It’s 4 a.m., and I’m feeding my son with one arm while reading through a product backlog with the other. I want to be clear that this is not a productivity flex. I would love to be asleep. But somewhere in these strange, quiet, sleep-deprived hours over the last few months, I noticed something I genuinely did not expect: I’m getting more done, and thinking more clearly, than I did back when I had all the time in the world.

I became a mother in March. I went back to remote work in June. And in the middle of all that, while learning to run on far less sleep than I thought a person could, I kept my full-time job going, kept a newborn alive and happy, and on the side started building Minilu, the baby-tracking app I never planned to make until he arrived and I suddenly, urgently needed one myself. Three things at once, on a fraction of my old hours. Every logical prediction says that should have flattened me. Instead, it sharpened me. I’ve been trying to understand why, because it runs against almost everything we’re told about mothers and work.

The old version of me had all the time and less reason to guard it

A year ago I had long, open-ended afternoons, and I used them more loosely than I do now. “I’ll get to it later” was a real sentence I said to myself, and later usually did arrive. But a thirty-minute job could quietly stretch to fill an afternoon, mostly because nothing was forcing me to decide what mattered first. Having plenty of time wasn’t a flaw. It just never made me choose.

Now every workable block is about twenty-five minutes wide, and I can see the edges of it before it starts. When your son might wake up at any moment, you stop opening seventeen tabs “to think.” You pick the one thing that moves the needle, and you do that. The constraint didn’t slow me down. It did the one thing all those open afternoons never had to: it made me choose.

Watching someone develop from zero develops you too

There’s a part of this I can’t reduce to time management. Watching a human being grow from literally nothing, from not being able to hold her own head up to recognizing my face and deciding it’s her favorite thing in the room, does something to you. It rewires how you see progress, patience, and what “good enough for today” means.

It turns out that isn’t just a feeling. Neuroscientists have a word for it: matrescence, and the research on it is striking. Becoming a parent reshapes the brain’s default mode network, the system tied to planning, self-reflection, and reading other people. One large study of nearly 28,000 adults even found that parents tend to have “younger-looking” brains in mid- and later life than people without children, and researchers think it’s the daily cognitive workout of caregiving, not pregnancy alone, that does it. Which lines up almost too neatly with how I feel: like some part of me is being trained every single day, and my work is quietly borrowing the results.

The part that made “all three at once” possible

I’ll be honest about the piece that carried the load, because it’s the whole reason the math worked: the recurring, unglamorous parts of my job mostly run themselves now. Over the past months, I’d handed a lot of them off to AI, from the daily task ritual I used to retype every morning to the weekly social report that used to eat half a day, plus the no-code workflows humming along in the background of my week.

I built those before my son arrived, mostly out of curiosity. I had no idea I was actually building myself a maternity buffer. When the hours got scarce, the work that didn’t need a human simply kept happening without one, and the hours I did have went to the things only I could do: the strategy, the judgment calls, the parts of Minilu that needed an actual person who understood both the product and the exhausted parent using it at 4 a.m. Scarce time only sharpens you if something else is absorbing the busywork. For me, that something was the automations.

The penalty that turned out to be a premium

We have a name for what’s supposed to happen here. It’s called the “motherhood penalty,” the well-documented assumption that once a woman has a child, she becomes a less committed, less capable employee, and gets quietly passed over for it. I braced for some version of it in myself. I kept waiting to feel dumber, slower, less serious about work.

The opposite happened, and the research is starting to catch up to why. A 2026 study in the Journal of Applied Psychology found what the authors bluntly called a “motherhood advantage”: parents, and especially mothers, tend to attach more meaning to their work. They see it as a way to prove something to their families, and their supervisors independently rated them as putting more effort in, not less. The penalty is real as a bias. It is not, apparently, real as a fact about the work.

What I’m taking from this

Here’s the one thing I’d hand to anyone staring down a season where their time is about to collapse, whether that’s a new baby, a new caregiving role, or a new anything: a hard constraint won’t wreck your output the way you fear, as long as two things are true. You need a reason that makes the remaining time non-negotiable, and you need to have already offloaded the work that doesn’t require you. Constraint plus purpose plus a cleared plate isn’t a recipe for burnout. In my experience, it’s the most focused I’ve ever been.

I don’t think less time made me better. I think less time forced me to finally act like the things I said mattered actually did.

So I’m curious, and I mean this as a real question: when in your life did having less somehow make you sharper, whether it was less time, less budget, or fewer people, and what did that season teach you that abundance never did? Hit reply. I read every one.

Until next Monday,

This one is different because it’s about the thing I actually like doing. Writing. Which is also, somehow, the thing I could never manage to do on schedule.

Here’s the honest version. I love writing. That’s the whole reason this newsletter exists. But loving something has never once protected it from a busy week. Every time work got heavy, the newsletter was the first thing to slide, because it was the only item on my list that no one else was waiting on. “I’ll write it this weekend” turned into three skipped weeks more times than I want to admit. So the task I audited this time wasn’t a chore I hated. It was a thing I loved, but I kept losing it to everything more urgent.

I didn’t want to automate the writing. I wanted to remove everything around the writing that made it too heavy to start.

The ritual

Putting out one issue, the old way, looked like this.

