Reach doesn't buy response.
Across 12 creatives, reach explained only 7% of emotional engagement. Buying more eyeballs guarantees distribution, but says almost nothing about whether anyone connects with the work.
We sit on 43B datapoints (growing ~1B/mo) measuring how people see, feel, and remember brands.
Diving into our meta audience benchmarks and sharing the findings here all week.
The cheapest creator wasn't.
Looking at sticker CPM vs Emotional CPM, the lowest rate on paper often delivers the least impact. Paying less for distribution gets expensive when viewers scroll past without feeling anything.
We sit on 43B datapoints (growing ~1B/mo) measuring how people see, feel, and remember brands.
Diving into our meta audience benchmarks and sharing the findings here all week.
Half the room wasn't listening.
In-feed tests with consented viewers show a brutal drop-off: people scrolled right past, or saw it and never engaged into memory.
You pay full price for impressions, but the audience never registered the message.
We sit on 43B datapoints (growing ~1B/mo) measuring how people see, feel, and remember brands.
Diving into our meta audience benchmarks and sharing the findings here all week.
Same engagement, wildly different offers - creators of colour consistently on the wrong end. Osman Badat (The Social Accountant), via @Digiday.
Markets with no measure of value price on assumptions. Opacity gives bias somewhere to hide.
Measured impact is a leveller.
13 years of work, 43 billion datapoints, and this week we cancelled our public AI launch.
The original plan was self-serve. Let the internet find the rough edges, break things in the wild, and report the bugs.
Instead, our services team gets the model first. They will run it behind the scenes on live client work, catch the edge cases in private, and run the real evals while keeping our gold-standard delivery intact.
If you are an existing client, expect a knock on the door soon.
@jamesnord Like the predictive sizing model you have.
Not tracking how prediction of view = inspection of property.
100K views can walk by the house, doesn't mean anyone purchases it. How is the model inspecting the quality of the view? Help me connect those dots.
@jamesnord's real estate analogy for creator pricing is right - with one addition.
House comps work because houses get inspected first. Creator comps built on follower counts are comps without the inspection.
Transparency on price needs measurement of value underneath it.
Six months after anyone leaves Element Human, I send them a message.
No matter how spicy the exit was at the time.
When you run a flat company, the biggest risk is people filtering bad news before it reaches you. I used to sit in leadership updates listening to clean summaries, and I'd stop the room to ask what our frontline delivery lead actually thought. She was the one dealing with broken workflows and client friction every day. I wanted the ground truth from the person doing the work.
When someone moves on, the internal politics disappear. By month six, the dust has settled, they are thriving in a new role, and they will tell you the exact operational warts you were too close to see.
If you hired smart people, they already know where the cracks are. You just have to be willing to ask when they have zero incentive to protect your ego.
I had one of those catchups this morning over coffee. Still taking notes.
Creator marketing is making the exact same mistake sports sponsorship made 20 years ago.
Everyone prices reach because it's easy to count. Then checks get huge and CFOs realize reach doesn't mean people gave a damn.
At IMG, we had to negotiate an "intangibles" row on a spreadsheet to price feeling.
@HarleyBlock is right: tech pricing inputs isn't the driver. Measured emotion is.
Yesterday we cancelled the public launch of our emotion AI, built on 13 years of work and 43 billion datapoints.
Our services team gets the tools first. They are running the model behind the scenes on live campaigns, catching the weird edge cases in private while keeping human delivery sharp.
If you already work with us, expect a knock on the door next week.
Agencies aren't the villain in creator pricing opacity.
Their margins are shrinking trying to solve it with guesswork at scale.
Pricing is built on views and clicks, but real brand work happens in between.
If the audience isn't in a state to remember the brand, nothing moves.
Measure audience readiness, and the guesswork stops for everyone.
4 creators, 1 campaign, identical attention. Everyone got "the view."
One drove zero emotional response and zero brand recall. The others: 4 - 6% uplift (where the median for all content is zero).
Same invoice. Not the same product.
That's the pricing problem @alyssa_merc nailed in @Digiday.
If an email actually matters, I never let Claude touch it.
Consuming empty calories all week makes you numb. When an LLM generates three polite paragraphs just to ask for a decision, you feel the laziness.
If you couldn't bother to write it, I won't bother to read it.
Anyone else seeing AI scale the waste in marketing?
Just reviewed 200 client calls and ran 47 product interviews and it seems the most expensive decisions are happening before the campaign even starts or the creator, content, and media investment is actually made.
@OVOEnergy hi there, struggling to get in touch. Webchat link is down, phone lines are always busy. What's the best way to get in contact that won't take hours?
Being conscious of your emotional responses makes you a better liar.
Be relaxed, don't raise your voice, don't deny without reasoning.
https://t.co/sUxLaiOsFj https://t.co/sUxLaiOsFj
New on the blog today: Why we want to teach emotions to machines - an excerpt from @JinghanOng's @TEDTalks
https://t.co/Bpjnf9FRGJ https://t.co/Bpjnf9FRGJ