Google’s agentic commerce vision

John Ebbert

On the first day of the annual National Retail Federation “Big Show” in New York City yesterday, Google delivered a blizzard of announcements focused on agentic commerce.

If that “blizzard” didn’t signify the importance of agentic commerce to Google, then Alphabet CEO Sundar Pichai’s appearance and opening remarks at the NRF show did.

Mr. Pichai began:

“It’s a really dynamic moment for every industry, especially retail. That’s due in part to the AI era we’re in.

We’ve been partnering with retailers and helping them grow through technology shifts for more than two decades. And we’re excited for a new era of partnership ahead and using our differentiated full stack approach to AI innovation to help…”

Read: “The AI platform shift and the opportunity ahead for retail” – Sundar Pichai, CEO, Alphabet on Google’s The Keyword blog

For starters, similar to recent industry announcements such as those by OpenAI and Walmart, Google is bringing the retail store to the chatbot courtesy of Gemini AI. There are two sides of this strategy”:

  • Shopping Agent – for the retailer and workflow
  • Business Agent – the retailer’s consumer-facing chatbot within Google Search, more on this in a bit.

On the Google Cloud blog, Google executives explained the business-to-business (B2B) offering for retailers:

  • “Our new Shopping agent, part of the Gemini Enterprise for Customer Experience (CX) solution, uses complex reasoning and multimodal capabilities to act as a proactive digital concierge — processing text, voice, and images to autonomously build carts and execute consented actions.”

Read:

From tipsheet: The phrases “agentic commerce,” “conversational commerce” and even “agentic retail” — courtesy of a Home Depot use case for Gemini — all appear in Google’s announcements.

AI is broadening the retailer’s lexicon.

LLMs & CHATBOTS

Google: UCP, Business agent, Direct Offers

Also announced at NRF’s Big Show, a new protocol known as Universal Commerce Protocol (UCP) will underpin Google’s agentic commerce strategy.

Google Ads & Commerce VP Vidhya Srinivasan said UCP is an open standard establishing “a common language for agents and systems to operate together across consumer surfaces, businesses, and payment providers.”

She added that UCP “will soon power” checkout in Google Search’s “AI Mode” and the Gemini app.

View:

Business agent

“Business agent” is Google’s new “branded agent” used in Google Search and, according to Srinivasan, is a “new way for shoppers to chat with brands, right on Search. It’s like a virtual sales associate that can answer product questions in a brand’s voice, enabling retailers to connect with consumers during critical shopping moments and help drive sales.”

Direct Offers

Also at NRF, Google rolled out “Direct Offers” for advertisers which effectively integrates coupons (my word) as a form of advertising and incentive in the chatbot environment.

Ms. Srinivasan called it a “Google Ads pilot” on LinkedIn and explained in the release:

“As shoppers turn to AI for discovery, retailers need smarter ways to deliver value. That’s why we’re continuing to test ads in AI Mode and we’re now introducing Direct Offers. This new Google Ads pilot allows advertisers to present exclusive offers for shoppers who are ready to buy — like a special 20% off discount — directly in AI Mode.

Here’s how it works: Imagine you search ‘I’m looking for a modern, stylish rug for a high-traffic dining room. I host a lot of dinner parties, so I want something that is easy to clean.’ Google already elevates the most relevant products to meet your search criteria. But often, you are only ready to buy if you’re getting a great deal. Now relevant retailers have an opportunity to also feature a special discount. This helps you get better value and helps the retailer close the sale.”

On LinkedIn, Google VP of Global Ads Dan Taylor emphasized, “‘Offers show right within AI Mode product recommendations, advancing the conversation rather than interrupting it.”

Read: “New Tech and Tools for Retailers to Succeed in an Agentic Shopping Era” (January 11) – Vidhya Srinivasan, Google’s VP/GM Ads and Commerce, on Google Ads & Commerce blog

From tipsheet:

  • One can imagine the importance of incrementality measurement as marketers try to understand the effect of various advertising touch points (including coupons) with the consumer which ultimately leads them to Google’s chatbot or any chatbot where retailers have a commerce integration.
  • To be clear, Google’s UCP is entirely different from LiveRamp’s User Context Protocol (UCP) which was introduced last year for enabling data interoperability for advertising in the age of AI. The open-source protocol was donated in November to IAB Tech Lab for further industry development.
  • Five Google Cloud clients simultaneously announced their various forms of Gemini AI for shopping integrations including (click company name for more info): Walmart, Home Depot, Kroger, Papa John’s and Honeywell.

LLMS & CHATBOTS

Developments

  • Nvidia Hires Googler Alison Wagonfeld as Its First CMO (January 8) – The Wall Street Journal (subscription)
  • “AI Diffusion Report: Global AI Adoption in 2025—A Widening Digital Divide”(January 8) – Microsoft
  • “Behind Anthropic’s stunning growth is a sibling team that may hold the key to generative AI” (January 10) – CNBC

TECH

Scope3 leaders make the AdCP case

Last Thursday, Scope3 co-founder and COO Anne Coghlan published a “Loom” video to answer “the main question I’m getting this week: how do I see the difference between IAB Tech Lab’s agentic roadmap and the goals of AgenticAdvertising.org and AdCP (Ad Context Protocol)?”

Scope3 is a founding member of the consortium that launched the Ad Context Protocol (AdCP) last October which aims to lay at least part of the plumbing for agentic advertising.

In the 3.5-minute video, Ms. Coghlan outlines an agentic advertising world where AdCP and ARTF co-exist. She makes her argument by identifying the difference in each protocol’s “level of abstraction” and “scope” as well as how one is focused — she says — more on real-time (ARTF) and the other is targeting non-real-time environments (AdCP) such as linear (see last week’s CES announcement).