I’d open a chat and ask for help with a draft, and get back something technically about my topic that sounded nothing like me: that flat, everything-is-a-bulleted-list, “in today’s fast-paced world” voice. So I’d spend twenty minutes re-explaining who I am: first person, no corporate throat-clearing, short sentences mixed with long ones, open on one small concrete thing, not a grand claim. Every single time, from scratch, like the last conversation never happened.

Then, once the piece was done, the other jobs started. Read it back and figure out what was actually worth pulling out. Write two or three short Notes to get people to read it. Write a LinkedIn version that isn’t just a pasted-in article, because that reads like a robot on LinkedIn. Make a cover image that doesn’t look like a stock-photo handshake. Each one a separate sit-down, each one me re-describing my taste to a blank box that didn’t remember it.

None of those steps were hard. But stacked together, on a busy week, they were heavy enough that “not this week” was always the easy answer. That’s the real thing that was killing my regularity. Not the writing, the weight around the writing.

What actually changed, and how it works

I’ll get concrete here because the how is the point of this one.

All of it runs in Claude Cowork. Think of it as Claude with hands: it can open my files, run steps in order, and keep working in the background instead of just answering in a chat. The important part isn’t the chat window. It’s that I can teach it something once and it stays taught.

The way you teach it is with skills. A skill is just a set of written-down instructions Claude loads when it’s relevant: your standards for a task, saved in a file instead of in your head, so you never have to re-explain them. I built three, one for each part of the loop that used to be manual.

The first is newsletter-writing, and this is the one that mattered most. It holds my voice. I built it the slow, unglamorous way: I fed it my own past posts and said this is the person you’re writing as. I added a short list of writers I actually like reading and said this is the neighborhood I want to live in; don’t copy them, just aim here. And then I corrected it, a lot. “Too formal.” “You stated a rule I’d never say out loud.” “This bullet list should be a sentence.” The difference from a normal chat is that every correction went back into the skill, so it stuck instead of evaporating when I closed the tab. After enough passes, drafts started coming back sounding like me on a good day. Still needing my hand, but starting from me instead of from nobody.

The second is substack-growth, and this is the analysis part. Once an article is finished, it reads the actual piece (not a topic, the real text) and works out which lines carry the idea, which are quotable, and what the hook is. From that, it drafts the Notes, the LinkedIn version, and a few alternative titles and subtitles, all in the same voice the first skill already knows. So the distribution isn’t a second creative project from a cold start; it’s the piece I already wrote, broken down and re-angled for each place it needs to live.

The third is blog-visual-mockups, which handles the images: cover art, simple diagrams, the occasional mocked-up screenshot, so I’m not fighting a design tool at the end of every issue.

Put together, the process now looks like this: I bring the idea and the actual thinking. The first skill drafts it in my voice. I edit by hand. That part is still mine, and I want it to be. Then the second skill turns the finished piece into everything I need to publish it, and the third makes the visuals. What used to be one heavy afternoon-that-kept-not-happening is now one focused writing session plus a review pass.

And I’ll be honest with you: this post ran through that exact system. I still went back in and rewrote lines by hand because some of them weren’t right. That’s the arrangement now, and I’m oddly fine with it.

What I’m taking from this one, and what you can steal

The first two audits taught me to look for the boring task. This one taught me to look at the task I love and ask a different question: not “how do I do this faster,” but “what’s the weight around this that keeps me from doing it at all?”

Because the fix here wasn’t a machine being me. It was me writing my taste down once (my voice, my format, my standards) in a place that remembers, so the energy I used to spend re-explaining who I am goes back into the only part that was ever really mine: the thinking, and the line that only lands if I write it.

If you want to try the same thing, the transferable version is small: pick one recurring piece of your own work, and instead of describing how you want it done this time, write those instructions down somewhere reusable. Feed it real examples of your own past work, correct it until it’s right, and keep the corrections. That’s the whole trick. Not “automate the thing.” Teach it your taste once, so it stops asking.

Three down. This is the one I was most afraid of, and it turned out to be less about giving something up and more about finally being able to keep a promise to myself to write every week.

If there’s something you love doing but keep failing to do regularly, I’d genuinely like to know what it is and what the actual weight around it is. Because I’d bet it’s not the part you love. Hit reply.

Until next week,

The ritual

Every week, and again at the start of every month, I put together a social media report. Numbers come in from a few different platforms. I’d sit down, pull them together, drop them into a Canva deck one slide at a time, double-check I hadn’t mixed up a number between platforms and then came the part that actually took focus: making sense of what those numbers meant and writing a short executive summary at the top, before exporting the whole thing as a PDF and posting it to our growth channel with a short note.

I never actually timed it until I thought about writing this. When I did the math, it landed somewhere around half a day. Not because any single step was hard, because there were a lot of small steps, and one of them (the summary) genuinely needed me to think, every single week, without fail.

What actually changed

That whole loop, reading the numbers, working out what actually moved and what it might mean, drafting the executive summary, updating the deck, exporting, sharing it in Slack, now runs on its own. Practically, it runs through Claude Cowork: it pulls that period’s data, opens the actual Canva deck and edits it directly inside Canva, exports the result as a PDF, and posts it into our Slack channel, the same sequence I used to do by hand, minus me sitting there for all of it. Every Monday for the week before, and on the 1st of the month for everything before that. I still open the result and glance at it before I fully move on with my day. That takes about half an hour, mostly out of habit.

