AJ Kumar

Behind the social media growth of @neilpatel @nontoxicdad, @bignikbh & more. 15 yrs building founder-led brands for experts.
↓ Follow to build yours.

[AJ Kumar @ajkumar]

Last day of May. 29 days into the daily posting experiment across five platforms.

Some of the posts hit. Some didn't. The metrics matter less than I thought they would on day one. What matters more is something I didn't expect.

Most of the value isn't in any individual post. It's in the system that makes posting unavoidable. Once you've shown up for 29 days in a row, day 30 isn't a question. The thing that compounds is the habit, not the post. The willpower stops being the variable.

Three things I'm carrying into June.

One. Long-form single posts on X are doing more work than I expected. That's the real anchor format for this audience.

Two. Threads is the platform I most underestimated. Replies are everything there. The conversation-first algorithm is real.

Three. Substack Notes that confess outperform Substack Notes that frame. The audience came for voice, not for tactics.

June. Same cadence. Different concepts. Better hooks. Less polish. More substance.

See you tomorrow.

[AJ Kumar @ajkumar]

Sunday morning. Joel Osteen built an audience of millions because he sounds like a friend, not a preacher.

He doesn't lecture from a position of authority. He arrives at eye level. The intonation is conversational. The vocabulary is plain. The cadence makes you feel like he's talking to you specifically, not to the room.

The Sunday morning preacher of the algorithm era is your favorite founder.

Same parasocial trust mechanism. Different stage. The brain doesn't draw a hard line between a real friend and a digital one. When content arrives at the right moment in a familiar voice, the same trust circuits fire. The same dopamine release. The same sense of "this person understands me."

Researchers call this parasocial connection. Voice familiarity, repeated exposure, conversational intimacy produce loyalty similar to in-person friendships.

Most founders try to sound credible. They sound credentialed and distant.

The ones who win sound like a friend who happens to be credible. Different posture. Same expertise.

[AJ Kumar @ajkumar]

21 days into posting daily across five platforms. Some honest observations.

Substack Notes. The ones that performed best were the most personal, not the most strategic. Confession beat framework every time. The audience is reading for voice, not for tactics.

X. Long-form single posts (300 to 800 words) outperformed threads consistently. Bookmarks were the silent winner. Posts I expected to flop because they were too long quietly racked up bookmarks for weeks.

Bluesky. Rewarded specifics. Vague hot takes got nothing. Real numbers, real names, real moments got reposts. The audience is allergic to anything that smells like a generic X take.

Truth Social. Steady but quiet. Low ceiling, low floor, predictable. A presence-keeping platform, not a growth-driving one. That's fine. That was the strategy.

Threads. Outperformed everything I expected from a platform I ignored for two years. The conversation-first algorithm is real. Replies pulled distribution that nothing else could.

Lesson. Don't pre-judge platforms based on what they were. Judge them based on what they reward right now.

Next 21 days. Same cadence. Better hooks. Less polish. More substance.

[AJ Kumar @ajkumar]

You think your content is failing because of strategy.

It's failing because of taste.

Strategy is the ability to know what to make. Taste is the ability to know what's good. AI gives every founder access to the first. Almost none of them have the second.

Coca-Cola's AI Christmas ads proved it. Strategy was perfect. Right channel, right season, right brand. Taste was nowhere. The output looked synthetic, lifeless, generic. The internet noticed instantly.

AI produces the statistical average of the internet. Taste is the human variable AI cannot replicate.

And taste isn't a gift. It's a trained ability. You build it the same way a chef builds a palate. By paying attention to what works, why it works, and what almost works but doesn't. By making bad work and keeping the receipts. By looking at great work in adjacent fields and noticing what they did that you didn't.

The founders pulling away in 2026 aren't the ones with the best AI stack.

They're the ones with the best taste.

Strategy is the floor. Taste is the ceiling.

[AJ Kumar @ajkumar]

Every system that processes information runs on one principle. Reduce the energy cost of retrieval.

The human brain. Google. ChatGPT. Three different surfaces. Same operating system.

