1 Zeta Global wired its data cloud into Palantir
Zeta Global announced a strategic partnership with Palantir on June 23 to build an enterprise AI infrastructure layer that connects operational and customer intelligence with marketing execution. The deep technical integration brings Palantir Foundry’s ontology, governance, and operational infrastructure into Zeta’s Data Cloud — which is being rearchitected on Foundry — with Athena, Zeta’s AI, sitting at the center.
When a marketing cloud rebuilds its foundation on a defense-grade data platform, it’s telling you where the value actually sits: the ontology, the map of how every piece of your data relates, is now the product — not the campaign tool bolted on top. The campaign UI is interchangeable; the governed, connected data layer underneath it is not. The practical question for any enterprise: is your customer data clean and connected enough that an agent could act on it safely, or would it just automate your worst assumptions faster?
2 MoEngage bought Aampe to put a reinforcement-learning brain inside the platform
MoEngage acquired Aampe, an AI infrastructure company that provisions a dedicated, autonomous AI agent for each individual customer, to bring 1:1 agentic decisioning to B2C marketing teams. The deal pulls Aampe’s reinforcement-learning engine natively into MoEngage so that workflow agents and decisioning agents run from a single unified system, and Aampe’s founding team — Paul Meinshausen, Schaun Wheeler, and Sami Abboud — joins to lead the agentic decisioning effort.
“Reinforcement learning” is the tell. Instead of you writing the rules — this segment gets this message at this hour — the system learns, per person, what actually works and adjusts itself. M&A in this category has stopped being about adding channels and started being about buying the decisioning brain. For marketers, “1:1 personalization at scale” stops being a slide promise and becomes an algorithm that runs the segmentation you used to build by hand.
3 Hightouch shipped Lifecycle Studio — agents that run the email-and-push grind
Hightouch launched Lifecycle Studio, the second studio in its Agentic Marketing Platform, giving marketers agents grounded in the full context of their brand, customers, and campaign history to build production-ready, cross-channel campaigns. The workspace also optimizes existing campaigns and turns market trends into new ones — and it lands on the back of the $2.75B valuation Hightouch reached on its $150M Series D in April.
Lifecycle marketing — the high-volume, rules-heavy work of email, push, and SMS — is exactly the kind of labor agents eat first, because it’s too constant and repetitive for a human to babysit. The phrase Hightouch is leaning on is “grounded in your context,” and that’s the whole game: a generic agent writes generic campaigns. Your owned data and campaign history are what make the agent’s output sound like you instead of like everyone else’s agent.
4 Contentful made “what AI says about you” a product you can buy
Contentful moved Palmata into general availability on June 23, a tool to help organizations understand, measure, and improve how their company is represented in AI answer engines. Powered by its proprietary Sounder Discovery Agent, Palmata continuously observes, learns, and synthesizes the signals that shape AI-generated answers by running ongoing analysis of public data. CEO Karthik Rau framed the stakes plainly: AI systems increasingly decide how businesses are discovered and evaluated before a customer ever visits a website.
Answer Engine Optimization went from buzzword to staffed function in under a year, and now your CMS vendor sells it as a monitored product. The shift is fundamental: you used to optimize to rank on a results page; now you optimize to be the sentence the model says. Start treating “what does ChatGPT say about us?” as a brand-tracking metric you check on a schedule — because right now most brands have no idea what answer is being given in the room they’re not in.
5 6sense closed the gap between ad spend and pipeline with AI-Recommended Leads
6sense introduced AI-Recommended Leads, a capability that surfaces high-quality, CRM-ready contacts for engaged accounts across 6sense advertising campaigns. It draws on the platform’s people intelligence, persona-targeting logic, and campaign engagement signals to identify high-confidence contacts, then hands them off as a downloadable list that routes straight to AI Email, a BDR team, or a sales rep — with measurable ROI as the headline claim.
B2B’s oldest leak sits between the ad you ran and the pipeline you hoped it built: you spent the money, then guessed who to call. Closing that loop with AI-picked contacts is what finally makes ad budget defensible in pipeline terms instead of impressions. The catch is the same as always — the leads are only as good as the intent signal underneath them, so a confident model on top of thin data just routes bad calls faster.
