One of my connections mentioned in a post on LinkedIn that since he turned off Premium, his posts have been getting less engagement. Since LinkedIn offers a month for free, and I have been trying to promote my recent book, I figured I would try my free month trial and see how many more views I could get. (Here I am not worried about Premium for applying to new jobs, it is possible it is totally worth it for that, I was not applying to jobs in this test so I do not know.)

Long story short, LinkedIn Premium does not appear to promote my material at all above the baseline.

Post Views

In a sample of 30 posts the month before I turned on Premium (turned on 3/24 in the evening, turned off 4/22 in the morning), my posts had an average of 3600 views (with a standard deviation of 7000, median 1400). Post-Premium, I had 23 posts, and the views were on average 2200 (SD 2900, median 900). Here is the full table of posts and links (Premium=1 means it was posted when my Premium subscription was turned on):

| Premium | Views | URL |
| ----:|----- :|:----- |
| 0    | 3659  | https://www.linkedin.com/posts/andrew-wheeler-46134849_llms-have-transformed-the-data-science-industry-activity-7426975341572984832-HGTA |
| 0    | 2526  | https://www.linkedin.com/posts/andrew-wheeler-46134849_no-guarantees-but-i-am-going-to-try-to-start-activity-7428418993553846272-vdjA    |
| 0    | 2290  | https://www.linkedin.com/posts/andrew-wheeler-46134849_much-of-the-hype-around-claude-code-is-having-activity-7428781380391567360-zkXr   |
| 0    | 545   | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-benefits-of-my-epub-version-of-activity-7429143771109302272-_V6T       |
| 0    | 1454  | https://www.linkedin.com/posts/andrew-wheeler-46134849_claude-code-has-the-ability-to-create-hooks-activity-7429506167527088128-01Um     |
| 0    | 1326  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-prompting-flows-i-find-convenient-activity-7429868558794436609-SDgG    |
| 0    | 1278  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-main-focuses-in-the-book-is-not-activity-7430230940444057600-pqK-      |
| 0    | 5988  | https://www.linkedin.com/posts/andrew-wheeler-46134849_while-skills-in-claude-code-are-all-the-rage-activity-7430955707358818304-iqjb    |
| 0    | 1726  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-mistakes-i-see-with-agent-based-activity-7431318102212100096-rI_W      |
| 0    | 1172  | https://www.linkedin.com/posts/andrew-wheeler-46134849_from-my-experience-as-an-educator-when-presenting-activity-7431680485585580032-8h7B    |
| 0    | 1360  | https://www.linkedin.com/posts/andrew-wheeler-46134849_although-the-llm-tools-are-currently-focused-activity-7432042882770944000-AfAG    |
| 0    | 5304  | https://www.linkedin.com/posts/andrew-wheeler-46134849_when-i-was-a-professor-at-ut-dallas-i-sat-activity-7432405268874817536-qrAI       |
| 0    | 30732 | https://www.linkedin.com/posts/andrew-wheeler-46134849_i-know-a-few-stats-folks-in-my-network-that-activity-7432767666781679617-HXnk |
| 0    | 1003  | https://www.linkedin.com/posts/andrew-wheeler-46134849_claude-code-does-not-have-an-image-model-activity-7433492422111776768-2dqY        |
| 0    | 884   | https://www.linkedin.com/posts/andrew-wheeler-46134849_while-i-have-a-section-in-the-book-devoted-activity-7433854815862013952-FIl-      |
| 0    | 888   | https://www.linkedin.com/posts/andrew-wheeler-46134849_in-the-book-i-have-a-dedicated-chapter-on-activity-7434217215450681344-hk3E       |
| 0    | 1868  | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-llm-book-is-compiled-using-quarto-so-activity-7434579595095580673-RHdx        |
| 0    | 807   | https://www.linkedin.com/posts/andrew-wheeler-46134849_llms-for-mortals-how-to-view-the-epub-activity-7434945455073169409-bNDu           |
| 0    | 1243  | https://www.linkedin.com/posts/andrew-wheeler-46134849_section-on-using-gliner-for-ner-activity-7435304382931513344-NPPC                 |
| 0    | 1745  | https://www.linkedin.com/posts/andrew-wheeler-46134849_my-first-book-data-science-for-crime-analysis-activity-7436029137695416320-aRmr   |
| 0    | 914   | https://www.linkedin.com/posts/andrew-wheeler-46134849_so-the-new-book-large-language-models-for-activity-7436376426100199424-1g8E       |
| 0    | 1593  | https://www.linkedin.com/posts/andrew-wheeler-46134849_agentic-coding-apps-like-claude-code-and-activity-7436738847054512128-qiXz        |
| 0    | 3415  | https://www.linkedin.com/posts/andrew-wheeler-46134849_many-people-are-turned-off-by-ai-writing-activity-7437101213717909504-QjOo        |
