Services Results About Writing MKT News Let's talk
Daily Briefings

MKT News.

This page collects daily marketing news briefings. Each day's briefing is added below, with the most recent at the top.

Published daily ~9:30 GMT · Last fetched 2026-04-12 21:24 UTC

April 10, 2026

Google March 2026 Core Update finishes rolling out today — start reading the shifts

Today is the expected completion date for Google's March 2026 Core Update (launched March 27), back-to-back with the Spam Update that finished March 24. That means the ranking volatility of the last two weeks should finally settle, and organic traffic data from this weekend onward is the first clean read you'll get. Don't make content or SEO changes yet — give it at least 3-5 days of stable data before concluding anything. If a client site dropped, check whether it was in a cluster affected by the Spam Update (spammy link profile) vs. the Core Update (content quality/E-E-A-T) — the fix is different. The December 2025 Core Update wiped ~15% of top-10 pages, so stakes remain elevated.

Source: Search Engine Land

Meta's Nielsen DMA → Comscore Markets transition has a hard June 22 deadline — audit your location targeting now

Meta is retiring Nielsen DMA targeting fields and moving to Comscore Markets for US geographic targeting. The hard deadline is June 22, 2026: campaigns still using the deprecated DMA fields after that date risk being paused automatically. This affects saved audiences, campaign templates, automated rules, and any third-party tool (Smartly, Madgicx, Revealbot) that writes campaigns via the Marketing API. Most agencies haven't touched this yet because the deadline feels far — but the audit itself takes a couple hours for a multi-client book. Do it this week: grep your saved audiences and automation configs for DMA references, map them to Comscore equivalents, and update templates before someone's Q2 budget stalls.

Source: Meta Ads Updates 2026

Canva acquires Simtheory and Ortto — marketing automation just consolidated one more layer

Canva announced a dual acquisition this week: Simtheory (AI agent collaboration platform) and Ortto (customer data + marketing automation). The strategic read: Canva is moving from "where you design the creative" to "where you design, schedule, automate, and measure the campaign." For agencies and consultancies, this matters because Canva is now a legitimate threat to standalone martech stacks at the SMB and mid-market end — expect more clients to ask whether they can collapse HubSpot + Canva + separate scheduling tools into one subscription. It also tightens the competitive pressure on Adobe Express and the creator-tool end of Meta's Advantage+ suite. Not an immediate action, but worth knowing when clients bring it up.

Source: TechCrunch

Google launches Advertising and Measurement Developers Hub — a single home for API, tag, and measurement docs

Google quietly launched a new Advertising and Measurement Developers Hub this week, consolidating Google Ads API, GTM, GA4, Measurement Protocol, Data Manager API, and server-side tagging documentation into a single destination. This is a housekeeping launch, not a feature launch — but if you've been losing time hunting across five different developer sites every time you integrate conversion uploads or server-side tagging, bookmark it now. For teams onboarding new engineers into conversion plumbing, it's the first place to point them. Small quality-of-life improvement, but these add up.

Source: Google Developers Blog

The bigger picture: Two of today's four stories are deadline/cleanup items (Core Update settling, Meta DMA migration) and two are structural (Canva's martech consolidation, Google centralizing developer docs). Taken together, they reinforce a pattern we've seen all week: the platforms keep shipping small enforcement dates and data-plumbing changes that reward agencies with tight operational hygiene and punish those still running on defaults. The Canva story is the one to keep an eye on over the next quarter — if they can actually stitch Ortto's automation into the Canva canvas, the mid-market SaaS tool that gets most disrupted isn't Adobe, it's HubSpot's Marketing Hub Starter tier.

April 9, 2026

Meta's engage-through attribution is rolling out now — update your reporting setup

Meta's attribution overhaul (announced in March) is now actively rolling out for campaigns optimizing toward website or in-store conversions. The key changes: click-through attribution now counts only genuine link clicks — not likes, shares, or saves. Everything else falls under a new "engage-through attribution" category (replacing the old engaged-view model) with a one-day conversion window. Video engaged-view thresholds also dropped from 10 seconds to 5 seconds, reflecting faster Reels-era conversion behavior. The practical impact: your click-through numbers in Ads Manager will drop (that's cleanup, not performance decline), while engage-through numbers will appear as a new line item. Update your client reporting templates now so the metric shift doesn't trigger false alarms. If you use third-party attribution tools, the good news is Meta's definitions now align more closely with how Google Analytics and others count clicks.

Source: Search Engine Land

Threads App Ads go global via Marketing API — app install campaigns now live

Meta's March 25 Marketing API update quietly enabled a significant inventory expansion: Threads now supports App Ads globally, including APP_INSTALLS and app event optimization flows. With 400 million monthly active users and carousel + video formats now supported, Threads is no longer just an organic brand presence play. The API update also added reply moderation tooling, so advertisers can view, hide, and respond to top-level replies on Threads ads — addressing the brand safety concern that held back early adopters. If you're running Meta app install campaigns, add Threads as a placement and let the algorithm test it. The CPMs are still relatively low compared to Feed and Reels, which means early movers get cheap inventory before the floor rises.

