SBR 75: Social Media Isn't Broken. BUT Your Strategy For It Is.

A couple of months ago we went to a fascinating talk by Inge Hunter from Clue Labs on how the algorithm actually works now. Not the marketing conference version (all case studies and optimism) but the real deal, with proper mechanics. She talked about how AI decides who sees what, how money flows and what the algorithm is actually rewarding.

It was, frankly, a bit unsettling. Not because social media is dying. It isn't. But because the way most brands are using it is now almost perfectly wrong. Here are our top insights from her talk and what to do about them.

The pipe is now the same for everyone

For years, paid and organic social operated as separate systems. You could build an audience for free, then layer paid on top when you had budget. That's gone. Meta has unified everything into a single AI architecture. Paid and organic now compete for the same distribution. The algorithm decides who sees your content and (newsflash), unsuprisingly it increasingly favours those who pay to play.

For brands with small budgets trying to replicate what worked five years ago, this is a structural problem. The DTC playbook that worked for early movers was built on economics that no longer exist.

The algorithm rewards the opposite of what most brands do

Social media platforms’ AI build what's called a "relevance profile" for every account. That’s a rolling picture of your brand, tone, and audience. It uses this to decide how widely to distribute your content. And consistency builds the profile. Deviation destroys it.

Departures in topic, style, or posting schedule shrink your distribution. The platform retains only 84 days of your data so every piece of off-brand content isn't just a missed post, it's actively damaging your account's ability to reach anyone. Most brands treat social like a place to be creative, timely, and varied. That might be rewarding for the creator, but I’m afraid the algorithm treats that as noise.

You're measuring the wrong things

Likes and follower counts feel good to you. It feels validating, right? The algorithm doesn't care about them much. What it actually values is: saves, shares, watch time over three seconds, profile visits, and link clicks. A post with twelve saves on a 500-follower account will outperform a post with 800 likes on a 50,000-follower account in terms of distribution.

The brief for every piece of your content should be "will someone save this to come back to, or send it to a friend?" Not "will our audience enjoy this." Those are different briefs. They produce different content and da-algo values them differently.

constraint is actually A SUPERPOWER FOR SCALING BRANDS

The behaviours the algorithm rewards: narrow focus, consistent tone, predictable rhythm, ruthlessly specific audience, are exactly what low-budget brands are forced into anyway. The brands getting penalised are the well-resourced ones producing glossy, varied, multi-audience content.

You can't outspend the algorithm. But you can outpatience it. Pick one audience. One tone. One posting rhythm. Run it for 84 days. Measure saves and shares. Adjust. Repeat.

So should you abandon social?

No. But use it for the right job. Stop using it as an acquisition channel. That's where the money goes and where the returns are disappearing fastest. Rather, use it as three things: (1) a proof layer (showing real traction to press, partners, and investors), (2) a retention tool (giving existing fans content worth saving and sharing), and (3) a credibility signal (making whatever else you're doing look legitimate).

Real growth has to come from somewhere the algorithm can't touch. Earned media. Word of mouth. Real-world presence. Owned channels, especially email, which remains the most reliable revenue driver for almost every brand we look at.

Social didn't stop working. It stopped being free. The brands that understand that distinction will spend their budget in the right places. The ones that don't will keep boosting posts and wondering why nothing's shifting.

All credit for these insights goes to Inge Hunter from Clue Labs, who runs a free weekly Zoom session on the algorithm with the very latest thinking. We'd strongly recommend following her and giving her a listen — she was absolutely fascinating. You can find her on Instagram and LinkedIn.