1 Amazon will now build your AI shopping assistant — in as little as 60 days
On May 27, AWS launched the Agentic Shopping Assistant, a service that lets retailers outside Amazon build their own AI shopping experiences using the same technology behind Alexa for Shopping (Amazon’s newly merged Rufus + Alexa+ commerce agent). Amazon is shipping it as a package of “architecture, starter code and learnings” tied to a retailer’s own storefront, catalog, customer data, and branding — with stand-up time “as little as 60 days.” The first public customer is Kate Spade (Tapestry), which built a gifting assistant; more than 300 million shoppers used Rufus in 2025, and Accenture estimates over $3 trillion in commerce could flow through AI agents by 2030.
The pitch is control: Amazon is telling retailers to own the agent rather than hand the customer relationship to a general-purpose AI — and it’s routing the service through AWS, not the retail arm, to look like infrastructure instead of a middleman. For brands, this is the real question of the year: is your storefront a destination, or a data feed into someone else’s agent? Building your own keeps the relationship; declining means competing for placement inside ChatGPT, Gemini, and Perplexity on their terms.
2 Meta launched paid “Plus” subscriptions — and teased an ad-free tier
Also on May 27, Meta rolled out consumer subscriptions globally: Facebook Plus and Instagram Plus at $3.99/month and WhatsApp Plus at $2.99/month, with perks like deeper analytics, story-rewatch stats, profile customization, premium stickers, and — notably — “wider audience reach.” Product chief Naomi Gleit said Meta intends to fold its offerings under a single brand, “Meta One,” and confirmed the company is working on “No Ads” plans for Threads and WhatsApp.
Two things matter for marketers. First, “wider reach” as a paid perk further blurs the line between organic and paid — the feed is now pay-to-play at the user level, not just the advertiser level. Second, an ad-free tier quietly removes paying, higher-intent users from your addressable audience; if your most engaged buyers opt out of ads, your Meta reach skews toward everyone else. Watch the take rate on “No Ads” the way you’d watch a targeting change.
3 Google killed the AI-Overview click signal — and your attribution just went dark
In early May, Google removed the method that let sites identify clicks coming from AI Overview results, with no replacement announced. The timing compounds a measurement problem that’s already structural: research across 50 B2B SaaS keywords in Q1 2026 found pages holding top-three rankings saw click-through rates fall 18–34% once AI answers appeared above the fold — even while rankings and impressions held steady. Buyers increasingly get their answer from an AI summary, then convert days later through branded search or a direct visit.
That influence never generates an attributable session, so last-click models simply miss it — and the gap widens every quarter. The fix isn’t a better analytics config; it’s a different measurement posture: add self-reported attribution fields (“how did you hear about us?”), build AI-referral segments in your CRM, and start treating share-of-voice inside AI answers as a real performance metric. If you only count clicks, you’re now blind to a growing slice of the funnel.
4 Rankings stopped predicting AI citations — the two finally decoupled
Google’s May 2026 Core Update landed alongside I/O, where the company said AI Mode crossed 1 billion monthly users and AI Overviews hit 2.5 billion, with Gemini 3.5 Flash as the default AI Mode model. The distribution story underneath: per Discovered Labs, top-10 Google rankers accounted for 76% of AI Overview citations in mid-2025 — but by early 2026 that share had fallen to roughly 38%. Strong rankings and AI-citation visibility have officially decoupled.
What drives a citation now is passage structure and information consistency across independent sources — not backlink count. Pages that answer a question cleanly, in self-contained chunks, with facts that match what other credible sources say, get pulled into the answer; the page that merely ranks #1 increasingly doesn’t. Practically, “SEO” is now three jobs — rank, get cited, and be trusted as an agent’s source — and they no longer move together. Audit your AI citation share separately from your rankings, because they’re telling you different stories.
5 Analysts say the ad market is splitting in two — and priced it accordingly
In late May, Redburn framed the moment bluntly: “AI is driving a bifurcation of the online advertising market,” with Meta pulling further ahead and smaller platforms finding catch-up “increasingly fruitless.” The price targets followed the thesis. Redburn lifted Snap to $10 (from $5) — crediting a subscription business already running above a $1 billion annualized rate and a core ad business it expects to swing from a ~$553M operating loss in FY25 to a ~$29M profit by FY27. It trimmed Reddit to $110 (from $125) on ad-revenue deceleration and data-licensing renewal risk, and cut Pinterest.
