—
The short answer
There’s a version of 2026 where every company is producing more content than ever before, and nobody can remember why they started. AI has made creation cheap and fast, which is genuinely useful. But cheap and fast without direction produces volume, not value. The companies pulling ahead right now aren’t the ones with the most AI tools in their stack. They’re the ones who got clear on their audience first, then used AI to serve that clarity at speed. Strategy still has to come from humans who understand the customer. Once that’s in place, tools like Plot or Midjourney or whatever else you’re using stop being shortcuts and start being amplifiers of something that actually means something.
—
Does AI change who you’re creating content for?
The mechanics of content production have changed dramatically. Writing a first draft, generating a visual concept, producing a video script, all of that is faster and cheaper than it was two years ago. The question of who you’re talking to and what they actually need to hear remains exactly as it was, and it still has to be answered before anything useful gets made.
At Shapes, we use Plot for a significant chunk of our written content, and we use AI tools to accelerate parts of the visual work. But every single piece still gets pinned against the same questions. Who is this for? What are they trying to solve? What do they already believe, and what do we need to shift? The ICP work, the persona thinking, the audience mapping. None of that has been replaced. If anything, it’s become more critical, because without it, AI just produces articulate noise.
—
So what does the data actually say is happening?
The 2025 Metricool data gives a useful picture of where content is landing right now, and it cuts through some of the received wisdom about AI-generated content simply flooding feeds to success.
Reach on social platforms isn’t growing uniformly. LinkedIn specifically saw only 17.68% of accounts grow their follower count in 2024, which is the worst performance of any major platform. YouTube was 41%, Instagram 39%, TikTok 34%. The implication is that content volume alone doesn’t move the needle on LinkedIn. The accounts that are growing are the ones publishing content that earns attention, not just content that exists.
Carousels are generating 45.85% engagement rates with an average of 791 interactions, significantly ahead of text-only posts. That result reflects a genuine audience preference. People on LinkedIn want to learn something specific in a structured way, and producing that well requires knowing what your audience is trying to understand, which brings everything back to the strategy before the tool.
—
Why does this matter so much right now?
Because 2026 is the year the AI content flood is fully visible to the people reading it.
The Metricool data shows that video content grew 53% year-on-year on LinkedIn, with an 87% engagement increase and impressions growing even faster at 73%. The algorithm is rewarding it. But the same data shows only 17.68% of LinkedIn accounts grew their following at all. Those two facts sitting together tell a clear story: the algorithm is pushing more content to more people, but audiences are still selective about who they follow and what they engage with.
The brands that are winning kept their strategic foundation tight while letting AI handle the production acceleration. The ones losing are treating AI output as strategy itself, producing content that looks polished but feels generic. Readers notice this, even if they can’t articulate it, and they move on.
—
What can you actually do about this?
Before opening any AI tool, build your audience brief. The ICP work doesn’t have to be a 40-page document. Three specific descriptions of real types of people you’re trying to reach, what they’re worried about this quarter and what they’re trying to achieve, is enough to give every content decision something to push against. Every piece you create, written or visual, goes back to that brief.
Use the Metricool format data purposefully. If carousels are your best tool for reaching company owners on LinkedIn, build your AI-generated content toward carousel format with a specific educational angle, rather than repurposing a blog post with no structural thought behind it.
Check the algorithm signals for your platform and match your format choices to them. Thursday late afternoon for LinkedIn. Video if your account is growing. Polls if you want reach without a large existing following, given the data showing 206% higher reach for a format used by fewer than 0.04% of accounts. The format choices matter, and they should follow from knowing your audience, not from picking whatever the AI produces most easily.
—
Want help building the strategy before the content?
That’s the work we do at Shapes. If you’re producing content with AI tools and you’re not sure it’s landing with the right people, the issue is almost always upstream of the tool. Get in touch and we’ll show you where the gaps are.
