After spending nearly $150 million of his own money to buy stock in his AI-enabled ad technology firm, The Trade Desk, early last week, CEO Jeff Green explained on Friday why he decided to make such a bullish bet on his own company.
In an op-ed on his company’s publication, The Current, he offered 10 reasons for the purchase beginning with AI:
- “The AI present.”
- “The bigger TAM.”
- “Wall Street is wrong.”
- “The trade press is wrong.”
- “Open will make a comeback.”
- “Our innovation in OpenTTD and agentic.”
- “Measurement.”
- “Amazon DSP is overrated.”
- “Our people.”
- “Clients and partners.”
Read: “Why I spent $150M on my own company??” – Jeff Green, CEO, The Trade Desk on his company’s The Current
From tipsheet: Perhaps most controversial of his opinions is that Amazon DSP will cease to exist in five years. He thinks Amazon will ultimately pursue its owned & operated properties only rather than including the open web in the mix.
I will suggest that the connected TV opportunity may entice Amazon DSP to continue beyond its O&O at the very least. But, Green might argue that companies like Netflix and Roku who are selling CTV inventory through Amazon DSP today will move elsewhere at some point.
Or, more simply, Green could argue that even CTV won’t be worth Amazon’s time.
Gonna be a great next five years (and beyond). Bring it on!
LLMs & CHATBOTS
Developments
- Codex Security: now in research preview (March 6) – OpenAI
- “Services: The New Software” (March 5) – Sequoia
- Partnering with Mozilla to improve Firefox’s security (March 6) – Anthropic
SEARCH
Google VP Liz Reid on Search’s future
“During our chat, Liz Reid toed the company line: what AI is doing to Google Search is ‘expansionary,’ not zero-sum, and people are asking more questions than ever. But she also said something I haven’t heard from a Google executive before. Asked whether Google Search and Gemini, the company’s standalone AI chatbot, will ever fully merge, she was unusually candid: ‘I don’t know the answer…’”
Read: “Liz Reid on where Google Search ends and Gemini begins” – Alex Heath, reporter, Sources
Listen: “‘What happens to Google when AI answers everything?’ with Google’s Liz Reid” (March 6) – Access on Apple Podcasts app
TECH
4 AI takeaways: Morgan Stanley conference
Last week, Wall Street investment firm Morgan Stanley (MS) had its annual, buzzy Technology, Media & Telecom (TMT) conference in San Francisco with 400 companies and 2,000 investors participating, according to MS Chief Client Officer Mandell Crawley on LinkedIn.
Here are four highlights relevant to the AI advertising and marketing industry…
1. Agentic commerce
Michelle Weaver, equity analyst, Morgan Stanley:
“It was mentioned a few times at the at the event last year, but this year, anyone who was an enabler of agentic commerce — from Visa to Affirm — were all talking about how they can benefit off of this [trend]. Agentic commerce was featured really heavily in Shopify’s presentation. We’re still in early innings for consumer adoption of agentic commerce, but I think it’s something to absolutely watch this year.”
2. Meta’s AI momentum
Meta CFO Susan Li expressed surprise at Meta’s “continued ability to find ways to improve its ad performance,” according to MBI Deep Dives newsletter’s conference coverage. AI is likely the fuel thanks to Meta’s enormous capex investment.
Ms. Li said at the conference:
“…we have an internal metric called iRev, which is basically how we measure the performance of ads. ‘And here’s the list… and here’s what they add up to.’
And it is, I think, one of the maybe modern wonders of the world that we have continued to generate basically ‘half after half’ a list of improvements that continue to generate IREV gains every half and those continue to compound on each other…that’s true on the organic side, too. And I would say the core business is very healthy.”
Glossary:
- ‘half after half’ = sequential six-month periods
- iRev = internal revenue per engagement or “how much revenue the platform generates from user engagement with content and ads,” according to Investing[dot]com.
Later, Ms. Li reminded the audience, “… our goal every day is to be the best place advertisers can come and spend their money relative to anywhere else.”
Read:
-
Morgan Stanley conference coverage – MBI Deep Dives (March 7)
-
Meta at Morgan Stanley Conference: AI Strategy and Financial Insights – Investing[dot]com. (March 4)
3. AppLovin eyes eCommerce, conversions
“CEO Adam Foroughi and CFO Matt Stumpf provided an optimistic outlook, specifically highlighting the potential for the company’s advertising technology to expand beyond its core gaming market into the much larger e-commerce sector. The CFO stated there is ‘definitely potential upside’ to their 20%-30% growth targets for gaming ads, and the CEO noted the addressable market outside of gaming could be ‘5x to 10x’ [larger]…”
Morgan Stanley analyst Matthew Cost observed of his interview with Messrs. Foroughi and Stumpf (lightly edited):
“I think AppLovin is probably the easiest ‘AI winner’ story in my space, and probably the one that’s the most disrupted right now for arguably the weakest reasons.
Obviously, there’s concern about the video game industry being disrupted, but we were able to ask the company the question, ‘What if video game production was disrupted? What if many more games are being made by many more people, what would it mean for AppLovin?’ And their argument is, as an advertising business that serves the video game industry, they’re in a position to provide discovery and to help get the right game in front of the right person at the right time. In a market where there’s more games and more competition, the value of what AppLovin does could actually go up.
(…) And AppLovin executives gave a new stat — which was the first time we’ve heard one on this metric in a year or two — which is that they’ve gone from 1% or below conversion rates to 1.3%. Pretty significant uplift in conversion.”