The brain compresses experience into patterns because biology is metabolically expensive. Google ranks structured content because crawlers can't afford to parse chaos. AI models cite organized expertise because token cost is real and the most efficient retrieval wins.

All three reward the same things. Clear labels, named frameworks, accessible language, predictable formats, structured headings, defined terms.

All three punish the same things. Dense paragraphs without breaks, undefined jargon, unstructured prose, vague claims, and content that requires the reader to do the organizing the writer should have done.

Most experts make their content harder to retrieve. They confuse density with depth. They confuse jargon with rigor. Then they wonder why search engines don't rank them and AI systems don't cite them.

Accessibility for humans and accessibility for AI turn out to be the same thing.

The same ARIA standards built for screen readers are now what AI agents use to read your website. Build for the lowest-energy retrieval and you win at every layer.

[AJ Kumar @ajkumar]

Most founders grade their personal brand the way a fan watches football.

This post won. That one flopped. That keynote went viral. That essay got buried.

The mental model is the reason their content works in bursts and fades.

A personal brand isn't a contest between posts or channels. It's a living system of interdependent signals. The sport that mirrors how it behaves isn't football. It's Formula One.

F1 isn't decided in a moment. It's decided across a season of compounding decisions. Strategy, engineering, pit stops, tire choice, weather adjustments, telemetry, driver chemistry, all running at once. A bad lap doesn't end the championship. A good one doesn't win it.

Same with a personal brand. One viral post doesn't make you. One flop doesn't unmake you. Your audience is building a stable mental model of you across hundreds of touchpoints. Long-form, short-form, podcasts, replies, keynotes, the way you show up in someone else's comments.

Stop watching the post. Start tuning the system.

[AJ Kumar @ajkumar]

All the AI generated content creation is tempting but the future of content is human-led.

[Variety @Variety]

YouTube Will Start Automatically Tagging Videos That Make 'Significant' Use of AI, and It's Making Labels for AI-Generated Content More Prominent https://t.co/URW4AbXoFd

[AJ Kumar @ajkumar]

I see the same pattern at every $20M client we evaluate before working together.

A consultant told them to "professionalize the voice." The voice flattened. Inbound dropped 30 to 60 percent within 90 days. Audience disengaged. The consultant called it brand evolution. The data called it suicide.

I've stopped being surprised. I've started leading with this in every diagnostic call.

Here's why it happens. Founders intellectually understand the case for sounding like themselves. Then they hire a brand person who has only ever worked inside Fortune 500 marketing departments. The brand person flattens the voice because that's the only voice they know. The founder defers because the brand person is supposedly the expert.

Within a quarter, the audience built around the founder's voice can't find the founder anymore.

The polish should move into the frameworks, the production, the offers. Never the voice.

If you stopped sounding like you, you stopped being the reason people came in the first place.

[AJ Kumar @ajkumar]

Three signal levels every personal brand needs to operate. Most founders only run one. That's why their growth caps.

Wide Signal. Captures the market. Easy entry. Broad relatability. Designed to make a stranger stop scrolling. Trump and Kardashian content live almost entirely at this level.

Sharp Signal. Builds authority. Accessible language carrying real insight. The Tyson move. Translates expertise so the audience feels smarter for following you. This is where topical authority compounds.

Deep Signal. Rewards the core. Narrow, advanced, high-stakes applications of your expertise. The 80/20 rule playbooks. The insider frameworks. The "now that you understand the basics, here's the last 20" material.

Most founders skip straight to Deep. They want to sound smart. They show up only to the people already in the room. Then wonder why nobody new ever joins.

You need all three. In sequence. Wide pulls strangers in. Sharp earns trust. Deep keeps them.

Skip a layer and the whole stack breaks.

[AJ Kumar @ajkumar]

Memorial Day.

Today I'm thinking about the people who built and didn't get to see what they built become.

Soldiers. Immigrants. Entrepreneurs who quietly broke ground for somebody else's harvest. The ones whose names aren't on the building.

Most of what any founder gets to build sits on top of work somebody else started and didn't get to finish. My grandparents, my parents, mentors who taught me things they had no obligation to teach. Customers who took chances on a younger version of me.