Translation for budgets: the consolidation toward Meta, Google, and Amazon isn’t a vibe — it’s being underwritten by the market. For challenger platforms, survival now runs through diversification (subscriptions, data licensing) rather than out-targeting Meta. For advertisers, the long tail of platforms is getting riskier to over-index on — test them as upside, not as the core of the plan, and make each one earn its line item every quarter.
6 Pipedrive bolted project management and AI briefs onto the CRM
In the week of May 29, Pipedrive launched project-management and communication capabilities so sales teams can run customer relationships and post-sale execution from one platform. The additions include workflow health monitoring, a Trust & Visibility dashboard, AI-generated project briefs, interactive Gantt timelines, Project Insights, and mobile project tracking — squarely aimed at SMBs juggling a half-dozen tools.
The interesting unit here is the AI-generated brief: the CRM is no longer just recording what happened, it’s drafting what should happen next. The pitch is consolidation — fewer tools, with AI handling the connective tissue between sale and delivery. Keep an eye on the “Trust & Visibility” framing, too: as AI takes over more of the busywork, buyers increasingly want a dashboard that shows what the agent actually did. Visibility into the automation is becoming a feature you can sell.
7 Target built its own creator ladder — gamified at the bottom, commissioned at the top
In early May, Target stood up two owned creator programs. Club Target (May 1) is a gamified, tier-based nano-creator program that rewards participation in weekly TikTok and Instagram challenges. Target Ambassadors (May 6) is invite-only for more established creators, powered by LTK, with higher commissions and deeper brand access. Together they form a ladder: enter at the gamified bottom, climb toward paid, commissioned partnership.
This is the structural shift in creator marketing made concrete — retailers building owned creator infrastructure instead of renting reach through agencies for one-off campaigns. The creator program is becoming a loyalty and retention system, with its own tiers, rewards, and progression, deduplicated across paid, affiliate, and organic. If you still treat creators as a campaign line item, the brands building rosters and ladders are quietly compounding relationships you’ll have to rent at a premium later.
8 Independent agencies keep buying creative shops to sell integrated judgment
In mid-May, Alloy — a marketing and technology partner — acquired Look Listen, an Emmy-nominated creative and digital agency, to strengthen integrated capabilities across strategy, PR, creative, media, and digital experience, adding storytelling, in-house production, and a dedicated studio. It’s a small deal, but it rhymes with the year’s loudest theme: consolidation toward owning the full stack.
As campaign execution commoditizes, the durable value an agency sells is integrated judgment — the ability to connect strategy to creative to production to measurement without seams. Independents are assembling that end-to-end capability to compete with the holding companies on capability while underselling them on bureaucracy. For brands choosing partners, the question shifts from “can you run the campaign?” (everyone can now) to “can you decide what the campaign should be, and prove it worked?”
9 Cannes is three weeks out — and the rules now force AI work to declare itself
With the festival running June 22–26, Cannes Lions 2026 has codified two governance moves worth your attention. It’s debuting the Creative Brand Lion, which — in CEO Simon Cook’s framing — shifts the focus from finished creative outputs to the internal systems, cultures, and capabilities that make great work repeatable. And it’s adding AI Craft subcategories across the craft-led Lions, judging human-plus-AI work on craft, intent, and genuine human contribution rather than letting it compete unmarked against fully human-made work.
This is the most consequential awards-governance call since the Effectiveness Lions, and the downstream implication is practical: provenance documentation — which models touched which assets, in what order, with what human review — is now an eligibility concern, not just a compliance one. The Creative Brand Lion sends the deeper signal: in a world where anyone can generate a thousand assets, the defensible advantage is the system that reliably produces good ones. Judge the engine, not the output.
10 The Trade Desk’s agents now buy the media — with Stagwell as first partner
On April 21, Stagwell became the first global marketing network to adopt The Trade Desk’s Koa Agents, alpha agentic-AI capabilities for media planning, buying, optimization, and measurement across the open internet. Instead of configuring campaigns step by step, marketers describe their goals and the agents execute, optimize, and adapt in real time. The first phase automates audience planning and activation inside Stagwell’s workflows, with closed beta for select clients later this summer.
This is agentic media buying moving from demo to deployment inside a major network — tasks that once took days of setup compressing into a prompt. Which is exactly why the trader’s job is changing, not disappearing: when the agent handles execution, value shifts to the quality of the brief, the richness of your first-party data, and the judgment to evaluate what the agent optimized toward. Across this whole list, that’s the pattern — the execution got automated, so the brief became the edge.