospecting, on the 2022 reasoning, most likely falls outside it.
What a headhunter actually looks at when they search LinkedIn
How much does my LinkedIn headline matter for search?
The headline is not read by LinkedIn Recruiter’s title filter; it appears only in keyword search, and the result card truncates it at roughly sixty characters. It matters, but less than the title field in your experience section.
Should I change my internal job title?
Put the market equivalent in the title field of your experience entry. You can explain the internal naming in the description. The field that makes you findable is the title field.
How many skills should I list?
At least five, and pick them from LinkedIn’s dropdown rather than typing free text — search only recognises standardised skills. Pin the three to five most relevant to your target title.
Can I improve my ranking in recruiter search?
There is no fixed rank. LinkedIn’s published work describes the ranking model as personalised to each recruiter, so the same query returns different orders for different people. What you can control is entering the list — the filter fields.
Do profile photos or connection counts affect search?
No LinkedIn statement or independent measurement could be found on this. Advice quoting precise figures has no source. The fields with measurable effect are title, skills and location.
How often should I update my profile?
Update the filter fields only when they genuinely change. Once a year, check whether the market’s name for your target role has shifted.
Do you show up when someone asks ChatGPT about you?
Can ChatGPT read my LinkedIn profile?
Not directly. LinkedIn’s robots.txt does not permit GPTBot or OAI-SearchBot. Where LinkedIn content appears in AI answers, it arrives indirectly through Google’s index.
What should I actually do to show up in AI answers?
First publish a crawlable page on your own domain with no login wall. Then get independent publications to write about you — that is where the large majority of citations come from. Third, put original data and cited sources into what you write.
Should I add an llms.txt file?
No. Google’s John Mueller said no AI system uses it, and server-log evidence agrees: zero requests to that file from the major AI crawlers.
How do I measure my own AI visibility?
Ask three types of question on each engine at least ten times. Count how often you appear, not where you rank. Measure your languages separately — source overlap between them is small.
What should I put on my own site?
A one-sentence definition, a few verifiable concrete records, links to your other profiles, and plain answers to common questions. Since query-answer matching is the highest-scoring measured factor, shaping headings like the questions people type helps.
I have a page but never appear in AI answers. Why?
Two likely reasons. One is technical: the search crawler may not be reaching the site, usually blocked by the content delivery network rather than robots.txt. The other, more common one is that nothing independent has been published about you — roughly 84% of citations come from editorial content.
How often should I measure visibility?
Quarterly is enough. Use the same questions, the same number of repetitions, and a fresh session each time; the benchmark is your own previous percentage.
Why no tool can post your LinkedIn comments for you
How do tools that promise automatic comment replies work?
Since reading comments through the API is not possible, they use browser automation or third-party scraping. Both breach LinkedIn’s User Agreement, and the risk sits with the user’s account.
Can a tool send connection requests for me?
Not through the API — that endpoint is restricted to approved partners with no self-serve route. Browser automation is prohibited outright by section 8.2 of the User Agreement.
So what can a tool legitimately do?
Publish posts, write comments, draft replies for your approval, and apply for analytics permissions. Naming the gap on the reading side and closing it with one click from you is better than hiding it.
What if a tool asks for my LinkedIn password?
Do not give it. Official integrations ask for OAuth permission, not a password. A tool asking for a password or session cookie is not acting on your behalf but in your place, and the account risk becomes entirely yours.
What happens to my account with browser-automation tools?
The consequence is usually staged: temporary limits on some actions, then feature restrictions, then suspension. The vendor is not party to any of those steps.
Can a LinkedIn-approved partner tool auto-reply to comments?
No. LinkedIn’s conduct rules for approved partners require the message to be tied to a member action, the member to be able to edit content before sending, and no scheduled automated sending.