| 0    | 928   | https://www.linkedin.com/posts/andrew-wheeler-46134849_pretty-much-every-day-there-is-a-new-prompt-activity-7437463609401794560-50H-     |
| 0    | 2185  | https://www.linkedin.com/posts/andrew-wheeler-46134849_much-of-the-hype-around-skills-is-imo-people-activity-7437826000958337024-t6i3    |
| 0    | 800   | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-benefits-of-my-llm-for-mortals-activity-7438550758763098112-hQUw |
| 0    | 870   | https://www.linkedin.com/posts/andrew-wheeler-46134849_large-language-models-for-mortals-preview-activity-7438913140639207424-tAli       |
| 0    | 1948  | https://www.linkedin.com/posts/andrew-wheeler-46134849_new-blog-post-using-claude-code-to-help-activity-7441087469388861440-TCKq         |
| 0    | 1160  | https://www.linkedin.com/posts/andrew-wheeler-46134849_given-all-the-rage-with-generative-ai-and-activity-7441449857480933377-uPFw       |
| 0    | 27842 | https://www.linkedin.com/posts/andrew-wheeler-46134849_stop-teaching-r-teach-python-when-i-was-activity-7441812266938826753-DywF         |
| 1    | 526   | https://www.linkedin.com/posts/andrew-wheeler-46134849_forecasting-the-future-is-difficult-especially-activity-7442537064803368960-qsVO  |
| 1    | 13096 | https://www.linkedin.com/posts/andrew-wheeler-46134849_when-using-llms-to-do-structured-data-extraction-activity-7442899426471407617-CpZz |
| 1    | 2394  | https://www.linkedin.com/posts/andrew-wheeler-46134849_ive-spoken-with-many-people-who-are-concerned-activity-7443039100477145090-v_3i   |
| 1    | 646   | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-main-audience-my-book-large-language-activity-7443261810511757312-R2Jc        |
| 1    | 3030  | https://www.linkedin.com/posts/andrew-wheeler-46134849_for-the-folks-that-were-not-happy-with-my-activity-7443401497444409344-q38H       |
| 1    | 437   | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-current-capabilities-of-googles-activity-7443624184120754176-YHbw      |
| 1    | 5275  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-error-i-am-seeing-devs-continually-make-activity-7443986571650973696-TZ-K     |
| 1    | 3815  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-biggest-issues-with-using-generative-activity-7444348969100664832-pX0v |
| 1    | 738   | https://www.linkedin.com/posts/andrew-wheeler-46134849_reports-of-rags-demise-are-overstated-activity-7444711358421491712-6BJu           |
| 1    | 425   | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-recent-litellm-distribution-attack-highlights-activity-7445073747968999424-NEmn    |
| 1    | 3752  | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-responses-to-me-writing-the-book-activity-7445436130080186369-7Fvx     |
| 1    | 1670  | https://www.linkedin.com/posts/andrew-wheeler-46134849_professors-that-follow-me-i-am-happy-to-activity-7446160904217407488-CXcZ         |
| 1    | 918   | https://www.linkedin.com/posts/andrew-wheeler-46134849_i-have-used-claude-code-the-longest-probably-activity-7447248077435920385-Ym8A    |
| 1    | 876   | https://www.linkedin.com/posts/andrew-wheeler-46134849_gio-has-a-new-post-out-on-examining-confidence-activity-7448697613513740289-hObZ  |
| 1    | 1959  | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-mythos-technical-blog-post-on-its-cybersecurity-activity-7449059395071709184-UJOC  |
| 1    | 2201  | https://www.linkedin.com/posts/andrew-wheeler-46134849_for-folks-that-use-jupyter-notebooks-one-activity-7449422388691243008-Twxn        |
| 1    | 626   | https://www.linkedin.com/posts/andrew-wheeler-46134849_one-of-the-recommendations-i-have-in-the-activity-7450147165781426177-vceO        |
| 1    | 333   | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-term-agent-is-almost-always-used-as-activity-7450509557677625344-42Ze         |
| 1    | 5787  | https://www.linkedin.com/posts/andrew-wheeler-46134849_agent-based-systems-require-bad-python-code-activity-7450871941458190336-gzSl     |
| 1    | 468   | https://www.linkedin.com/posts/andrew-wheeler-46134849_broadly-there-are-two-types-of-agent-based-activity-7451234329390678016-ZU1h      |
| 1    | 1267  | https://www.linkedin.com/posts/andrew-wheeler-46134849_i-get-periodically-asked-what-is-the-best-activity-7451596716354707456-aFMF       |
| 1    | 346   | https://www.linkedin.com/posts/andrew-wheeler-46134849_the-saying-a-picture-is-worth-a-1000-words-activity-7451959111132299264-Tk7o      |
| 1    | 480   | https://www.linkedin.com/posts/andrew-wheeler-46134849_it-is-important-to-have-independent-benchmark-activity-7452318150701953024-ftcM        |