Half a day down to half an hour. That’s the whole story, honestly, more than any clever technical detail about how it works.

What I’m taking from this one

The first automation I built saved me a small daily annoyance. This one saved me actual hours, every single week, that I used to just accept as the cost of the job. I hadn’t clocked how much of my week this one task was quietly eating until it stopped eating it. And honestly, I didn’t expect the part I’d have to think through the summary, the “what does this actually mean” part, to be the part that got handed off too. I assumed I’d still be doing that bit myself.

I think that’s the real thing I’d tell you to look for in your own week: not the task that annoys you the most, but the one you’ve stopped noticing takes as long as it does, because you’ve been doing it that way for so long it just feels like “part of the job.”

Two down, and this is the one that’s actually changed how my week feels. Mondays used to have a report hanging over them, and now they mostly don’t.

If you’ve got a report, a deck, or anything recurring that quietly eats half a day of your week, I’d genuinely like to hear about it. Hit reply.

Here’s where I started.

The ritual

Every morning at 9:00, a Slack bot on our team - we call him Komuncan - posts the same line in #general: “Good morning! Let’s write today’s agenda.” And every morning, the whole company types their to-do list underneath it.

It’s a nice habit, genuinely. Everyone can see what everyone else is working on. But I wanted to keep my list more organized and actually be able to check at the end of the day what got done and what didn’t. So I asked Claude to help with exactly that. I didn’t set out to build anything clever. I just figured this is a chat interface; I’m allowed to just ask it to help with the literal, boring, repetitive part.

What I actually did

I asked Claude to help me keep a running list of my tasks instead of relying on my memory and a Slack thread. We landed on a shared Canvas - basically a doc both of us can read and write to - as the place my task list actually lives now, instead of disappearing into a scrolled-away channel.

Then, because the company ritual specifically happens in that Slack thread, I asked if it was possible to have Claude post there for me, automatically, a few minutes after Komuncan does. Turns out: yes. Every morning at 9:15, it finds that thread on its own and replies with my list of yesterday’s finished items crossed out, anything I didn’t get to automatically carried into today so it doesn’t quietly disappear.

At 6 PM, it checks in with me and asks what actually got done. I want to be honest that this part still needs me; it’s not deciding on my behalf whether my day was productive; it’s just asking. I don’t know yet if that’s the “correct” way to do this, but it felt important not to let the one reflective part of my day get fully automated away, too.

What I’m taking from this, so far

I don’t think I’ve found a grand rule yet. What I can say is that this specific task recurring, same format every time, currently living only in my memory and a scrolling thread — turned out to be a pretty gentle place to start. It didn’t require anything technical from me beyond describing what I wanted, in plain language, more than once, until it worked the way I meant.

If you’re earlier in this than even I am, here’s the only advice I feel confident giving right now: find the one thing you retype in almost the same format every single day, and just ask if it’s possible to stop retyping it. You might be surprised at how small the first useful automation actually is.

I’m sure I’ll have more of these as I keep going; some will probably work better than this one, and I wouldn’t be shocked if a few fail in ways I can’t predict yet. I’ll write about those too.

If you’ve been experimenting with this kind of thing yourself, I’d genuinely like to hear what you’ve tried. Hit comment and tell me. I’m learning this in public; might as well learn from you too.

I wrapped up certificates in Claude with the Anthropic API, Introduction to Model Context Protocol, Introduction to agent skills, and Claude Code in Action.

You might be wondering why a marketing leader is spending their time learning about context protocols and command-line interfaces instead of, say, the latest TikTok algorithm update.

Here is the truth: we’ve hit a ceiling with the traditional AI chat interface.

Typing a prompt into a web window and waiting for a response is fine for brainstorming an email subject line or outlining a blog post. But when you want to build actual marketing operations, systems that categorize thousands of customer feedback tickets, generate localized ad variations on the fly, or automatically pull competitive pricing, copying and pasting doesn’t scale. I wanted to understand how to move AI out of the chat window and wire it directly into the plumbing of my campaigns.

Learning the Anthropic API showed me how to programmatically send and receive data, bypassing the manual interface entirely. The agent skills course was a massive lightbulb moment: it’s about giving the model actual tools. Instead of just asking Claude a question, you can give it the ability to search your product database, read a live webpage, or trigger an email sequence.

The Model Context Protocol (MCP) was particularly interesting for B2B and SaaS. In marketing, our biggest hurdle with AI is context. The model doesn’t inherently know our brand guidelines, our past campaign performance, or our highly specific customer segments. MCP provides a secure standard for connecting these models directly to our private data sources without compromising security. And Claude Code? It showed me how AI can sit right inside a development environment to handle repetitive technical tasks, bridging the gap between what marketing wants and what engineering has to build.

By understanding the technical constraints and capabilities of these systems, I discovered that everyone can design significantly better marketing workflows. I don’t necessarily want to become a full-time software engineer, but knowing exactly how the pipes connect means I can architect processes that actually work, rather than just throwing vague ideas at my team.

