There’s a conversation I keep having with founders right now, and it goes something like this.
They’ve been using AI to accelerate marketing. Content is going out faster than ever. Ads are being tested in volume. SEO briefs are produced in minutes. The team feels productive.
But something isn’t working. The pipeline looks busy. Conversion is flat.
They’ll usually frame it as an AI problem. Maybe the output isn’t good enough. Maybe they need a better prompt. Maybe the model isn’t right for their use case.
It’s not an AI problem.
It’s a strategy problem, and AI just made it visible.
The Amplification Effect Nobody Warned You About
AI is the most powerful amplifier the marketing world has ever seen. But amplification doesn’t care about direction. It scales whatever you point it at.
If you have clear positioning, a defined ICP, and messaging that resonates with a specific buyer psychology, AI can help you produce that at 10x the volume and half the cost. That’s a genuine competitive advantage.
If you don’t have those things, if you’re operating on vibes and broad assumptions about who you’re selling to, AI will help you produce that at 10x the volume, too. You’ll just be broadcasting noise faster than you ever could have manually.
The bottleneck in 2026 isn’t content production. It’s strategic clarity. AI didn’t create that bottleneck. It removed every excuse for not confronting it.
This is the thing most startup growth conversations are still dancing around. Teams are debating which AI tools to use, what the right posting cadence is, and whether to double down on SEO or go all-in on paid. But the underlying question, who exactly are we trying to reach, what do they actually care about, and why would they choose us, often doesn’t have a crisp answer.
When content was expensive and slow, you could survive a fuzzy strategy by just doing less. Every piece had to count, so teams were forced into tighter thinking by resource constraints. AI removed that constraint. And for a lot of startups, it removed the last thing that was forcing strategic discipline.
What AI-Generated Traction Actually Tells You
One pattern I’ve noticed repeatedly in 2026: startups that adopted AI early are seeing either strong compounding growth or accelerated plateau. There’s very little in between.
The ones compounding had something in place before they scaled with AI. They knew their buyer. They had tested messaging that converted. They understood which channels accumulated value over time versus which ones were pure rent. When they introduced AI, they were feeding a system that already worked.
The ones plateauing reversed the order. They started with AI-enabled production and expected the volume to generate insight. It doesn’t. Volume tells you what gets clicks. It doesn’t tell you why, or whether the right people are clicking, or whether the people who click are converting into customers who stay.
If your AI-assisted content is generating traffic but not leads, that’s a positioning signal.
If your AI-assisted ads are generating clicks but not conversions, that’s a messaging signal.
If your pipeline looks full but deals are stalling, that’s an ICP signal.
None of these are problems an AI tool will solve on its own. Some problems require a human with strategic clarity to diagnose and fix. The data AI generates is pointing to the diagnosis. Reading it correctly is the actual work.
Three Things AI Cannot Do For Your Growth Strategy
I want to be specific here, because I think a lot of founders are still holding out hope that the right AI workflow will solve what are fundamentally strategy problems. It won’t. Here’s why.
1. AI cannot define who your customer actually is
It can generate personas based on patterns in existing data. But personas built from past data reflect who you’ve historically reached, not necessarily who you should be reaching. If your early traction came from a narrow cohort, your network, early adopters, people who found you through a specific channel, then AI-generated personas will keep pointing you back at that cohort. Breaking into a new segment requires human judgment about who the opportunity actually is.
2. AI cannot tell you why your best customers chose you
This is the most valuable piece of strategic information a startup can have, and it lives in conversations, not in data. When I work with founders on growth strategy, one of the first things I do is interview their best customers. Do not survey them. Talk to them. Ask what was happening in their business the week before they signed up. What alternatives did they consider? What made them feel confident enough to move forward? The language that comes out of those conversations is irreplaceable. No AI prompt generates it. You have to go get it.
3. AI cannot decide which bets to make
Strategy is ultimately about saying no. Which channels are you not going to invest in? Which audiences are you not going to chase? Which messages are you going to stop running because they attract the wrong buyers? AI is very good at generating options. It’s not able to make the judgment calls about which options are worth pursuing, given your specific stage, team, and market dynamics. That’s leadership. That’s strategy. That’s a human decision.
What Good AI-Assisted Growth Actually Looks Like
The startups using AI most effectively in 2026 are doing something that looks almost obvious in retrospect: they’re using AI to execute against strategy, not to replace having one.
Concretely, that looks like this:
- They’ve done the positioning work first. They know their ICP at a level of specificity that would make a generalist uncomfortable. They have messaging that reflects real buyer language from real conversations.
- They use AI to produce variations of that messaging at scale, for ads, landing pages, email sequences, social, SEO content, but the core strategic substance is human-originated.
- They use AI analytics to identify signals in performance data faster than they could manually, then route that signal to human decision-makers who can act on it with judgment.
- They treat AI as infrastructure, not as strategy. It runs processes. It doesn’t set direction.
This sounds straightforward. In practice, it requires genuine discipline, especially for early-stage teams under pressure to move fast. The temptation to skip strategy and go straight to production is real. AI makes that temptation more dangerous than it’s ever been, because now you can produce the wrong thing at an industrial scale before you realize it’s wrong.
Where to Start if You’re Feeling This
If you’re a founder reading this and something is landing a little too accurately, here’s what I’d suggest before your next AI-assisted campaign:
- Pause production for a week. Look at your best customers from the last 12 months. Three calls. Real conversations. What do they say about why they chose you?
- Rewrite your ICP based on what you hear, not what you assumed going in. What’s the specific trigger that brought them to market? What’s the outcome they’re actually optimizing for?
- Run your existing marketing against that new ICP definition. How much of it is speaking directly to that person? How much of it is generic?
- Brief your AI tools with the specific language and context you just collected. Let that research drive production, not the other way around.
This process takes time. It’s not glamorous. It won’t show up in your tool stack or your OKRs. But it’s the work that makes everything downstream, including everything AI touches, dramatically more effective.
One Last Thing
AI is genuinely transforming what’s possible in growth marketing. The ceiling for what a lean startup team can execute is higher than it’s ever been. But the floor for what happens when strategy is absent is lower, too. Mediocrity now scales.
The founders who build durable growth in this environment will be the ones who invest in strategic clarity before they invest in production velocity. Everything else follows from that.
If you’re working through this and want an outside perspective on your growth strategy, positioning, or acquisition mix, that’s the exact kind of work I do with early-stage startups and VC-backed scaleups.
→ Get in touch at anthonynealmacri.com/contact
