Content @Mayple, 0 to 50k/mo visitors, I want to be your next podcast guest 😎

I'm an experienced well-rounded marketing manager. I grew a B2C home decor company to over $1M in
annual revenue. I worked for several SAAS B2B companies managing email marketing, content marketing, and affiliates. 

I've created and managed blogs, affiliate programs, referral programs, influencers, cold email
outreach and social media. I have a deep passion for CRO, A/B testing and UI/UX. I'm a humble self-starter, quick learner.

Here is my content writing portfolio - https://clearvoice.com/cv/BenKazinik 🙌
Email me - ben.kazinik@mayple.com

Ben Kazinik

Nemanja Zivkovic

Volcano of creativity | I fix wasted marketing spend when the pipeline’s stuck | B2B systems where brand, sales & growth collide into one engine | Biblical (like Oasis)

You have a great product, but your message is messy. Your team is busy, but the pipeline is stuck. You’ve tried every "growth hack" in the book, yet you’re still the one doing the heavy lifting to close every deal.

I’m Nemanja, and I help you build a business that runs like a system.

I don’t just "do marketing." I did, as marketing goal = busines goals. But for the last few years I do much more. I help you align your brand, your sales, and your operations so they all pull in the same direction. I believe that in a world of AI-generated noise, being human is your only real competitive advantage.

Why work with me? 

I think like an owner. I’ve done this for 120+ companies and 80+ startups, helping them stop chasing clicks and start building equity.

If we're talking about real results, my own agency, Funky Marketing, closed 65 inbound B2B deals in 18 months in the very 1st year with zero ads (nothing against ads I just use them the right way - to scale what's already working without them).

I live in Serbia, but even my 1st job (ugh, a long time ago - check my work experience below) was for a Canadian company working on North American markets. Global perspective from the start. 

I’ve scaled systems for brands like Microsoft, Wix, Deloitte, and Generali across the US, Europe, and the Middle East.

In time, I learned that I love to have skin in the game, so I often partner with serious companies through revenue share or equity. If you don’t scale, I don’t win. And as an ex-basketball player, I hate losing.

Published the 1st book in the Serbian language talking about LinkedIn (soon in English, too!).

Pronounced as one of 10 Marketing Voices to follow on LinkedIn, by LinkedIn! Hey, that was huge!

Built a few communities, the biggest one with 600+ people online and 350 coming to offline events every two weeks. Proud of that and currently building a global one (pssst, more info soon).

When I’m off the clock, I’m probably listening to good music (search dilimanac on Spotify), wandering in the woods, traveling, or getting lost in a comic book or an epic fantasy world. 

I also spend my time mentoring founders at the Founder Institute and BioSense because I believe in giving back to the ecosystem that helped me grow.

So what do I actually do?

I clarify your message so buyers "get it" instantly.
I build systems that keep working even when you aren't. 
I turn your business into a revenue engine that compounds over time.

No interns, no buzzwords, no fluff. 
Senior operator who cares about your bottom line as much as you do.

Ready to simplify? Let's talk.

CEO & Founder

We create demand generation programs primarily for B2B tech-based companies in the growth phase and enable them to generate consistent revenue growth, increase their deal size and shorten their sales cycles using inbound and content strategies.

People think marketing is easy. Hint: it’s not.
People see marketing as black and white, like sales. Hint: it’s not.
Marketing is very very gray. And oftentimes counterintuitive, which is why so many people don’t get it.

We keep delivering on what we promise.

LinkedIn je upravo omogućio nešto što bi svaki ozbiljan B2B profil trebalo da pogleda odmah.

Sad možeš da vidiš tačno koliko impresija - i koliko klikova - dolazi iz koje sekcije tvog profila.

Imam dva dashboarda ispred sebe. Evo šta kažu.

Dashboard 1: Impresije po sekciji

About - 28%
Featured - 24%
Activity - 23%

Tri sekcije. 76% svega što se dešava na mom profilu.

Experience? 4.9%.
Skills? 1.4%.
Recommendations? Ispod 1%.

Dashboard 2: Klikovi po sekciji

Intro - 45.5%
Activity - 44.5%

Dve sekcije. 90% svih klikova.

Experience - 9.8%.
Sve ostalo - šum.

Godinama su LinkedIn “eksperti” govorili da optimizuješ Experience sekciju. Da dodaješ Skills. Da skupljаš Recommendations.

Niko nije merio. Nije ni moglo da se meri.

Sad imamo podatke.

I podaci kažu: ljudi ne dolaze na tvoj LinkedIn profil da čitaju CV.

Dolaze kao na sales page.

Skeniraju headline. Proveravaju šta si postovao/la. Za nekoliko sekundi odlučuju da li si vredan/na razgovora.

Tvoj About je headline i value proposition.
Tvoj Featured je ponuda ili dokaz rada.
Tvoja Activity je razlog zbog kojeg su uopšte kliknuli.

Experience sekcija? Sitno slovo.
Skills? Footer.
Niko ne klikće footer.

Prosečno vreme pregleda mog profila ove nedelje: 23 minuta.

To nije slučajni scroll. To je neko ko procenjuje da li da počne razgovor.