If you want to start shifting your marketing team from “prompting” to “system building,” here is how to start thinking about it:

• Audit your copy-paste loops: Spend a week writing down every time you or your team manually copies data out of a tool, pastes it into an AI chat window, and copies the result back. These are your prime candidates for API automation.

• Define your required “tools”: When an AI struggles with a marketing task, it’s usually because it lacks access to real-time information. Ask yourself: what tool does it need to finish the job? A calculator? A web scraper? A CRM lookup? These are “agent skills” you can build.

• Use AI as a router, not just a writer: Some of the best marketing use cases have nothing to do with generating content. Use the API to read incoming lead data, structure it, and route it to the right sales rep based on context.

• Talk to your developers differently: If you understand the basics of APIs and how context windows work, you can write far better technical briefs. You stop asking for magic and start asking for specific data flows.

Let’s look at a concrete scenario. Suppose you run a SaaS company and want to track feature requests from support tickets to inform your next product marketing push.

The old way: Every Friday, a junior marketer exports a massive CSV from Zendesk, pastes chunks of it into an AI chat window, and asks for a summary of feature requests. It takes two hours, the context window constantly breaks, and the output format changes every single week.

Learning the technical side of these tools was a bit humbling at times, but it completely changed how I look at marketing automation. We are moving from a world where we “talk” to AI to a world where we build with it.

I’m curious about what repetitive, manual marketing task you wish you could just wire an agent to handle in the background? Hit reply and let me know. Let’s see if it’s possible to build it.

Until next week,

Instead of just guessing what AI could do, the researchers created a brilliant new metric called “observed exposure”. This metric blends the theoretical capabilities of AI with real-world usage data from their Claude platform to see what workplace tasks are genuinely being automated today.

The biggest takeaway from the study is that we can all take a deep breath, because AI is still far from reaching its theoretical capabilities.

Who is really feeling the impact?

When we think of vulnerable workers, we often picture blue-collar jobs or entry-level roles, but the data shows the exact opposite. The workers who are most exposed to AI automation are actually more likely to be older, female, highly educated, and higher-paid. In fact, the report highlights that people with graduate degrees are almost four times more likely to be in the most highly exposed group compared to the unexposed group.

So, what are these highly exposed jobs? The professions leading the pack right now include Computer Programmers (with a massive 75% of their tasks covered by AI capabilities), Customer Service Representatives, and Data Entry Keyers. On the flip side, jobs requiring physical presence or specialized non-digital skills, such as cooks, lifeguards, and mechanics, have virtually no AI exposure.

Are the mass layoffs happening?

Here is the best news from the report: despite the explosive growth of tools like ChatGPT since late 2022, there has been no systematic increase in unemployment for workers in these highly exposed jobs. The catastrophic job losses some predicted simply haven’t materialized in the aggregate unemployment data yet.

A subtle shift for younger workers

While older, established workers seem to be holding onto their jobs just fine, the research found a concerning trend for the next generation. There is suggestive evidence that hiring for young workers (ages 22 to 25) in highly exposed occupations has noticeably slowed down. Specifically, the rate of young workers finding new jobs in these vulnerable fields has dropped by about 14% since the post-ChatGPT era began.

So, while AI isn’t causing mass layoffs, it might be quietly closing the door on entry-level opportunities in certain white-collar fields.

Looking ahead

Anthropic’s research shows that the AI revolution in the workplace is definitely happening, but it’s more of a gradual integration than an overnight economic apocalypse. As capabilities grow, we will inevitably see more tasks automated, and the US Bureau of Labor Statistics already projects slightly weaker employment growth for these highly exposed jobs through 2034.

For now, your job is probably safe, but if you’re a recent grad trying to break into coding, finance, or data entry, you might find the job market just a little bit tougher to crack than it used to be.

Read the full 2026 Anthropic Research Paper

But if you are a new entrepreneur, you need to hear the truth: that straight line is a complete myth.

According to seasoned entrepreneur and investor Howard Love, the actual path of a successful startup looks like the letter “J”. When you first start, you don’t go straight up; instead, you take a steep dip into what is often called the “Valley of Death”. Products take longer to build, customers don’t understand your initial offering, and cash starts burning fast. But don’t panic! This dip is totally normal.

In his book The Start-Up J Curve, Love breaks down the startup journey into six predictable phases. By understanding these phases, you can figure out exactly where you are, avoid deadly mistakes, and eventually ride that curve straight up to success. Let’s walk through them.

1. Create: It’s Not Just About the Idea

The first phase is where the magic begins. You have a vision, and your optimism is through the roof. However, Love shares a dirty little secret: your brilliant idea is only worth about 5% of the startup’s total value.

Why? Because execution is everything. You should treat your initial idea simply as a “hypothesis,” a guess that you are going to test in the real world. In the Create phase, your focus shouldn’t be on protecting a “perfect” idea, but rather on finding a real problem to solve, securing initial funding, and putting together a solid founding team to execute the vision.

2. Release: Get the Damn Thing Out There!

This is where many first-time founders freeze. They fall into the trap of perfectionism. They want to add just one more feature or do one more redesign before anyone sees it.

Love’s advice? Stop tinkering and get your product into the market. Your goal here is to release a Minimum Viable Product (MVP). It won’t be perfect, and that is exactly the point. You aren’t releasing it to make millions of dollars on day one; you are releasing it to get real feedback from real customers. Time is your enemy, so launch fast and listen closely.