More: “AppLovin Corporation (APP) Presents at Morgan Stanley Technology, Media & Telecom Conference 2026 Transcript” (March 4) – Seeking Alpha (subscription)
4. Model improvement momentum
Stephen Byrd, equity analyst, Morgan Stanley, said about the conference:
“I think the discussion we’ll have 12 months from now, in terms of the scope the capabilities of AI will be far greater than where we are today. And so I would just urge all of our investors to think very carefully about that non-linear rate of model improvement, what that might really mean, because it’s really going to be a big driver of stock performance over the next 12 months.”
DATA
LiveRamp’s agentic plan at RampUp
At LiveRamp’s RampUp conference last week, the data collaboration technology company announced new “agent-powered” marketing updates to its platform. LiveRamp’s Chief Product Officer Matt Karasick said in the release:
“We’re making it possible for AI agents to do what marketers have been doing manually — build audiences, measure cross-media performance, and optimize spend — but faster and within the governed environment our customers already trust.”
An expansion of its agentic platform plan first announced at CES in January, LiveRamp detailed the two newest agentic partners:
-
“SemantIQ, health and life science marketers can build and activate healthcare provider audiences from the LiveRamp Clean Room.”
-
“Using Newton Research, marketers can unlock instant measurement insights from LiveRamp’s Cross-Media Intelligence using simple natural language questions.”
LiveRamp promises “more agents to come.”
Read: “LiveRamp Launches Agentic AI Upgrades to Power Smarter Growth, Planning, and Measurement” (March 3) – press release
More: “After two days at the conference, here are three takeaways that stood out.” (March 6) – Sell-side advisor James Deaker on LinkedIn
MARKETING
Thought leadership on GEO
- “When People Never See Your App: Designing Brands for the Agentic Economy” (March 2026) – Bain & Company
- “The dual-domain strategy is architecture, not a tactic.” (March 5) – Rishabh Jain, CEO, Fermat Commerce on LinkedIn
- “How companies can navigate three major shifts from LLMs encroaching on search.” (March 7) – Ganna Pogrebna, consultant, regarding her HBR article on LinkedIn
RESEARCH
Agentic commerce opportunity
- “The Collapsible Funnel: Part 2 — A New Model for the Shopper Journey in the Age of AI“ (March 6) – Andrew Lipsman on Media, Ads + Commerce (sign up required)
- “What It Means That The Leader In ‘Agentic Commerce’ Just Pulled Back” (March 7) – Emily Pfeiffer, Principal Analyst on the Forrester blog
AGENTS
Meta’s Manus and its annual letter
Manus gently put itself back in the news again late last week.
As you may recall, in less than three months, Manus — the startup acquired by Meta in December — has already started to optimize the workflow in Meta’s most important technology product: its AI-enabled ad platform.
What can it do for Meta in another year? Probably quite a lot. A new annual letter from Manus on Friday suggested as much.
The letter began with a quote for Manus’ leader:
“Manus was born from an idea that seems almost unremarkable today: AI shouldn’t just think, it should also do.”
Read: “Manus Year One: Our First Annual Letter” (March 6) – Manus
From tipsheet: It’s interesting to note the emerging autonomy of the Manus team in public messaging as well as its sense of independence within Meta broadly.
It may speak to the terms of their transaction with Meta as well as the entrepreneurial culture at the startup. Meta CEO Mark Zuckerberg likely has his hands in this directly, too. He likes what he sees and wants to keep it together, focused.
The anecdotes shared in the letter provide the sense that Manus, an AI agent platform, just needs to know which problems to solve for and it will do the rest.
Perhaps this is Meta’s AI “red team” for ads or any other company product where workflow can be streamlined.
The publication of a second Manus annual letter in March 2027 will be telling.
AGENTS
Former TTD-er on Amazon MCP integration
Ash Gangwar, founder and CEO of Neural Ads announced a new Amazon Ads integration for the startup.
He explained, “The way media teams interact with their platforms is changing — Neural Ads integration with the Amazon Ads MCP server means AI agents can now talk directly to the platform in plain English, get answers, and get things done. For the first time, strategists, planners, and traders all have the same level of transparency and control over campaign outcomes.”
Read more from Mr. Gangwar on LinkedIn. (March 7)
More: “We just made Amazon Ads conversational.” – Neural Ads on LinkedIn (March 6)
Gangwar was formerly GM, CTV at The Trade Desk, according to his LinkedIn profile.
From tipsheet: Amazon Ads rolled out the “open beta” of its MCP server integration in early February.
LLMs & CHATBOTS
‘Why OpenAI-Criteo partnership’
“Instead of starting with Amazon, Google, or Meta, companies that would bring their entire ad stack, OpenAI partnered with a commerce-focused platform that brings advertiser demand without controlling a consumer discovery surface.
That gives OpenAI room to experiment with how sponsored recommendations appear, how commercial signals influence ranking, and how attribution works when the answer itself becomes the recommendation.”
Read more from Mr. Moy on LinkedIn.
Moy runs business development and ads partnerships at Reddit.
PEOPLE MOVES
Now hiring
-
AI-native business intelligence firm Rill Data has added Devan (Fearman) Kane as CMO, Jay Stevens as CRO, among others – Devan Kane on LinkedIn
-
“I’m excited to share that I’ve joined A+E Global Media as VP, Programmatic.” – Joseph Lerner on LinkedIn
MORE
- Axel Springer acquires the Telegraph (March 6) – Axios
- “The Brand Age” (March 2026) – Paul Graham, Founder at Y Combinator, on his personal blog
- YouTube Americas Leader Tara Walpert Levy Says Measurement Proves Creators Do TV Ads Best (March 6) – AdExchanger
- “it is heartening to see major media and measurement companies like Comscore and NBCU integrate HyphaMetrics into their businesses.” (March 7) – Kris Magel, VP, FreeWheel on LinkedIn
- “Automating creative to unlock scale” (March 4) – “Brian” on AppLovin’s Axon blog