Ako trošiš vreme na poliranje dna svog sales page-a dok su ti headline i sadržaj prepušteni slučaju - podaci ti govore nešto.

Popravi vrh. 
That’s where revenue lives.

No alternative text description for this image

LinkedIn je promenio feed algoritam. Ne "unapredili smo rangiranje." Kompletna rearchitektura sistema koji odlučuje šta vidi 1,3 milijardi profesionalaca.

Pročitao sam engineering blog od prve do poslednje linije. Evo šta stvarno znači.

Stari sistem je bio mozaik. Sadržaj je dolazio iz pet odvojenih izvora: hronološki index mreže, trending postovi, collaborative filtering, industry trending, embedding sistemi. Svaki sa sopstvenom infrastrukturom i svaki optimizovan odvojeno.

Radilo je, ali holistička optimizacija je bila gotovo nemoguća. I ranking model je tretirao svaki post nezavisno, bez konteksta šta si čitao juče ili prošle nedelje.

Šta su sad izgradili 👇

Unified LLM retrieval sistem koji razume semantičko značenje, ne ključne reči. 

Ako engaguješ postove o "small modular reactors" a tvoj title kaže "electrical engineer," novi sistem povezuje te tačke jer LLM donosi world knowledge iz pre-training korpusa.

Zatim Generative Recommender (GR) model koji tretira tvoju interakciju kao sekvencu - "profesionalnu priču ispričanu kroz postove sa kojima si se engageovao tokom vremena." 

Obrađuje 1.000+ istorijskih interakcija da razume trajektoriju, ne samo preference.

Engaguješ ML sadržaj u ponedeljak, distributed systems u utorak. Stari model je video dve nezavisne odluke. Novi razume kuda putanja ide.

Engineering detalj koji je prošao ispod radara 👇 

Inicijalno su ubacivali sirove engagement count-ove direktno u promptove. Post sa 12.345 pogleda izgledao je kao "views:12345" - model ga tretira kao bilo koji text token.

Korelacija između popularnosti i similarity scores? 
-0.004. Praktično nula.

Rešenje su našli u percentile bucket-i omotani u special tokens. 

"views:12345" postaje <view_percentile>71</view_percentile>. Jedna promena. Korelacija skočila 30x. Recall@10 poboljšan 15%.

Sirovi brojevi su šum i znače mnogo jer ne registruju istoriju. 
Strukturirana reprezentacija je signal.

Šta ovo znači ako praviš sadržaj na LinkedInu?

Tematska dubina > tematska širina. 

Sistem modelira trajektoriju tvojih interesovanja. Konzistentna ekspertiza u specifičnoj oblasti daje algoritmu jasnu sliku ko si i kome da te prikaže.

Engagement bait je arhitekturalno kažnjen, ugrađeno u način na koji ranking model evaluira sekvence ponašanja. Nije još isčišćen iz feeda, ali očekujemo da se to desi.

Doseg van mreže postaje realan. 

Semantic matching znači da relevantan sadržaj može da nađe profesionalce koji te nikad nisu videli, ako su im interesovanja semantički alignovana sa tvojim temama.

Platforma više ne nagrađuje distribuciju. Nagrađuje semantičku relevantnost i konzistentnu ekspertizu.

Ako gradiš iskren autoritet u specifičnoj oblasti, ovaj update radi za tebe.

Ako pokušavaš da izigraš algoritam, playbook se upravo promenio da ti to onemogući.

Jeste li primetili promene u performansama vaših postova u poslednjih mesec-dva?

LinkedIn just released something every B2B seller should look at right now!

You can now see exactly how many impressions each profile section is generating.

I just found this on my profile. New metrics!

Here’s mine 👇

About - 28%
Featured - 24%
Activity - 23%

That’s 76% of everything happening on my profile! 

Three sections.

Experience? 4.9%.
Skills? 1.4%.
Recommendations? Under 1%.

For years, LinkedIn “experts” told you to optimize your Experience section. 

Add more Skills. 
Collect Recommendations.

Nobody measured it!

Now we have data.

And the data says: people don’t visit your LinkedIn to read your CV.

They come because of what you write (Activity), who you are (About), and what you show (Featured).

The rest is background check. 
Not a sales driver.

Here’s what this actually means ⤵️

Your LinkedIn profile is a sales page.

And like any sales page - the headline, the hero section, and the social proof at the top do all the work. 

Nobody converts because of the fine print at the bottom!

I blame LinkedIn for adding redommendations at the bottom, but it is what it is.

About = your headline and value prop.
Featured = your offer or proof of work.
Activity = the reason they showed up in the first place.

If you’re spending time polishing sections that drive less than 5% of your profile impressions - you’re optimizing the footer of a sales page nobody scrolls to.

Fix the top. 
That’s where the revenue lives.

Every B2B brand is about to start posting on LinkedIn.

Most will fail. 

Not because of bad content. Ok? 
Because of bad positioning.

Heike Young made an excellent point about the LinkedIn/LLM opportunity - EGC, exec voices, paid creators, and customers posting consistently. 