Most founders skip the gratitude part because it doesn't feel like progress. It is.

Pause today. Build tomorrow.

[AJ Kumar @ajkumar]

Every founder I've worked with above $5M tells me the same thing on the first call.

"I wish I had started posting three years earlier."

Every. Single. One.

The cost isn't measured in followers you didn't get. It's measured in trust you didn't compound. Customers who chose someone louder. AI systems that don't have a stable entity for you yet. Three years of compounding interest on silence.

The founders who started early aren't more talented. They're not better writers. They're not luckier. They just stopped negotiating with themselves about whether the moment was right.

It's never the right moment. The brand guidelines aren't done. The strategy isn't perfect. The team isn't fully built. The platform isn't ideal.

All of that is true. None of it matters.

The cost of waiting is invisible until you can see what three years of compounding would have produced. By then it's too late to recover those three years.

Start ugly. Start now. Start before you're ready.

Three years from now you'll wish you had.

[AJ Kumar @ajkumar]

Three patterns I see in the founders breaking through right now. None of them are about going viral. All of them build the kind of entity confidence AI systems cite.

1. Sub-30-minute long-form video, weekly. The anchor format. Long enough for a viewer to form a stable mental model of the founder. Short enough to actually finish in one sitting. Every shorter asset downstream borrows trust from this anchor.

2. One named framework introduced per quarter. ROAC. V.O.L.T. Founder Signal. Mirror Test. Named ideas compound. Every time you reference the framework, you're planting a flag the AI systems and search engines start associating with you specifically.

3. Two podcast appearances per month on other people's shows. Borrowed audience. Borrowed authority. Third-party citation in the entity layer. The founders winning aren't the most famous. They're the most cited.

Most founders are still measuring follower counts. The ones who break through aren't.

Different game. Different scoreboard.

[AJ Kumar @ajkumar]

Short-form is engineering, not art.

Every viral short-form video is a stack of four neurological gates engineered in sequence.

Register. Fires in the first 0.5 seconds. Text hook, visual hook, spoken hook, opening frame motion.

Retention. Runs from second 2 onward. Format, pacing, prediction loops between every segment.

Resonate. The payoff. Substance plus a reframe that exceeds the brain's prediction.

Reinforce. The signature. A recurring format and identity that compounds attention into authority.

Going viral without Reinforce is a sugar rush.

I had a client video on Kobe Bryant hit over a million views and bring in zero clients. We had Register, Retention, and Resonate. We didn't have Reinforce. The viewer didn't know what to expect next, what we stood for, or why to follow.

Most creators win the first three and skip the fourth. That's why their best post never compounds.

[AJ Kumar @ajkumar]

Politics aside, two of the most influential communicators of the last 20 years use the same strategy.

Trump speaks at roughly a 4th-grade reading level. Short sentences. Repeating simple adjectives. Concrete imagery. He says "we're going to build a wall." Not "we will pursue thorough border infrastructure reform." He uses nicknames instead of arguments. He repeats phrases until they're memorable.

Neil deGrasse Tyson translates astrophysics into dinner-party language. He's the most successful science communicator alive because he doesn't sound like a scientist when he speaks. The substance stays advanced. The language stays accessible.

They could not be more different. They use the same strategy.

Accessible language travels further than sophisticated language. Every time, in every market, across every platform. The experts who refuse to simplify cap their reach without realizing it.

Einstein already said it. "If you cannot explain it simply, you do not understand it well enough." He wasn't being humble. He was being precise.

Complexity isn't proof of expertise. Translation is.

[AJ Kumar @ajkumar]

Founders losing attention aren't uninteresting.
They're using a broadcast voice in a medium that became two-way 10 years ago.

Every platform is now a relevance engine. The audience doesn't read your post the way someone reads a press release. They read it the way a friend would. Conversational, present, opinionated, alive.

When you sound like a corporate brand, the audience builds a corporate-brand mental model of you. When you sound like a person, they build a person mental model. The trust mechanism is identical to a real-world friendship.

The mistake at scale is over-corporatizing. Founders cross $20M and assume the bigger the brand gets, the more polished the voice should be. The opposite is true. The polish should move into the frameworks, the production, the offers. Not the voice.