6 Sprout Social added native Snapchat publishing
Sprout Social launched a Snapchat publishing integration, now available to all customers, letting brands plan, schedule, and automatically publish Stories and Spotlights to Snapchat directly from Sprout. CMO Scott Morris pitched it as removing workflow complexity and helping brands keep a consistent presence across platforms, turning sustained connection and expanded reach into business impact.
Snapchat keeps getting left out of the “big platform” social toolset, so native publishing support inside a major suite is a real signal that brands want its young, highly engaged audience back in the rotation. For social teams the immediate win is mundane but real: one less platform you manage by hand. The thing to watch is whether the engaged-audience story actually converts into reach for brands, or whether Snapchat stays a place teens talk to friends and tune out ads.
7 Bloomreach turned the storefront into a conversation with Loomi
Bloomreach enhanced Loomi, its conversational AI shopping assistant that turns complex searches into guided shopping conversations. The upgraded agent remembers shopper preferences, reasons through complicated requests, fetches real customer reviews, and walks a buyer from first search straight to checkout — while feeding a self-serve analytics dashboard that surfaces trending themes, catalog gaps, and performance for the brand.
This is the product grid dissolving into a chat. When a shopper talks to an agent instead of scanning a page, your merchandising logic, reviews, and catalog data become the raw material the agent reasons over — which makes product-data quality the new conversion lever. But the analytics layer is the sleeper: every conversation is a stated-intent dataset you never had from a search box, telling you in plain language what shoppers want and where your catalog falls short.
8 Markup AI shipped Content Guardian Agents to QA the AI’s output
Markup AI launched Content Guardian Agents, a suite of AI quality agents built for marketing teams publishing at scale. The agents evaluate content for brand voice, audience fit, accuracy, human voice, and “AI citability” — whether a model is likely to quote it — with a Google Docs browser extension available today plus native API and MCP server integrations to embed the checks inside existing LLM and CMS workflows.
Once AI writes the first draft, the scarce job flips to checking it, and Markup AI is betting that QA gets automated too. Two of those scored metrics are quietly revealing: “AI citability” means you now write for the model as a reader, and “human voice” as a measurable score is the industry admitting that AI-written content reads like AI. The healthy pattern is agents draft, agents QA — but a human still owns the voice, because that’s the one thing the scoring can flag but can’t supply.
9 Dunnhumby pooled the retail-media walled gardens into one buying route
Dunnhumby launched a network alliance that connects brands with audiences across multiple retailers through a single route, with Tesco, B&Q, John Lewis, and Waitrose taking part in pilots. The promise is that brands and agencies can run campaigns across several retailers without rebuilding plans and processes for each one — a meaningful unlock in a channel that’s already the fastest-growing in advertising at roughly $63 billion globally.
Retail media’s biggest friction has always been fragmentation: every retailer is its own walled garden, with its own login, its own data, and its own specs. Pooling them into one buying route is the move that finally makes the channel manageable for a brand that isn’t the size of Amazon. For D2C brands now steering 15-25% of digital budget into retail media — up from 5-10% a few years ago — fewer walls means more of that spend actually reaches shoppers instead of being eaten by setup.
10 Aurasell shipped an Agent Builder — because the week’s real product is the builder, not the agent
Aurasell, an AI-native go-to-market platform, released Agent Builder, letting GTM teams build, orchestrate, and operationalize agentic workflows across their stack without writing code. It runs on three pillars: Aura Context, a graph connecting conversations, channels, and signals into a real-time view of the customer; Aura Trust, enterprise security built on RBAC and policy controls; and Aura IQ, GTM intelligence built by operators.
The pattern of the week wasn’t “an agent” — it was the agent builder. Vendors have figured out that one prebuilt agent never fits a real workflow, so they’re shipping the no-code tool to build your own instead. That moves the bottleneck from “can we get an agent” to “do we understand our own process well enough to describe it.” And note which pillar enterprises will actually scrutinize: not the context graph or the intelligence, but Trust — the governance and access controls that decide whether you can let an autonomous agent loose on real customers.