3. Morph: Embrace Radical Change

Here is a wild statistic: roughly 90% of startups have to drastically change their original business plan. When your product hits the market, customers might use it in ways you never imagined, or they might not care about the feature you spent months building.

This phase is all about “morphing” or pivoting. It can be painful to let go of your original vision, but you have to drop your ego. Your job is to adapt the product based on what the market is actually telling you. Don’t try to force a product on people who don’t want it. Change it until you find that sweet spot where customers genuinely love what you are doing.

4. Model: Nail It Before You Scale It

Once you have a product that people love, you need to figure out how to make actual money. The Model phase is about proving the economics of your business.

A lot of entrepreneurs think that a great product automatically equals a great business, but that’s not true. You need to find a business model where the revenue you make from a customer is higher than the cost of producing your product and acquiring that customer. Look for a model that has high margins, is easy for customers to buy (low friction), and can be repeated over and over again. Whatever you do, do not scale your business yet. Scaling a flawed business model is the fastest way to kill your startup.

5. Scale: Pouring Fuel on the Fire

Okay, you have a product people love, and a business model that makes a profit. Now it is time to go big.

Scaling means massively increasing your operations. During this phase, your focus shifts to three things: people, processes, and money. You will likely need to raise serious venture capital to fund this expansion. You will also need to change how your team operates. In the early days, you needed “generalists” who could do a bit of everything; now, you need to hire experienced “specialists” to handle specific departments. You have to build systems and rules because you can no longer manage every single detail yourself.

6. Harvest: Enjoying the View

You survived the Valley of Death, you morphed, you nailed the model, and you scaled. Now you are in the Harvest phase. Love describes this as the most fun part of the journey.

You are generating cash and making what he calls “puffball decisions”—decisions that are generally favorable no matter which option you choose. During this phase, you get to decide how to reward yourself and your investors. Will you take the company public (IPO)? Will you sell it to a larger corporation? Or will you keep running it and enjoy the long-term wealth of compound growth?.

Two Golden Rules for the Road Ahead

As you start your journey, keep these two extra pieces of advice from Love in your back pocket:

• Ramen Profitability is a Superpower: Don’t stress too much about raising massive amounts of money on day one. Having just enough money to survive (enough to eat ramen noodles) forces you to be disciplined, focused, and fast. Scarcity is often a better motivator than a bloated bank account.

• Make Failure Your Friend: In a startup, operational failures, like a marketing campaign that flopped or a feature that crashed, are normal. Don’t hide from them. Acknowledge them quickly, learn the lesson, and move on. “Fail fast” so you can succeed sooner.

The startup journey is an emotional rollercoaster that will test your patience, your bank account, and your sanity. But by keeping the J Curve in mind, you will know that the initial dip isn’t the end of the road; it is just the beginning of your climb.

Stay flexible, listen to your customers, and enjoy the ride!

The Yum! Brands (Food Trends) report has a brilliant formula for this: The Vibe Equation. They found that “Value” is no longer just Price / Quality. It’s now Price / (Quality + Vibe).

If a brand delivers a strong enough emotional lift, a moment of comfort, a laugh, or a “treat yourself” dopamine hit, consumers will ignore the logical price tag. Adobe backs this up with their “All the Feels” trend, noting that softness, comfort, and even “Surreal Silliness” are becoming the primary drivers of loyalty.

While AI handles the boring, logical tasks, your brand’s only job is to provide the thing AI can’t: A feeling.

Your Action Plan: How do you operationalize “vibes” without sounding like a teenager? Here are three takeaways from the reports:

• Audit for “Glimmers,” not just Grinds: The GWI and Consumer Trends reports highlight a shift toward “micro-joys.” Look at your customer journey. Is every touchpoint purely functional? Find one spot—your welcome email, your loading screen, your packaging—and insert a moment of pure, illogical joy. (Think: A funny micro-copy line or a surprise visual).

• Embrace “Surreal Silliness”: Adobe predicts a rise in “weird” aesthetics, chaotic, colorful, and slightly absurd visuals that break the polished “corporate Memphis” look. Why? Because it signals humanity. If your B2B ads look too perfect, they look like AI. Try getting a little messy.

• Sell the “Ritual,” Not the Product: The Food Trends report notes that people aren’t just buying coffee; they are buying a “Ritual of Control” in a chaotic world. Don’t sell your software/service as a tool. Sell it as the calm moment in their crazy day. Position your product as the anchor.

A Quick Example: The Consumer Trends 2026 report mentions the explosion of “Labubu” (those quirky plush toys). Logically? It’s a useless expense. But in the Vibe Economy, it’s a massive hit because it offers community and a shared emotional language.

A B2B version of this? I saw a project management tool recently that didn’t advertise “efficiency.” They advertised “The feeling of closing your laptop at 5 PM on a Friday.” They sold the vibe of relief, not the logic of features. And it worked.

As we move into 2026, the brands that win won’t be the ones with the best specs. They’ll be the ones that make us feel something in a numb world.

I’m curious what the last purchase you made purely for the “vibe,” even if it didn’t make logical sense? (Mine was an overpriced ceramic mug because it just felt like a good Monday morning).

Hit reply and tell me yours.