Sharper. Not colder.

Most founder content fails this test before the substance even gets read.

[AJ Kumar @ajkumar]

I worked with Kimberly Snyder when her blog had a few thousand monthly readers.

Within a few years she was at 500,000 monthly readers, three New York Times bestsellers, and a wellness brand doing eight figures.

The shift wasn't a content strategy. It was a Founder Signal shift.

Expressive becomes Intentional. Not human becomes corporate.

She didn't flatten her voice when she scaled. She sharpened the production, the frameworks, the offers, and kept the voice expressive. Same person. Cleaner signal.

Most founders go the other direction. They cross a revenue threshold, hire a consultant, professionalize the voice, and watch their inbound dry up in 90 days. The consultant calls it brand evolution. The data calls it suicide.

The mistake at scale isn't sounding like a founder. It's sounding like everyone else's founder.

If you stop sounding like you, you stop being the reason people came in the first place.

[AJ Kumar @ajkumar]

WWE isn't a wrestling company.
It's a brand platform that launches a cast of characters. Each character is tuned to a different audience segment. Every storyline routes attention back to one core IP.

WWE swapped "Federation" for "Entertainment" in 2002. Repositioned from a sport to a media platform. Now operates the largest sports channel on YouTube. Over 100 million subscribers.

Ramsey Solutions reaches 20 million weekly listeners across a five-character roster. Eliminates the key-person risk in a single-founder show. iShowSpeed and Logan Paul's WrestleMania 42 run proved the cast-of-characters playbook still scales in 2026.

Most founders are still trying to be the only character in their own show. They wonder why the second character on their team can't get traction. The answer is the platform isn't built for it.

A brand platform owns infrastructure, distribution, and audience. A cast of characters owns the emotional connection and parasocial trust.

Hire a showrunner. Not a content manager.

Different role. Different leverage.

[AJ Kumar @ajkumar]

Sunday morning. The thing almost nobody warns founders about when they start building a personal brand.

Most quit at month 9.

Observable second-order signals start around month 12 to 18. Cascades depend on third parties repeating and citing your ideas, which takes its own time to accumulate. The math punishes anyone who measures the curve before it bends.

Year one is investment. Year two is interest. Year three is compounding.

Most founders never see year three because they quit in year one. They were measuring the wrong scoreboard the whole time. Looking at follower counts on day 90 and concluding the strategy doesn't work.

The strategy works. The timeline is just longer than the impatient version of you wants it to be.

If you're at month 6 and questioning the whole thing, you're exactly where everyone else who quit was. Don't quit.

[AJ Kumar @ajkumar]

The advice I give every new founder client when they ask me about cadence.

You don't need to post daily. You need to post 3 substantive things per week. Every week. For two years.

Most founders read "daily" advice from creators in the creator economy and assume it applies to them. It doesn't. Creators monetize attention directly. Founders monetize trust converted into customers. Two different scoreboards.

Three substantive posts per week, sustained for 24 months, is more than enough to build category authority. The math compounds quietly. The willpower required to post daily for two years almost never lasts.

The founders who build durable brands aren't the ones with the highest output. They're the ones with the longest consistency.

Pick a cadence you can hold for two years. Then hold it.

Everything else is decoration.

[AJ Kumar @ajkumar]

Most founders measure their content with the wrong scoreboard.

Views, likes, followers. Vanity metrics that look like momentum and feel like progress.

I track ROAC. Return on Attention Created. It maps how attention actually moves through four neurological gates.

Register. Earn the stop in the first 0.5 seconds.
Retention. Earn the watch in the next ten.
Resonate. Earn the save with substance plus a reframe that exceeds the brain's prediction.
Reinforce. Earn the return with a recurring format and identity that compounds attention into authority.

Going viral without Reinforce is a sugar rush.

I had a client video on Kobe Bryant hit over a million views and bring in zero clients. We had Register, Retention, and Resonate. We didn't have Reinforce.

Most founders win the first three and skip the fourth. That's why their best post never compounds.

The scoreboard you measure determines the strategy you build.