I know she knows what she's doing and talking about, as the same day I did a workshop with the Microsoft SQL product leaders about how to use their positions and become evangelists for their products and promote their features, she wrote a post saying the exact same things I said to them the same day, with the difference that I was in Serbia and she was in LA. I was hired externally, she was in-house. 😁 

Anyways, the thing is that the company needs to act like a symphony, and not a single instrument. Are we clear on that?

But a symphony still needs a score.
Most companies will skip that step. 

They'll activate employees, brief the executives, set a posting cadence, and start amplifying. 

But amplifying what, exactly?!

I call it Surround Sound. When your CEO, your sales team, your subject matter experts, and your customers are all saying the same true thing from different angles, to different networks, in different formats, with different proof points, that's when LinkedIn authority compounds.

Surround sound without a clear message doesn't build authority. It just distributes the vagueness further.

In 120+ positioning reviews across 15+ industries, the pattern is always the same. Companies activate employees before they've decided what those employees should stand for. 

Everyone posts. 
Nobody owns anything. 

The company becomes "active on LinkedIn" - not known for anything specific.

Here's what actually happens in LLM search when your brand is vague:

1. The AI finds your content
2. Can't place you clearly
3. Cites someone else who said the same thing with more conviction and consistency.

Before you activate anyone - one question 👇 

What is the one true thing your company stands for that your closest competitor would never say?

Answer that first! 
Then build the surround sound around it.

Think! If you don't, here's what will happen.

Activate everyone. 
Brief no one on what to actually say. 

That's not a symphony. 
That's expensive chaos or however we call it.

A choir is only as powerful as the song they all know by heart.

LinkedIn rebuilt its feed from scratch, a full architectural overhaul of the system that decides what 1.3 billion professionals see every time they open the app.

Hristo Danchev published the engineering blog last week. I read every line. Here's what actually matters.

The old system was a patchwork. 5 separate retrieval sources (network activity, trending posts, collaborative filtering, industry content, embedding systems) each with its own infrastructure, each optimized independently.

It worked, but holistic optimization was nearly impossible. The ranking model scored every post in isolation with no context for what you read yesterday, last week, or last month.

What they built instead 👇

A unified LLM-based retrieval system that understands semantic meaning, not keyword matches.

If you engage with "small modular reactors" posts but your title says "electrical engineer," the new system connects those dots because the underlying LLM knows these topics are related from its pretraining. The old system didn't.

Then a Generative Recommender (GR) model that treats your entire interaction history as an ordered sequence. It processes 1,000+ historical interactions through a transformer architecture to understand trajectory, not just preference.

Monday you engaged with ML content. Tuesday, distributed systems. The old model saw two independent decisions. The new one understands where you're heading.

They initially passed raw engagement counts directly into prompts. A post with 12,345 views appeared as "views:12345" - treated like any other text token.

The correlation between popularity and embedding similarity? 
-0.004. Basically zero.

So they converted counts into percentile buckets wrapped in special tokens. "views:12345" became <view_percentile>71</view_percentile>. 

One change. Correlation jumped 30x. Recall@10 improved 15%.

Raw numbers are noise. 
Structured representation is signal.

What does this mean if you create content on LinkedIn?

Topical depth now beats topical breadth. Consistent expertise in a specific domain gives the algorithm a clear signal of who you are and who should see you.

Engagement bait is architecturally penalized, baked into how the ranking model evaluates sequences of behavior.

Reaching beyond your network becomes real. Semantic matching means genuinely relevant content can find professionals who've never seen you - if your topics align with their interests.

Cold-start is structurally solved. New members get relevant feeds immediately based on profile information alone, before any engagement history accumulates.

The platform is moving from rewarding distribution tactics to rewarding semantic relevance and consistent expertise.

If you've been building genuine authority in a specific domain, this update works in your favor.

If you've been playing the algorithm, the playbook just changed underneath you.

Most people publishing "thought leadership" aren't leading thought. They're performing it. The difference is harder to see than it sounds.

Thought performance is optimized. It looks bold. 

It's structured to seem provocative. But read it twice and there's nothing that could genuinely offend, challenge, or change anyone's mind.

"Hot take: execution matters more than strategy." 

"The truth nobody talks about: relationships drive revenue." 

"Controversial opinion: culture starts at the top."

These are safe ideas wearing the costume of a courageous take.

Real thought leadership costs you something. 

Think about it. It's the post that loses you followers from one camp because you said what you actually believe. 

It's the conference talk where you publicly disagree with a sponsor's premise. 

It's the client conversation where you say "your strategy is wrong" - and mean it.

I've met maybe a dozen actual thought leaders in B2B in 7 years. The rest are content machines optimized for reach, engagement, and the dopamine hit of an impressive impressions number.

Nothing wrong with content, but don't confuse it with leadership.

Leadership means you were willing to be wrong in public. 
Performance means you engineered it so you never had to be.

taplio

The Battle

Ben Kazinik

Nemanja Zivkovic

Followers

Posts

Engagements

Likes

Performance ScoreNemanja Zivkovic's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Ben Kazinik

Nemanja Zivkovic

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Nemanja Zivkovic's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreNemanja Zivkovic's LinkedIn Performance