Stay human,

There was a specific thrill in hitting “ON” and watching data zip across the screen without me having to touch a keyboard. I told everyone I had “automated my reporting.”

But last week, that feeling evaporated.

I looked at a lead scoring workflow I built two years ago. It was a 14-step monster. A few months ago, one of our tools changed a field name from Email_Address to Contact_Email.

The result? The entire system silently failed for three weeks.

I realized I hadn’t built a system; I had built a digital machine. It was powerful, yes, but it was incredibly fragile. If the wind blew the wrong way, the whole thing collapsed.

This fragility is exactly why the next phase of AI is going to change everything. We are leaving the era of “If This Then That” and entering the era of Agentic AI.

The Shift: From “doing the task” to “owning the outcome.”

If you’ve been reading the tea leaves for late 2026, the industry is screaming one phrase: Agents, not Chatbots.

Here is the simplest way to understand the difference, and why it matters to your Monday morning:

• Traditional Automation (Zapier/Make): You are the architect. You define every single step. “If a user fills form X, wait 2 minutes, then send email Y.”

  • Pro: Predictable.

  • Con: Brittle. If the form changes, the bot breaks.

• GenAI (ChatGPT/Claude): You are the prompter. You ask for text/images. “Write a follow-up email for this user.”

  • Pro: Creative.

  • Con: Passive. It sits there waiting for you to ask.

• Agentic AI: You are the manager. You define the goal. “Here is a lead. Research them, decide if they are high-value, and draft the appropriate welcome sequence. Ask me if you’re unsure.”

  • Pro: Autonomous and resilient.

  • Con: Scary. (We’ll get to that).

We are moving from tools that can generate text to tools that can sense, reason, and act.

Why this is a “Wow” moment for marketers

Imagine you want to repurpose this newsletter into a LinkedIn post.

The Old Way:

1. Copy text.

2. Paste into ChatGPT.

3. Prompt: “Rewrite this.”

4. Copy result.

5. Open LinkedIn.

6. Paste.

7. Find an image.

8. Hit post.

The Agentic Way: You have a “Content Repurposing Agent” running in the background. It “watches” your newsletter folder. When a new file appears, it:

1. Read it to understand the tone.

2. Check your LinkedIn analytics to see what format is performing best this month (Carousel? Long text?).

3. Draft the post based on that data.

4. Pings you on Slack: “I’ve prepared a draft based on today’s newsletter. I formatted it as a text-only post because those had 20% higher engagement last week. Approve?”

You hit “Approve.” Done.

The agent didn’t just follow rules; it made a decision based on context.

Your “Agentic Readiness” Playbook

You don’t need to build a complex agent today. But you do need to start thinking like an Agent Manager, not just a creator.

Here is how to prepare:

1. Audit for “Decision Points”: Look at your current workflows. Where are you currently acting as the “router”?

• Example: “I read the support ticket, see it’s angry, so I assign it to Sarah.”

• The Shift: This is where an agent fits. It can analyze sentiment and route automatically.

2. Define “Freedom to Act” Levels: Not all agents need the keys to the car. Think of them like interns.

• Level 1 (The Scribe): Can research and draft, but a human must hit “send.” (Start here!)

• Level 2 (The Copilot): Can reply to internal team members or low-risk queries.

• Level 3 (The Agent): Full autonomy to execute actions (e.g., refunding a customer under $50).

3. Keep the “Junior” Loop: If AI does all the grunt work (research, summarizing, basic drafting), how do your junior marketers ever learn? Make sure your human team still does the “reps” occasionally. You can’t manage an agent if you don’t know how to do the job yourself.

Closing thoughts

The future of marketing isn’t about who can write the best prompt. It’s about who can build the best system.

We are graduating from being “writers” to being “editors-in-chief” of our own AI newsrooms.

Question for you: If you could hire an AI intern today to take complete ownership of one annoying task, without you ever checking it, what would it be? (Mine is definitely chasing invoices.

Hit reply and let me know.

See you next week,

When I was consulting with a few startups during their growth phase, one of the mentors I worked with recommended a book to me: “Thinking in Bets” by Annie Duke.

At first, I hesitated. I had already found a PDF version online, but I’m someone who loves the feel of turning real pages. I ended up ordering the physical copy from Amazon, wondering whether it would be worth it. It turned out to be one of the best reading decisions I’ve made in a long time.

Annie Duke, a former professional poker player, draws powerful parallels between poker and life. Her central argument is simple yet profound: we’re all playing at the table, whether we realize it or not. Every choice we make — in business, relationships, or personal life — is a bet made under uncertainty.

As entrepreneurs or decision-makers, we’re constantly trying to “make the right call,” often with incomplete information. This book doesn’t promise perfect decisions — instead, it teaches you how to think more clearly about them.

Here are a few insights that stayed with me:

🎯 Every Decision Is a Bet

We never have all the information. The best we can do is weigh the potential gain, possible loss, and alternatives. The question Duke poses — “Is it worth the bet?” — reframes decision-making as an act of calculated courage rather than blind certainty.

⚖️ Don’t Judge the Decision by Its Outcome

A poor result doesn’t always mean you made a bad decision — and a lucky outcome doesn’t mean you were right. What matters is the process and context in which you made your decision. This mindset helps entrepreneurs separate luck from logic and avoid hindsight bias.

