You Don't Need a Marketing Team. You Need a Story. AI Should Be Doing the Rest.

The rant

I'm going to say something that will get me uninvited from a few marketing dinners.

Your marketing team isn't a team. It's a tax. Specifically, it's the 2022 Marketing Production Tax. And you've been paying it for four years past its expiration date.

Six people. Five tools. Three vendors. One agency on retainer. Brand reviews. Wireframe approvals. HubSpot tickets. Sprint planning for a homepage update that should take 20 minutes and somehow takes a month.

That stack made sense in 2022. It does not make sense now. And the reason most founders haven't realized this yet is that the world changed in January 2026, and most marketing leaders haven't checked since 2024.

I'm just the messenger

I know this post is going to make a lot of marketing leaders hate me. I walk on the graves of CMOs every week. It feels that way some weeks.

But here's the data. The average tenure of a B2B CMO is around 13 months. Shorter than any other C-suite role by a wide margin. Their boards are already firing them faster than they can hire them. That isn't my opinion. That's what is happening.

The reason CMOs are getting fired faster than every other executive isn't that they're bad at their jobs. It's that the job most of them were hired to run, the 2022 production cycle, is no longer the job that creates leverage in a 2026 company. The boards know it. The founders know it. Most of the CMOs themselves don't yet, because nobody told them the production cycle is collapsing underneath them in real time.

And I'm not the first person to say any of this. The entire Fractional CMO industry, a billion-dollar category that didn't exist 15 years ago, exists on the premise that most companies don't need a full-time CMO. That argument has been mainstream for a decade. I'm just adding the sentence nobody has said out loud yet: AI is now doing to the Fractional CMO what the Fractional CMO did to the full-time CMO. The trajectory hasn't changed. Just the speed.

What changed in January 2026

AI image generation crossed the threshold for production-quality marketing assets in late 2025. By January 2026, the models trained on every great design pattern of the last 30 years can produce a homepage hero, a product page, a webinar landing page, a campaign banner... not a draft, not a mockup, the actual shipped asset... in minutes.

Your marketers don't know this. They tried Midjourney in 2024, got something that looked like a fever dream, and decided AI couldn't do design. Then they stopped checking.

This is the most expensive blind spot in B2B marketing right now. The capability gap between what AI could do 12 months ago and what it can do today is bigger than the gap between a flip phone and an iPhone. And most marketing leaders are still benchmarking the iPhone against the flip phone they tried in 2024.

The CMS lie

Here's the secondary tax most founders forget they're paying.

In 2018 you bought into the promise that HubSpot, WordPress, and Webflow would make your website "easy to update." Plug-and-play. Anyone on the team can change copy. Marketing owns the website. No more bottleneck on engineering.

Look at your homepage right now.

How many sections has it had for the last six months? When was the last time anyone changed the hero? Does your footer still say 2025?

I'll wait.

The CMS promise was never about ease. It was about feeling easy enough that you'd stop asking whether the actual production cycle had gotten faster. It hadn't. You replaced a developer with a marketing ops person and called it transformation.

The reality is that templated CMSes have 8 to 12 locked sections, brittle styling, and a content workflow that requires three people to approve a comma change. The bottleneck isn't the tool. It's the people and the process the tool requires. And nobody has the political will to update a homepage when updating it means scheduling a meeting with five stakeholders.

That's why your footer still says 2025.

The new math

Here is what I am watching founders do in April 2026, with no team.

One person. One LLM trained on the company's story (call it a Brand Twin). One afternoon. Twenty production-ready pages shipped to the live site.

No mockups. No wireframes. No Figma reviews. No "let's iterate on the hero." The LLM has been trained on great design patterns. It's been trained on your brand voice. It generates the page. The founder reads it, says "tighten the second sentence," the LLM tightens it, the founder ships it. Live URL. Real traffic. Real conversion data flowing back the same day.

This is the math nobody is putting on the wall yet. One marketer with a Brand Twin in 2026 outproduces a six-person team with an agency on retainer in 2022. By an order of magnitude. With better personalization. At a tenth of the cost.

