Social platforms started grading you on who made it | Leveret Drive

Social platforms started grading you on who made it

This stopped being a taste debate the moment it started affecting distribution.

Jon Usborne
· 7 min read
A phone held up filming a real moment

For about two years the AI content debate on social was a matter of taste. Some people found it cheap, some people found it efficient, everybody argued in the comments, and nothing material changed.

Then it started affecting reach. Different conversation.

What shifted

Over 30% of consumers now say they avoid brands that use AI-generated ads, and platforms have been adjusting their ranking models in response, penalizing synthetic content and rewarding original work (Storrito, 2026).

That is the part that matters for anyone running an account. You can hold whatever opinion you like about AI creative. If the distribution system has an opinion, your opinion is a hobby.

The production trap

Here is the bind teams are actually in, and it is real.

You need volume. Social does not reward one great post a month. The reason everyone reached for generative tools is not laziness, it is that the calendar demands more than the budget covers. That math has not changed.

So the instinct is to make the AI output better. Higher polish, better prompts, more realistic. That is exactly backwards. The accounts performing well are not producing better synthetic content. They are producing original material at a lower production standard.

Authentic over polished, not AI-generated over human-made. That is the trade, and it is a much cheaper trade than it sounds.

What lower standard actually means

It means a phone instead of a camera. It means one take. It means the office as it currently looks rather than the office after somebody cleaned it.

  • A 40-second clip of someone explaining a thing they just figured out.
  • A photo of the actual work, badly lit, taken because it was interesting and not because it was scheduled.
  • A screenshot of a question a customer asked, answered plainly.
  • The job that went wrong and what you learned, which nobody posts and everybody reads.

None of this needs a production budget. It needs someone with permission to post without three approvals, which is usually the actual blocker.

The second shift nobody is planning for

There is a quieter change happening alongside it. People are using platforms with more intent and less drift. 42% of consumers say screens feel overwhelming, and searches for digital detox rose 400% year over year (Ovative, 2026). The response has not been to log off, it has been to log on for a reason. One platform for inspiration, another for discovery, another for research before buying.

That breaks the strategy of posting the same asset everywhere. If people show up to a given platform with a specific job in mind, content that ignores the job gets scrolled regardless of how good it is.

Practically: the same idea, reformatted for what people came there to do. Research on one, discovery on another, proof on a third. Not five copies of one caption.

Posting the same thing everywhere used to be efficient. Now it is just five ways to be ignored.

What we would change this quarter

Two things, and neither costs money.

First, cut the polished output in half and replace it with material shot on a phone by someone who was there. Watch the numbers for six weeks before deciding whether it worked. It usually does, and it usually surprises the person who resisted it.

Second, give one person the authority to publish without a review chain. Every approval layer adds a day and removes a little of the specificity that made the post worth reading. The account that posts real things slowly loses to the account that posts real things quickly.

The tools are not the problem. The approval process usually is.

What to take from this
  • Over 30% of consumers say they avoid brands using AI-generated ads, and platforms are adjusting ranking models accordingly (Storrito, 2026).
  • Authenticity is now a distribution input, not just a brand preference.
  • The fix is original material at a lower production standard, not higher-polish synthetic content.
  • People use platforms with intent now. Reformat one idea for what each audience came to do instead of cross-posting.
  • Approval chains cost more reach than production quality gains. Give one person permission to publish.

Questions people ask

Are social platforms penalizing AI-generated content?

Platforms began adjusting ranking models in 2026 to reduce distribution of synthetic content and reward original work, following consumer resistance that saw over 30% of consumers report avoiding brands that use AI-generated ads (Storrito, 2026).

How do we keep publishing volume without using AI content?

Lower the production standard rather than the volume. Phone footage, single takes, unpolished photos of real work, and plain answers to customer questions cost almost nothing to produce and outperform polished synthetic content.

Should we still post the same content across every platform?

No. Audiences increasingly arrive at each platform with a specific purpose, such as inspiration, discovery, or research before buying. Reformatting one idea for each of those jobs performs better than distributing identical assets.

What is the biggest blocker to authentic social content?

Approval processes. Every review layer adds delay and removes the specificity that makes a post worth reading, so giving one person authority to publish typically improves performance more than increasing production quality.

Jon Usborne, founder of Leveret Drive
Written by

Jon Usborne, Founder

Jon runs Leveret Drive, where the team builds strategy, brand, websites, video, and demand systems for businesses that want to be the obvious choice.

Keep reading
Follow in Google

Want this kind of thinking pointed at your business?

Tell us what you are working on. We will tell you what we would do first.

Chat with us →