A jerky brand's first-ever streaming campaign: $7 CPA against a $25 order value. 354% more orders.
Not platform-reported. Independently verified by A/B test.
CTV just became buyable by anyone. The hard part now is proving what you actually paid for.
New from Alkimi Research 👇
CTV used to have a bouncer. It checked whether you had the six-figure budget and the trading desk to place a buy at all.
That bouncer is gone. Anyone can run CTV now.
But the same programmatic plumbing that opened the door imported the open web's oldest habit: a fee stack you can't see.
New LabTalks below 👇
Imagine a fable 5 quality model that’s 3-4x less expensive in less than 6 months. And an Opus 4.8 grade model that can run on a local device in less than 12 months. Greater than 50% chance that these events will happen. Worth keeping in mind when you make predictions about the future.
IAB Tech Lab's Agentic Real Time Framework moves AI from the sidelines into the transaction itself.
Agents run as containers inside a platform's own infrastructure. Identity resolution, fraud detection and bid shading all happen live, before the bid clears, not after.
We're building our agentic marketplace on the same principle: agents inside the transaction, with a verifiable record of what happened.
Read more ↓
Financial markets hit this wall in the late 1960s. Settlement paperwork was closing the NYSE one day a week.
The fix wasn't better reconciliation. It was a shared record both sides could reference. The Depository Trust Company, 1973.
Advertising is at the same point now. With the same choice.
Our research with WPP shows why 👇
This matters more for advertising than people realise.
Today's ad verification burns 8-11% of media spend running thousands of redundant copies of the same data. And still can't prove any of it.
On shared verifiable state? Under 0.1%. And the verification actually verifies.
At agent speed, this stops being a nice-to-have. It becomes the only architecture that
works.
The missing primitive is verifiability. A shared record both parties can independently check and neither can quietly alter. That's what a blockchain is for.
Our research simulated 90,000+ agentic deals. Separate record-keeping produced disagreement in 95.3% of transactions. A shared, on-chain record cut that to a fraction of a percent.
Guardrails are what a system builds when it cannot verify. Verification is what lets the guardrails come down.
Full piece 👇
2/ Trust in AI-driven media buying sits at 68% when a human reviews the output.
Remove the human and it drops to 26%.
Same technology. Same capability. Entirely different confidence level.
1/ The ambition is there. Three quarters of IAB UK members are testing, piloting, or scaling agentic systems.
The infrastructure to support it is not.
47% of advertisers don't trust AI agents because of a lack of transparency in how decisions are made. Among IAB UK members, that rises to 67%.
74% of the UK ad industry is experimenting with agentic AI.
4% has actually built around it.
IAB UK's State of AI report lays out a gap that should worry anyone planning the next two years of media buying. 🧵⬇️
Some categories face severe restrictions on where they can advertise at scale. That makes premium inventory and full attention even more valuable.
Using 100% Share of Screen, Alkimi delivered full-page ownership across Bloomberg, WSJ, and the Financial Times. Zero competitor presence. 100% viewability per takeover.
The results: 41% ad recall, +17pp awareness lift above benchmark, and quarter-on-quarter improvement across every tracked metric.
The client has run continuously across multiple quarters.
https://t.co/5SqY0sWcbw
95.3% of the records did not match.
We simulated 90,202 deals negotiated between buy-side and sell-side AI agents, each keeping its own record. The same architecture every DSP and SSP runs on today.
The agents weren't faulty. The structure was. Two parties, two databases, two versions of the same deal.
Today humans manage that gap through reconciliation. Slow, but survivable. Agents remove the speed limit, and a monthly headache becomes a structural problem.
The fix: one shared record both agents work from.
Whitepaper below 👇
Agentic advertising isn't possible without a single source of truth. Every agent, every trade, every settlement has to resolve against the same record, or the whole thing collapses back into the mess we're leaving behind.
@SuiNetwork and @WalrusProtocol are how we get there. The settlement layer and the data layer that make agent-to-agent advertising actually workable at scale.
Labs is the community heart of all of this. Built with you, and the best part is still ahead.
A $24bn holding company CEO just called ad-tech intermediaries a "toll" on his Q1 earnings call.
For a decade the language was neutral. "Fragmentation." "Complexity." "Stack consolidation." The vocabulary protected the people taking the toll.
When it's a stack, you optimise it. When it's a toll, you route around it.
The conversation just shifted.
Digital advertising was supposed to be the medium you could finally prove worked. That was the whole pitch.
A click could be a bot. An accident. A person who was never your customer.
The biggest brands in the world still target everyone because they can't trust what they're measuring.
That's the problem we're fixing.
Everyone accepted that your ad would share a page with your competitors. That's just how the industry works.
Alkimi runs the exchange. We auction the whole page at once. One brand. One moment. Zero competition.
Fragmentation was never inevitable. It was just never challenged.