🔢 Speak in Probabilities, Not Absolutes

Instead of saying “This will work”, say “There’s a 60% chance this will work.” Thinking in probabilities forces you to acknowledge uncertainty and refine your predictions over time. It’s a small shift in language that builds mental flexibility and realism.

🔄 Update Your Beliefs as You Learn

Great decision-makers don’t cling to being right — they update their views as new data comes in. As Duke puts it, “Saying ‘I might be wrong’ is a superpower.” For entrepreneurs, this means adapting fast, not defending old assumptions.

🧭 Build a Decision System

Instead of relying on gut feelings alone, create a repeatable structure:

• Define “exit criteria” before you start (“If X happens, we’ll stop”).

• Conduct premortems — imagine what could go wrong in advance.

• Reverse-engineer your path from the goal backward.

• Surround yourself with a small, honest group that gives feedback on your reasoning, not just the results.

💬 Why This Book Matters

Reading Thinking in Bets reminded me that decision-making is not about being right — it’s about being thoughtful under uncertainty. Especially for founders and marketers like myself, every campaign, hire, and strategic move is a bet. The goal isn’t perfection, but consistent improvement in how we think.

This book deserves a place on every entrepreneur’s shelf — not just because it teaches better decision-making, but because it helps you make peace with not knowing.

If you’ve ever questioned a choice you made — or replayed a “what if” scenario in your head — this book will help you see decisions in a new light. After all, life is one long series of bets. The key is learning to play them wisely.

We’ve all been there. It’s 9:00 AM. You have a massive deliverable due, a strategy deck, a migration plan, or a high-stakes launch email. You sit down, crack your knuckles, open a fresh Google Doc… and stare.

Suddenly, reorganizing your Chrome bookmarks by color feels incredibly urgent. Or maybe you decide it’s the perfect time to audit your entire Notion workspace. (Guilty.)

This is “Blank Page Syndrome,” and for years, it absolutely paralyzed me.

Most marketers try to solve this by opening ChatGPT and typing: “Write a marketing strategy for a B2B SaaS company.”

And what do they get?

“In today’s fast-paced digital landscape...”

Generic, soulless, corporate fluff. It sounds like everyone else because it is everyone else. It lacks your nuance, your context, and your specific constraints.

But recently, I stumbled on a workflow that completely changed how I start big projects. I call it the 15-Minute Journalist Method.

It turns “chaos” into “clarity” before my second cup of coffee is cold.

The Core Shift: AI as an Interrogator, Not a Writer

The mistake we make with AI is treating it like a Junior Copywriter (”Here is a topic, write this for me”).

Instead, you need to treat it like a Ruthless Journalist (”Here is my messy brain; interview me until it makes sense”).

You already have the strategy in your head. You have the messy notes, the shower thoughts, the 3 AM anxieties, and the gut feelings about why the current plan is failing. You just don’t have the structure.

So, stop trying to write. Start talking.

The 15-Minute Protocol

This is the exact workflow I used last week to draft a Scope of Work that I had been avoiding for three days.

Step 1: The “Ugly” Ramble (5 Minutes) Close your laptop. Open the voice memo app on your phone (or use a tool like Otter.ai or AudioPen).

Hit record and just talk.

Do not filter yourself. Do not try to sound smart. Vent about the project. “Okay, we need to launch this feature by March, but the engineering team is behind. We need to target CFOs, but I’m worried our messaging is too fluffy. It needs to be serious about ROI. Also, we have no budget for paid ads, so we need a heavy organic play...”

Get it all out. The messier, the better.

Step 2: The Transcript Transfer (2 Minutes) Take that raw, rambling text transcript. It will look terrible. It will be full of “umms” and half-finished sentences. That is perfect.

Paste it into your AI tool of choice (Claude is my favorite for this because it handles nuance well, but ChatGPT-4o is great too).

Step 3: The “Ruthless Editor” Prompt (3 Minutes) Do not ask it to write the draft yet. If you do, it will hallucinate details you didn’t provide. Instead, use this specific prompt:

“I am going to paste a raw brain dump for a [Strategy Document / Email / Campaign].

Do NOT write the draft yet.

Instead, act as a senior editor. Read my notes, analyze my logic, and identify the gaps. Then, ask me 3–5 specific, hard questions to clarify the direction. Force me to make decisions.”

Step 4: The Interview (5 Minutes) The AI will come back with questions like: “You mentioned targeting CFOs, but your notes focus on ‘speed of implementation.’ Do CFOs actually care about speed, or do they care about cost savings? Which is the primary hook?”

This is the magic moment.

When you answer these questions, you aren’t “writing”, you’re solving problems. You’re making strategic choices in real-time.

Once you’ve answered the questions, give one final command: “Okay, now summarize our conversation into a structured outline.”

Boom. You have a solid, logic-tested structure. The blank page is dead.

Why this works

When I first tried this, I realized something painful: I wasn’t procrastinating because I didn’t know how to write. I was procrastinating because I didn’t know what I thought yet.

Writing requires two different brain modes:

1. The Architect (Structure, logic, strategy)

2. The Builder (Words, grammar, flow)

Trying to do both at once is exhausting. This method separates them. The “Ramble” lets you be the Architect without worrying about the bricks. The “Interview” forces you to stress-test your blueprints.