See the video on LinkedIn. (January 8)

Scope3 CEO O’Kelley weighs in

Yesterday, Scope3 CEO Brian O’Kelley provided his “take” on the need for agentic advertising and where AdCP fits.

In a post titled, “Agentic Advertising is for Allocation” on his “BOK on Ads (and Climate)” Substack, Mr. O’Kelley concluded:

“Today, most advertisers buy effectively from 3-5 platforms because execution costs scale with complexity. Instead of requiring a team to scale, agents make it a workflow. Diversify to 20 suppliers and portfolio theory does the rest.

Agentic advertising provides automation, tools, and intelligence to parts of the industry that haven’t been served well by ad tech. Agentic advertising is the allocation layer for a $1T industry that could be $2T if we improve discovery, execution, and allocation at scale. This is the performance play: not adding more compute to tweak individual bids; reallocating spend at a macro level to find billions of dollars of opportunity.”

Read more. (January 11)

From tipsheet: In spite of how “early days” agentic advertising is, the importance of getting it right now is evident among industry participants – particularly those whose livelihood depends on effective monetization of the open web.

But, a difference of opinion remains on “what protocol is next” between supporters of the Agentic Real-Time Framework (ARTF) overseen by IAB Tech Lab and AgenticAdvertising[dot]org’s Ad Context Protocol (AdCP).

SELL-SIDE

Amazon DSP executive responds on AdCP

Former Rubicon Project ad tech executive Neal Richter, who is currently Amazon DSP’s Director, Advertising Science/Engineering, provided his perspective on AdCP yesterday.

Responding to Mr. O’Kelley’s BOK on Ads Substack post on LinkedIn, Mr. Richter said:

“I’m glad you finally see this distinction between allocation and valuation. It’s why Rubicon Project back in 2014 purchased iSocket and ShinyAds to automate the outside the bid stream work. We were too early, but the designs and workflows were prescient, yet normalizing between all the APIs to provision the orders for execution was difficult. The real value was in the negotiation and packaging of inventory, and once a deal was cut, provisioning the deal for execution into buyer and seller systems. I’m still not sure that prolific inventory like the general web needs this, yet the high value Pubs and STV broadcasters do.

[IAB Tech Lab’s] Tony Katsur and the late Josh Wexler were the business leaders on this product. Many lessons learned. Including the drafting of the Programmatic Direct API of the IAB Tech Lab, and the AdCOM spec to separate the RTB transaction layer from the inventory and campaign objects. That said, I’m confused as to why AdCP needs to be a branch off an old version of MCP.”

Mr. O’Kelley responded in the comments:

“Neal Richter, not sure I get that last point. AdCP is agnostic to MCP version and is live with A2A as well.(…)

Per Claude: The claim that AdCP is based on an ‘out of date version of MCP’ is incorrect.”

See more on LinkedIn. (January 11)

EVENTS

CES observations on AI opportunity

On LinkedIn, industry executives shared their views in the context of last week’s CES show in Las Vegas and the AI opportunity in advertising and marketing.

Michael Komasinski, CEO, Criteo

  • “A consistent theme across CES this week was that the next wave of innovation is unlocking incremental growth, not replacing existing channels…” Read more. (January 10)

Tal Jacobson, CEO, Perion

  • “1) AI is moving from an assistive layer into the operating system. 2) CTV and retail media data are starting to connect more directly through clean rooms and shared planning environments. 3) And conversational discovery is beginning to behave like a real performance channel….” Read more. (January 9)

Kristy Quagliariello, VP, Programmatic, Klick Health

  • “As I come out of my post-CES haze, one thing is clear: agencies are becoming AI-powered systems. Not tools. Not dashboards. Systems that plan, buy, optimize, and learn…” Read more. (January 11)

SELL-SIDE

Akamai on piracy prevention; AI crawlers next?

Over the weekend, content delivery network Akamai firmly set its foot into the protection of publishers’ intellectual property.

Akamai CEO and co-founder Tom Leighton delivered on opinion piece titled “Safeguarding the Digital Frontier: Why Global Anti-Piracy Efforts Are Essential” on LinkedIn.

He began:

“Internet piracy is a big business. Stopping it has nothing to do with protecting free speech or geopolitical concerns. It has everything to do with billions of dollars: profit made by pirates and those that enable them on the Internet; and losses suffered by legitimate rights owners and ultimately the law-abiding viewers who have to pay higher prices as a result.”

Read more on LinkedIn. (January 10)

h/t Matthew Scott Goldstein

From tipsheet: Mr. Leighton’s argument above is a step away from the AI bot crawler protection discussion started by CEO Matthew Prince of CDN Cloudflare on behalf of publishers last June. Mr. Prince alleged at the time that AI bot crawlers were attempting to steal publisher content on a massive scale and that his company would aim to provide a solution.

From here, Akamai is well-positioned to provide a content marketplace for AI bots, too, if the content marketplace business model should prove itself to be remotely sustainable.

Back in the early days of programmatic, Akamai quietly pulled out of its advertising business (an eCommerce cookie pool) when it sold Akamai Decision Solutions in 2013 to MediaMath. Time for a comeback?

MORE

  • “How does Netflix predict what you’re in the mood for?” (January 11) – Analyst Eric Seufert, Mobile Dev Memo on LinkedIn
  • “AI-driven traffic declines are reshaping the web: Resilient portfolios — particularly playable game publishers — are emerging as scaled, durable sources of web supply…” (January 11) – Jounce Media CEO Chris Kane on LinkedIn.
  • Snowflake Announces Intent to Acquire Observe to Deliver AI-Powered Observability at Enterprise Scale (January 8) – TechCrunch
  • “Inside the brand and agency scramble for first-party data in the AI era” (January 9) – Digiday (subscription)