Source: Social Media Today

Google PMax raises video asset limits to 15 and adds portrait image support

Google increased the Performance Max video asset limit to 15 videos per asset group and now supports 9:16 portrait images natively. Combined with last week's Veo integration (3 images → 10-second video), this means you can fill PMax asset groups with significantly more video variations without hitting caps. The portrait image support matters for mobile-heavy campaigns — PMax can now render vertical creative across YouTube Shorts, Discover, and Gmail without awkward cropping. If you've been holding back on video diversity in PMax because of asset limits, that constraint just loosened.

Source: Google Ads API Changelog

Google March Core Update expected to complete tomorrow — hold your SEO changes

The March 2026 Core Update (launched March 27) is expected to wrap around April 10 — tomorrow. Combined with the Spam Update that completed in record time on March 24, any ranking shifts you've observed over the past two weeks have been impossible to attribute cleanly. The December 2025 Core Update wiped ~15% of top-10 pages, so stakes remain high. Final reminder: don't make reactive content or SEO changes until the rollout completes and you have at least a week of stable data to analyze. Check Search Console tomorrow afternoon for initial signals.

Source: Search Engine Land

The bigger picture: Meta is doing two things simultaneously that together reshape how advertisers measure and reach audiences: tightening attribution definitions (engage-through vs. click-through) while expanding inventory to new surfaces (Threads app ads, WhatsApp Status). The attribution cleanup is overdue — it brings Meta closer to how the rest of the industry counts conversions, which actually makes cross-platform reporting easier. The inventory expansion means more places for your budget to go, but also more complexity in placement-level analysis. Meanwhile, Google quietly expanding PMax's creative capacity continues the pattern of the last two weeks: more automation surface area, more inputs required from advertisers. The platforms are building bigger engines — they just need you to feed them more fuel.

April 8, 2026

Meta's AI ad overhaul demands 3x more creative — and most brands aren't ready

Marketing Brew's deep dive yesterday revealed the practical fallout of Meta's Andromeda ad retrieval system and Advantage+ automation push. The new system matches ads to broad audiences rather than narrow segments, which means advertisers need far more creative variations — one agency saw a client's 300-asset request balloon to 1,000 after accounting for personas and concepts. The problem: most big brands are refusing to let AI generate their creative due to legal and brand control concerns. As one SVP put it, they're constantly playing "Whac-A-Mole to figure out what's the new thing they didn't tell us about that they've turned on." Meta's goal of full automation by end-2026 looks unrealistic, but the creative volume requirement is real right now. If you're running Meta campaigns, start stress-testing your creative production pipeline — the volume demands are only going up.

Source: Marketing Brew

Google March 2026 Core Update still rolling — expect completion by April 10

The first Core Update of 2026 (launched March 27) is still in progress and expected to wrap around April 10. Combined with the Spam Update that completed in record time on March 24, any organic ranking shifts you're seeing this week are nearly impossible to attribute cleanly. The December 2025 Core Update wiped ~15% of top-10 pages entirely, so stakes remain high. Keep monitoring organic traffic on content-heavy client properties through end of week — don't make reactive changes until the rollout completes and the dust settles.

Source: ppc.land

Reminder: Google Lookalike user list uniqueness enforced April 30 — audit now

With three weeks until Google enforces its uniqueness check on Lookalike user lists in Demand Gen campaigns, this is your audit window. After April 30, duplicate lists created from the same seed (common when testing different expansion rates) will be rejected at the API level, breaking automated workflows. If you run Demand Gen through scripts or third-party tools, consolidate duplicate lookalike segments now rather than scrambling at the deadline.

Source: Google Ads API Changelog

The bigger picture: Meta's Andromeda story is the one to watch. The platform is systematically removing advertiser control levers while demanding more creative input — a combination that squeezes agencies from both sides. Brands that can scale creative production (whether through AI tools, creator partnerships, or just better workflows) will get rewarded by the algorithm. Brands that resist will see performance degrade as the system optimizes around them. This isn't a future prediction — it's happening in Ads Manager right now.

April 7, 2026

The agentic marketplace is here — AI agents are starting to buy media without the bid stream

Swivel and Olyzon partnered to build a fully agentic marketplace where AI agents discover, negotiate, and execute ad buys — including inventory that was never in the bid stream. They're communicating via AdCP, an open-source protocol built on top of MCP that lets agents across different companies talk to each other. Meanwhile, SSP Kargo launched Project Kera, a chat-based planning and buying interface for media buyers. This isn't theoretical anymore: Wpromote is actively testing agentic SSP tools in live campaigns. The implications are structural — if agents can bypass the traditional DSP-SSP pipeline, the fee structure of programmatic advertising gets renegotiated. Worth understanding now, even if your clients aren't touching it yet.