The team isn't the bottleneck. The team IS the bottleneck. There's a difference between those two sentences, and most founders haven't figured out which one applies to them.

A real example: the Tuesday I never logged into GoHighLevel

This is not theoretical. I lived this last week.

I had an idea for a six-email nurture sequence for B2B founders who'd downloaded a piece of content but hadn't booked a call. Not a complicated idea. The kind of idea I would have spent two weeks on with a marketing team in 2022.

Tuesday morning I opened a Claude agent I'd built. I told it the concept. The audience. The offer. What I wanted each email to do. Out loud, in conversation, the way you'd brief a smart colleague.

It wrote all six emails. In my voice. Not a generic "AI-sounding" voice. MY voice. Because I'd trained it on hundreds of pages of my own writing, my own decks, my own podcast transcripts, my own client calls. That's the Brand Twin part. The part most founders skip because they don't know it's the unlock.

Then it wrote the GoHighLevel integration. Set the trigger conditions. Set the timing. Mapped the segments. Returned the URLs of the live workflows ready to fire.

I never logged into GoHighLevel. Not once. The thing that used to be a four-day project for a marketing ops manager took me 45 minutes of conversation.

Here's the part most founders need to sit with. The "marketing operations manager" job that every B2B company was hiring for in 2023 just collapsed in front of me on a Tuesday morning. Not someday. Last week. And I'm not running a team. I'm having a conversation.

A business IS an experiment

Here's the bigger frame this is pointing at.

A business is an experiment. Every page you publish is a prototype. Every email is a hypothesis. Every campaign is a test you're running on the market to see what it returns.

That has always been true. It's just that the production cost of running an experiment in 2022 was so high that most companies could only afford to run two or three a quarter. So they made each one a Big Deal. Mockups, wireframes, brand reviews, sign-offs. The cost of being wrong was the entire production cycle.

In 2026, the cost of running an experiment is fifteen minutes of conversation with an LLM. Which means you can run twenty experiments before lunch. Which means the businesses that win are the ones that prototype, ship, learn, and prototype again at speed. Not the ones with the prettiest mockups.

Sam Altman has been talking publicly about a betting pool among Silicon Valley CEOs for the first one-person billion-dollar company. The conversation isn't whether it will happen. It's when. The premise underneath that conversation is the one most B2B founders have not fully internalized. Scale is no longer about more human bodies. It is about systems and intelligence. The ten-person company in 2026 will outproduce the fifty-person company from 2022. The two-person company will outproduce the ten-person company by 2028.

This is not a productivity story. It is a structural story. The shape of a company is changing. And the marketing team is the first thing that will look unrecognizable on the other side.

What this means for you

If you're a founder reading this and you have a marketing team of more than two people, here's what I'd ask you to do this week.

1. 1Footer Date Test. Open your homepage. Check the footer. If it says 2025 (or earlier), you have a Marketing Production Tax problem. The only question is how big it is.
2. 2Last Edit Test. Ask your marketing team when the homepage was last meaningfully changed. Not a copy tweak. A real change. A new section, a new offer, a new positioning beat. If the answer is more than 60 days ago, your marketing has stopped being a living surface and has become a brochure. Brochures are a 2022 concept. Surfaces are a 2026 concept.
3. 3Brand Twin Test. Open Claude or ChatGPT, tell it about your company, ask it to write a 400-word landing page in your voice. Read what comes back. If it sounds like a generic SaaS company, your story isn't trained anywhere a machine can find it. That's the work. Train your story into AI once, and the entire production cycle collapses behind it.

The 2022 Marketing Production Tax is not a small thing to be paying. It's $200K, $500K, $1M a year depending on your size. And every dollar of it is going to a production cycle that is no longer the bottleneck. The bottleneck is your story. Whether your story is documented clearly enough that an AI can produce in your voice without you in the room. Train the story. Build the Brand Twin. Wire it to the systems. The team gets smaller. The output gets bigger. The next billion-dollar company is going to have fewer than ten people, and most of them are not going to be in marketing. If you can't answer "I have a story trained into AI that produces in my voice on demand," that's the work. We do it in one session. Open Kitchen