1 Zoom is buying Common Room — buyer intelligence moves inside the meeting stack
Zoom signed a definitive agreement to acquire Common Room, the AI-native go-to-market intelligence platform used by GTM teams at Atlassian, Anthropic, Autodesk, Notion, Okta, and Snowflake. Common Room unifies first-party data across CRM, product, marketing, and engagement systems into a person-level view of every buyer, and its RoomieAI agents handle account research, message personalization, and prospecting. Terms weren't disclosed; the deal is expected to close within weeks.
The interesting part is where the intelligence lands: inside Zoom Revenue Accelerator, so reps know which accounts are in-market and why to reach out before the call ever starts. The handoff between "marketing knows something" and "sales does something" is exactly the seam this deal stitches shut. If your GTM stack still passes leads over a wall, you're now competing with teams whose wall no longer exists.
2 Profound launched Aim — AI search monitoring that turns into work, not dashboards
Profound announced Aim, an always-on background agent that monitors AI search signals — brand visibility, prompt trends, citations, sentiment, and competitive benchmarks across ChatGPT, Gemini, and Perplexity — and converts them into prioritized project briefs. Aim routes the work to specialized Profound Agents for research, content, and optimization, then measures whether completed work actually moved the targeted metric. Early users include Plaid's organic growth team.
This is the AEO category growing up: from "here's a dashboard showing you're invisible in ChatGPT" to "here's the brief, the agent, and the before/after measurement." Two months after raising at a $1B valuation, Profound is betting the moat isn't the data — it's closing the loop between insight and execution. Marketers should ask every analytics vendor they pay: what happens after the chart?
3 Runway shipped Agent 2.0 — feed it your Meta metrics, get the next batch of ads
Runway released Agent 2.0, an agent built for marketing teams that takes a struggling paid campaign or an upcoming launch, analyzes it, asks clarifying questions, and builds creative with you in one conversation. Connect your Meta or TikTok performance data and it studies what's winning, then generates the next round of ad variants to test.
The creative-performance loop — analyze results, brief the team, produce variants, relaunch — is the most labor-intensive cycle in performance marketing, and Runway just compressed it into a chat. For lean teams this is leverage; for creative agencies billing by the variant, it's a pricing problem arriving fast. Either way, "make more of what works" is becoming a button.
4 Google made images cost 3 cents and video 10 cents a second
Google launched Nano Banana 2 Lite, its fastest and cheapest image model — four seconds per image at roughly $0.03 — alongside Gemini Omni Flash, a video model priced around $0.10 per second of output. Chain them and a still product photo becomes a moving clip in a single pass.
Creative production cost is collapsing to rounding-error territory: a thousand campaign visuals now costs about $30. That doesn't make creative strategy cheap — it makes it the only expensive thing left. When every competitor can generate infinite variants, taste, brand codes, and knowing what to say are the bottleneck. Budget accordingly.
5 X opened its real-time firehose to any AI agent via MCP
X now hosts official MCP servers, letting AI assistants inside Cursor, Claude Desktop, and VS Code search its full archive, pull live trends, manage bookmarks, and draft posts using your own account — no custom API code required. It's built on Model Context Protocol, the open standard Anthropic released in 2024.
Real-time social data has been locked behind expensive API tiers since 2023; now it's a plug-in for whatever agent your team already runs. Social listening, trend-jacking, and community research just became things an agent does in the background, not a SaaS subscription line item. The platforms that feed agents the best data will quietly win the agentic workflow era.
6 Cleverbridge joined Visa's Agentic Ready program — the payment rails keep getting real
Cleverbridge, the e-commerce and subscription platform behind thousands of software merchants, joined issuer testing in Visa's Agentic Ready program and adopted the Trusted Agent Protocol to test AI-initiated transactions. The protocol ensures a purchase made by an agent on a consumer's behalf carries the same trust and authorization signals as a human-initiated one.
This is the unglamorous story that matters: agentic checkout fails without verified payment rails, and those rails are being laid merchant by merchant, right now. Subscription businesses should watch this closely — renewals, upgrades, and repurchases are precisely the transactions consumers will delegate to agents first.
7 Hostinger turned a product photo into a checkout link — no website required
Hostinger launched an e-commerce platform for small sellers and creators that converts a product photo into a functional checkout link in minutes, with no storefront needed. The context: AI-driven visits to retail sites grew 4,700% last year, and 76% of Gen Z already discover products on social media.
The storefront is dissolving into wherever the buyer already is — social feeds, chat threads, AI answers. Hostinger's bet is that small merchants need a commerce backend that follows the demand rather than a website that waits for it. If your conversion strategy still assumes a homepage visit, the data says your youngest customers never planned to make one.
8 ESW: consumers let AI compare prices, but won't let it pay
New ESW research found 47% of US consumers are comfortable using AI for price comparison, but only 24% are comfortable with AI-powered payments. AI is welcome in discovery and research; the wallet is still human territory.
Hold this stat against everything above — Visa's protocols, Hostinger's checkout links, agentic commerce infrastructure everywhere. The rails are being built roughly twice as fast as trust is forming. The practical play: deploy AI aggressively where consumers already accept it (comparison, discovery, recommendations) and make the payment step conspicuously human-controlled, transparent, and reversible until the trust curve catches up.
9 Edge226 acquired AnyClip — performance marketing keeps buying video intelligence
Edge226, an AI-powered cross-channel performance platform for mobile apps and games, acquired AnyClip, a video intelligence company built for media companies and publishers. The deal combines AnyClip's ability to analyze and activate video content at scale with Edge226's campaign execution infrastructure.
It's a small deal that fits a big pattern: performance platforms are buying specialized AI — especially video AI — rather than building it, because video is where ad spend is going and where analysis is hardest. Expect more of these tuck-ins; the mid-sized independents with real video IP are acquisition targets in a consolidating stack.
10 Audyence and Demandbase killed the B2B spreadsheet handoff — 43x faster to launch
Audyence, the programmatic marketplace for cost-per-lead inventory, launched the first native Demandbase integration: account segments sync directly into Audyence and activate CPL campaigns against those exact accounts in minutes — up to 43x faster than the traditional six-week purchasing motion. Audyence estimates up to 47% of what companies pay per lead is overhead — manual workflows, commissions, reseller markups — not the lead itself.
Every B2B demand team knows the ritual: export the segment CSV, clean it, email publishers, reconcile at month-end. That ritual just became a competitive disadvantage. When account intelligence flows straight into activation, the teams still doing spreadsheet gymnastics are paying nearly half their lead budget for friction.