Earlier this year ICONIQ surveyed a hundred and fifty B2B go-to-market leaders and found that AI-forward startups put about 13% of their headcount into marketing versus 17% everywhere else, run 20 to 30% leaner across the board, and generate close to double the revenue per person. The conclusion most people are drawing from that is that startups don’t need a marketing hire anymore, and it’s easy to see why. A few hundred dollars a month buys a stack that writes the blog, drafts the newsletter, schedules the posts and builds the dashboard, which is more or less the job description of the junior generalist most startups used to hire first.
Buried further down in the same report is the number nobody is quoting. Marketing headcount is still growing 23% a year once companies pass $10M in revenue. The hire isn’t disappearing. It’s arriving later, in a different shape, with a different job, and what that job turns into is the question we’ve built our company around.
Everything is going through the wash
Start with what it feels like to be on the receiving end of marketing right now.
We talk to investors who open a hundred decks a month and tell us that lately there’s so much happening on every page, so many words and stats and boxes, that they can’t tell what the company does, so they feed the deck into an AI and ask it to explain in plain language. A founder used AI to generate a deck that an investor used AI to decode, and somewhere in the middle the actual thinking got lost.
That’s the pattern everywhere. Content goes through the wash three times and comes out grey. Every cold email reads like every other cold email. Every logo is slowly becoming the same logo. Someone says GEO is the new thing so the team buys a GEO tool, someone says founder content so the founder starts posting daily and hasn’t talked to a customer in a month, and none of it is connected to anything the business actually learned about itself.
People are exhausted by it, and exhaustion is where taste comes back. In a few years, “a human actually made this” is going to be the most valuable thing you can say about a piece of work. Not because AI gets worse. Because it gets everywhere, and when everyone can generate a hundred things a week, the thing a person clearly spent a day on becomes the rare thing again. Shopify spent a year telling every employee to use as much AI as possible, measured them on it, and then a few months ago quietly said it was creating more work than it saved. That correction is already starting, and it’s going to spread.
The real problem isn’t the output. It’s the input.
Here’s the part we think most of the industry has backwards.
The reason marketers reach for the generic output isn’t laziness or lack of talent. It’s that they don’t have the information to do anything better. Most of a marketer’s week is spent hunting. What did the founder say on that customer call. What changed in the positioning. Which campaign actually worked last quarter, and why. What did the competitor ship on Tuesday. You go looking, you ask around, you find last week’s version of the doc, you make the call with 60% of the picture, and then you generate something and hope.
You don’t know the one thing that’s going to work for your business until you’ve tried enough things to see the pattern, and you cannot see the pattern when the information lives across six tools and four people’s heads. Every AI marketing tool on the market today makes the generating part faster. Almost none of them make the knowing part better, and that’s why the output all looks the same. Same generic context in, same generic content out.
What we think a marketing team looks like in five years
Not a department. A few people, some full time and some part-time, because marketing has always been seasonal and lumpy and pretending otherwise is why so many teams are the wrong size. Spread across time zones. Some of them vendors, some of them fractional, some of them the founder on a Sunday.
All of them plugged into one brain that knows the business better than any of them does alone.
Not because someone maintains it. Because it learns. The positioning shifts and the brain already knows. A campaign runs and it learns what worked. A customer says something on a call and it’s there for the content person the next morning without anyone sending a Slack message about it. Everyone on the team makes decisions from the same picture, and nobody has to go looking for it or remember to check.
Which is why we think dashboards mostly disappear. A dashboard is a thing you have to remember to look at, and the whole point is that you shouldn’t have to remember. The brain lives wherever the team already is. In Slack, in a text, on a phone call, inside whatever tools they open first in the morning. It works around the clock, and it shows up with what you should be doing next, sometimes before you knew to ask.
Project management, as a category, is more or less dead in that world. The tools that kept a team on task were only ever a workaround for the fact that nobody had the same information at the same time. Give everyone the same picture and most of that scaffolding falls away, and what’s left is a small number of people making good decisions quickly and then spending their time on the parts that genuinely need a person. The video only the founder can film. The launch copy where the hard part is deciding what not to say. The conversation with the customer who’s about to churn.
What we’re building
Supercurve is the brain in that picture. A teammate for your marketing team that knows your business, learns from everything the business does, and gets smarter every time you use it.
It is deliberately not another tool to add to the pile. We don’t think the world needs a better GEO tool or a better social scheduler, and we’re not trying to be the eleventh tab open in your browser. We’re building the shared context underneath all of that, the thing every other tool and every person on the team pulls from, so the work that comes out the other end is specific to your business instead of specific to whatever the model saw most of on the internet.
The bet is simple. Marketing teams are going to get smaller and that’s fine. The teams that win won’t be the ones with the fewest people. They’ll be the ones where everyone, human or otherwise, is working from the same understanding of the business, so the humans can spend their time on the only part that was ever really theirs.
P.S. We wrote this using Supercurve because it knows our philosophy and it knows our voice.