I would have expected a multiplier (e.g. typically 3k views, now you have 6k or 9k views per post). So you could nitpick that I have differential timing for the posts, and the pre-premium posts have some contamination (if they promoted my older posts when I activated Premium). But those are not large enough to make a difference in my findings relative to what I expected.

The posts are quite comparable in content, mostly focused on my book and LLMs. It is possible my audience is oversaturated with that content, but I think it is just as likely that LinkedIn Premium doesn’t really promote your work to any substantive extent. (I have additionally obtained more followers in this period, so that should bias the results to have more views, not less.) At least here there is no evidence I should continue to pay $20 a month to increase my reach on LinkedIn.

Posts are bursty, and in the end I have very little ability to forecast what will or will not be popular. In the pre-period, my most popular post was on a blog post I did on log-probabilities (30k views). I definitely try to post more technical stuff on LinkedIn than the typical social media influencer, so that limits the reach.

I also had a rage-bait post on professors should teach python and not R with just under 30k views. (That was a bit of social media manipulation – have a controversial opinion that divides people, you get a bunch of thumbs up and a bunch of comments.) I do not have that many potential rage-bait post topics!

In addition to this, I also did the month for free for LinkedIn Premium for my business Crime De-Coder page. The same with my business page, I did not see any increased views, increased followers, etc.

Profile Views

Although I have not seen LinkedIn explicitly say Premium boosts your posts (besides actually paying for advertising), I have seen LinkedIn explicitly advertise that Premium profiles get more views:

So how do profile views look? I did get more the week I signed up, but it was trending upward previously, and reverted to the trend after week one anyway. (A few days short, I cannot access the chart week by week since turning off Premium.)

For a bit of background, I spent most of my time posting on my LinkedIn business Crime De-Coder page, and only posted on my personal page maybe once or twice a month. But since publishing LLMs for Mortals (in February of 2026), I have posted more on my personal. Which you can see increased my profile views before I signed up for Premium.

Likely the past additional profile views are for that rage-bait Python vs R post that was popular, not due to anything Premium did.

This appears to be extremely misleading advertising on LinkedIn’s part. If they just look at Premium vs not, it is likely Premium users are more active. This should just say the explicit “boost” profile views get, like ranked higher in searches.

$100 ad credit

With premium, you get $100 ad credit for posts a month. I used this to boost my original LLM for Mortals launch post, which was stale at that point and not accumulating any additional views.

The metrics on the post were as LinkedIn said they would be. Despite having 80+ likes when I first created it, the post only had 3700 views. Spending $100 on the credits got me an additional ~3500 views and supposedly ~50 additional website clicks. (I am confused how this is calculated, as I can see the actual link in the post was clicked fewer than 10 additional times with the campaign.)

I knew going in that adverts on LinkedIn are not a net benefit given my book purchase conversion rates. What I will call “high trust” referrals, I have something like a 1/100 purchase rate for the book. For other mediums, it is more like 1/1000. As far as I can tell, these seem pretty typical for a higher dollar value book purchase ($50+).

I have debated on setting the purchase price for the epub to much lower. $50 is in line with current offerings from O’Reilly, and in my informal demand curve tests is where I think it should be. But I don’t think any realistic conversion rate would make LinkedIn advertising make sense for my book.

For reference for influencers though, this gives a rough estimate comparable to LinkedIn’s direct advertising. Basically my average post is worth $100 according to LinkedIn. I only have around 3k followers currently on LinkedIn, so I imagine folks with followings 10x that can likely do direct advertisements to their audiences for more like $1k and up.

Wrap Up

I still think LinkedIn is the best social media site currently to promote my work and business. It is not just about the raw view counts, but also about conversion to people buying my book or reaching out for additional consulting gigs.

I will continue to use LinkedIn for this, but paying for a Premium LinkedIn account does not appear to be worth it for these reasons. Even if the views were increased, it is possible that they are not good connections for these end goals.

There are additional things you get with Premium (can send cold messages to people you are not connected to, supposedly higher priority when applying to jobs). Those are maybe worth the $20 a month for some people. But focusing on what LinkedIn advertises for “boosting” your posts and profile, I did not personally see any evidence that would justify spending even $1 a month for the Premium features.