OpenAI’s Statsig acquisition and advertising strategy

Was OpenAI’s $1.2 billion Statsig acquisition last week “next steps” in building an advertising strategy for its answer engine, ChatGPT? There are signs. Read the OpenAI blog post (9/2) for background.

First, Statsig’s optimization tools and its executive leadership team—including the company’s founder, Vijaye Raji (formerly of Meta), who will become CTO of Applications at OpenAI— offers plenty to think about. Mr. Raji’s LinkedIn bio states that his early Meta days were explicitly in advertising when he worked on Facebook Audience Network and “Mobile App Install Ads.”

Also, the acquisition puts Mr. Raji on the same team as CEO of Applications, Fidji Simo, also formerly of Meta and a key player in monetization strategy at the social giant.

An anon named “chambers” on Hacker News offered insights about Statsig post-acquisition last week:

“Statsig’s core value is their experimentation platform —the automation of Data Science. Big Tech teams want to ship features fast, but measuring impact is messy. It usually requires experiments and traditionally every experiment needed one Data Scientist (DS) to ensure statistical validity, i.e., ‘can we trust these numbers?’. (…)

Separating real results from random noise is the meaning of “statsig” / “statistically significant”. I think it’s similar to how companies define their own metrics (their sense of reality) while the platform manages the underlying statistical and data complexity. The ideal outcome is less DS needed, less crufty tooling to work around, less statistics learning, and crucially, more trust & shared oversight. But it comes at considerable, unsaid cost as well.

Is Statsig worth $1B to OpenAI? Maybe. There’s an art & science to product development, and Facebook’s experimentation platform was central to their science. But it could be premature. I personally think experimentation as an ideology best fits optimization spaces that previously achieved strong product-market fit ages ago (…)”

According to Statsig’s website, the company specifically addresses three industries today: eCommerce, B2B Saas and Gaming.

Pitching its eCommerce solution, Statsig speaks clearly to an advertising-like opportunity:

“Fine-tune your online storefront with Statsig’s feature gates, delivering the right experience to the right customer at the right time.”

And on Statsig’s own blog post announcing OpenAI’s acquisition, the company explained, “… we’ve built a platform that spans product experimentation, feature flags, product analytics, session replays, marketing experiments, web analytics, and more.” Read more. (9/2)

Marketing experiments – makes sense as OpenAI “tests & learns” about balancing user trust and the brand advertiser’s needs.

From tipsheet: Ads are coming to OpenAI. It’s just a matter of time and what format.

Agentic commerce is “black box” commerce

On the Mobile Dev Memo podcast last week, analyst Eric Seufert commiserated with fellow analyst Andrew Lipsman about agentic commerce – neither is a believer in how it is defined today.

Mr. Lipsman explained (lightly edited):

“First of all, I think this is a hype cycle, unlike which I’ve seen in a while. It’s become something of a runaway freight train over the last several months. And I found myself scratching my head at it most of the time.

I am not a believer. And I feel I felt like I’m kind of on an island, which is why when I saw your LinkedIn post, it was nice to see somebody also articulate a bit of skepticism with something that everybody thinks is a foregone conclusion.

The definition that I’m using, and I’ll distill it to one sentence, is “agentic commerce
is a revolutionary model of shopping where autonomous AI agents act on behalf of consumers to handle the entire buying process.

So the entire buying process to me is the operative phrase because it’s taking the pain out of shopping and buying. Theoretically, and to me, it’s not really agentic if it’s just certain parts of the process, right?

At that point, very quickly, you just get to the evolution of search.

And listen, AI is gonna change a lot of parts of the path to purchase, but it will never actually go out and anticipate your needs and go and buy things for you. That’s just not gonna happen.”

Mr. Lipsman is a former eMarketer analyst and runs a newsletter called “Media, Ads + Commerce” on the retail media space.

Hear more on the Apple Podcasts app. (9/4)

From tipsheet: This is a great discussion on the reality of the agentic commerce opportunity today. I couldn’t agree more with Seufert and Lipsman who articulately explain that (seller) agents selling to (consumer) agents isn’t out there and seems hard to believe.

On the other hand, perhaps this is all semantics. Let me explain.

First, I submit to you that agentic commerce today is merely affiliate marketing and publishing in a “black box”… the “black box” is the answer engine’s AI.

The AI has figured out two things:

1. What the consumer wants (in a buyer agent kinda way) and

2. What product fits which the consumer (in a seller agent kinda way).

The OpenAI/Shopify partnership (from April) teases this type of affiliate marketing/publishing opportunity. Both companies are hoping for a significant revenue boost, and —I’d say— it’s coming.

We are moving into a “black box” world. I don’t see transparency as a winner. AI only gets more powerful, dense, convoluted and expensive to ever unravel.

That said, you want an AI transparency product? I’m sure OpenAI (with its Statsig acquisition?) or an ad tech company will make you something so you feel better. And I bet it will be better than any attribution and/or analytics product we’ve seen before.

Please note: I am not advocating for the “black box.” I understand why this is a scary proposition: loss of control. But, I’m observing that we’re moving into a new reality.

This is the rubicon that OpenAI is about to cross with ChatGPT:

• ChatGPT is going to generate referral fees on product transactions it helps facilitate.
• ChatGPT’s AI will be recommending these products because they are the best available option for the consumer.
• ChatGPT will be making a choice for the consumer and we/consumers won’t know why.
• And we (consumers) are going to live with it because it “feels” like the right choice and because, ultimately, we trust the AI.

There… I said it.

Agentic commerce is “black box” commerce.

And there still will be advertising. Billions, then trillions.

---

Headcount

• “Whenever I talk with people about AI, one of the first questions I get is almost always: what will this mean for my job? I believe AI will open up more opportunities than any technology in history – but it will also change how work gets done…” – Fidji Simo, CEO of Applications, OpenAI on her company’s new jobs initiative. Read more on LinkedIn. (9/5)

---

Policy and politics

An NC senator’s words were manipulated by AI in an ad. Now she’s suing (9/7) – Charlotte Observer

Funds

• OpenAI expects business to burn $115 billion through 2029: Report (9/6) – The Economic Times
• “Anthropic to pay authors $1.5B to settle lawsuit over pirated chatbot training material” (9/5) – NPR
• Sources: ASML is committing €1.3B for Mistral’s €1.7B Series C funding round, becoming the top shareholder, and is expected to get a board seat (9/7) – Reuters (subscription)

---

More

• Scope3 Lays Off Staff as Part of Shift Toward Agentic Advertising (9/5) – Adweek
• Vizrt Launches New AI-Powered Features to Speed up AR Sports Production (9/5) – TV Tech
• The Varanasi Municipal Corporation (VMC) adopts AI tech to crack down on illegal out-of-home ads – The Times of India
• European Commission fines Google €2.95 billion over abusive practices in online advertising technology (9/4) – European Commission
• The Complete Guide to Using AI as a Marketing Professional in Germany in 2025 (9/6) – nucamp