A Quick Example

I was recently working on a positioning doc for a client in the fintech space. I was stuck. I kept typing “We help companies save money...” and deleting it.

I did the Journalist Method.

I rambled into my phone while walking the dog. I fed it to Claude.

Claude asked me: “You kept saying ‘we’re different because we’re transparent,’ but you didn’t give a single example of transparency. What is a specific transparent practice you do that competitors don’t?”

That question stopped me cold. I realized we didn’t actually have a differentiator yet. We had to go back to the product team and build one (a live dashboard for fees).

If I had just forced myself to “write the copy,” I would have written a vague, weak marketing claim. The process forced the strategy to get sharper.

Try this today

If you have a task on your to-do list that you’ve migrated from Monday to Tuesday to Wednesday... try this:

1. Stop typing. Pushing keys isn’t working.

2. Dictate for 3 minutes. Use your phone. Pace around the room.

3. Get interviewed. Let the AI find the holes in your thinking.

You’ll be surprised how fast you can get to a “Draft 1” that actually feels like you.

Question for you: What is your favorite way to “waste time” when you’re avoiding a big project? (Aside from color-coding spreadsheets, which is clearly productive work... right?)

Hit reply and tell me.

See you next Monday,

I just wanted to tell you about the Social Hackers Academy (SHA) in Athens. Our mission is all about helping vulnerable people learn how to code. I worked as the Head of Marketing there in 2019 for a year, and it was an amazing experience!

At SHA, we focus on teaching software engineering to make sure our students are ready for jobs in the tech industry. It’s all about giving people the skills they need to succeed.

In my role, I got to work with a fantastic team. We used social media and other digital tools to share inspiring stories from our students. It’s all about building a community that cares about everyone, especially those who need a little extra support. It’s not just about coding – it’s about changing lives and creating a place where everyone feels valued.

Starting at Social Hackers Academy (SHA) in Athens was both exciting and a little scary. I’d never been to Athens before and hadn’t met my colleagues in person, so adjusting to a new city and job had its challenges. Even though I had to get used to everything from afar, I quickly got into SHA’s vision of using technology to help people. Our team used social media and stories to show the real people behind our work, building empathy and support for those in need.

One of the best parts of working at Social Hackers Academy (SHA) was hearing their stories and seeing how happy they were when they were supported. That’s what kept me going every day. I had a really good experience teaching computer skills to a group of women. We started with the basics, like how to use a mouse and keyboard, and then moved on to making presentations. It was so rewarding to see how excited and confident they became as they learned new things. I loved being part of their journey and helping them feel more comfortable in the digital world.

These moments showed me how important SHA’s work is, and it made me even more passionate about supporting people from different backgrounds. It’s amazing to see how education can change lives, and I’m grateful to have been a part of it.

At the end of the year, I had to go back home due to family reasons. But when I returned, something amazing had happened. I felt like a completely different person – happier and more self-confident than ever before.

I packed my suitcase and started a whole new life in a different city. It was a fresh start, and I was excited to see what the future held. As I settled in, I made some incredible friends who added so much joy to my life.

It was the pivotal moment that shaped my journey.

Your inbox is already a mess, I know mine is.

So, why am I doing this?

I’ve spent 14 years in marketing. I’ve been in the big agencies, the SaaS startups, and the C-suite meetings where I spent most of my time just trying to get departments to talk to each other.

The biggest thing I’ve learned is that marketing is never a straight line. Most of the advice you see online tries to make it look neat: “How I made $10k in 48 hours” or “The 10x Playbook.”

Real life is messier. Strategies fail because the operations are broken. Campaigns flop because the data was wrong. That’s why I’m starting Monday Marketing. I’m not here to be a guru; I just want to help you connect the dots between the chaos.

The “Hybrid” Problem

My main theory is that you can’t just be good at one thing anymore. Ten years ago, you could just write copy or manage ads.

Today, if you write copy but don’t understand how the automation triggers the email, you’re flying blind. If you build systems but don’t understand the strategy, you’re just building a faster way to send bad messages.

We have to be “Hybrids” now. We need the mix of strategy, ops, creative, and AI. When those four work together, you actually get a marketing engine that works.

Where is your gap?

I do this quick check whenever I start a new project. Look at your current setup and be honest, which bucket is the weakest right now?

• Strategy: “I don’t know who we’re actually talking to.”

• Ops: “I’m doing everything manually, and I’m exhausted.”

• Creative: “Our brand sounds like everyone else.”

• AI: “I know I should use it, but I don’t know where to start.”

You don’t need to fix everything at once. You just need to fix the one thing slowing you down.

I remember working with a team that had beautiful, creative, but zero ops. They were manually writing every single lead email because they were afraid automation would sound robotic. We put their five best emails into a sequence and set a simple rule: if a human replies, the automation stops. They saved 10 hours a week, and conversions actually went up.

Sometimes you don’t need to “work harder.” You just need to fix the plumbing.

What’s next

I’ll keep these short. Next Monday, I’m digging into "The Blank Page", how to use AI to start a strategy document without getting generic, garbage results.

What I’d do now

Hit reply and tell me which of those four buckets is your biggest headache today. I’d love to know what you’re actually dealing with.

I’ll see you next week,