Source: AdExchanger

Google Ads API quadruples batch upload limits — automation workflows just got faster

Google increased the request size limit for BatchJobService AddBatchJobOperations from 10.48 MB to 41.94 MB (individual mutate operations still capped at 10.48 MB). If you manage large accounts with heavy image assets or bulk campaign changes through the API, this means fewer chunked requests and faster execution. Not glamorous, but for agencies running automated workflows at scale, this removes a real friction point. Check your scripts — you may be able to simplify upload logic that was working around the old limit.

Source: Google Ads Developer Blog

Two agency holdcos audit The Trade Desk — mid-market DSPs smell blood

Two major agency holding companies announced audits of The Trade Desk's fee structures, and mid-market DSPs are aggressively pitching to capture any ad dollars that come loose. Yahoo launched a "DSP Push" with new AI-powered tools specifically targeting the mid-market advertisers that TTD and DV360 have dominated. The competitive landscape in programmatic is shifting: if your clients use The Trade Desk through an agency holdco, ask whether their fees are being reviewed. If they're mid-market and feeling priced out of premium DSPs, the alternatives just got more credible.

Source: Digiday / AdExchanger

Google improves Offline Conversion Import — fewer expired event errors, cross-account flexibility

Google is rolling out infrastructure updates to Offline Conversion Import that improve attribution matching (reducing the "expired click" errors that plague long sales cycles) and add flexible import requests that make it easier to push conversions across multiple accounts. If you've been fighting OCI error rates on B2B or high-consideration accounts, this is directly relevant. Revisit your OCI setup — the error rates you accepted six months ago may no longer apply.

Source: Google Ads API Changelog

The bigger picture: The most important story today isn't a feature update — it's the agentic marketplace. For years, programmatic has operated through a fixed pipeline: advertiser → agency → DSP → exchange → SSP → publisher. AI agents communicating through open protocols threaten to collapse that chain. It's early, but the fact that real agencies (Wpromote) are testing this in production, and that an open standard (AdCP) already exists, means this isn't vaporware. Meanwhile, the holdco audits of TTD signal that even within the existing pipeline, the fee structures advertisers have accepted are under pressure. The programmatic supply chain is getting squeezed from both ends.

April 6, 2026

Google AI Mode ads hit 25% of results — and they cost 35% more per click

Google's shopping ads inside AI Mode have graduated from pilot to broad rollout. Sponsored product listings now appear in roughly 1 in 4 AI Mode conversations, labeled "Sponsored" below organic recommendations. Early data: 18% higher engagement than traditional search ads, but at a 35% CPC premium. Only Performance Max and AI Max for Search campaigns are eligible — standard Search campaigns don't qualify. If you haven't audited your Merchant Center feeds recently, do it now: AI Mode pulls heavily from structured product data, so feed quality directly determines whether you show up in this new surface.

Source: Search Engine Land

OpenAI's ChatGPT ads hit $100M ARR in six weeks — ads inside conversations are real now

OpenAI disclosed that its US ads pilot, launched February 9, crossed $100 million in annualized revenue by late March — just six weeks in. Over 600 advertisers are running. Ads appear at the bottom of ChatGPT responses for free and Go-tier users, clearly labeled and excluded from under-18 users and sensitive topics. Less than 20% of eligible users see ads daily, meaning there's massive room to scale inventory. Testing is expanding to Canada, Australia, and New Zealand. The measurement story is still weak (as flagged March 30), but the demand signal is unmistakable. Worth monitoring — not yet worth a serious media plan allocation until attribution improves.

Source: CNBC

Meta ships AI video generation from single images — beta testers report 10% CTR lift

Meta rolled out a suite of AI creative tools: generate video ads from a single product image, create UGC-style ads using AI avatars, auto-translate voiceovers across languages, and auto-convert product catalogs into video format. Beta testers are reporting ~10% higher click-through rates on AI-generated video vs. static creative. For e-commerce advertisers drowning in creative production costs, this is a direct workflow improvement. Test it on lower-stakes product lines first to calibrate quality expectations.

Source: Digital Applied

LinkedIn's algorithm now uses LLMs to understand post meaning — keyword stuffing is dead

LinkedIn updated its feed algorithm to use large language models that parse the actual topic and meaning of posts, not just keyword matching. This means the platform can now distinguish between a thoughtful post about "attribution" and one that just name-drops the word for reach. For organic content strategy and LinkedIn Ads targeting alike, this rewards genuine expertise over gaming. If you've been optimizing LinkedIn content around keyword density, shift to topic depth instead.

Source: SwipeInsight

The bigger picture: The ad industry's center of gravity is shifting from search bars to conversations. Google is monetizing AI Mode at scale, OpenAI proved there's $100M+ in demand for conversational ad inventory, and Meta is using AI to solve the creative bottleneck that's held back video adoption. The advertisers who'll benefit most are those with clean product data (for AI Mode), strong creative assets or willingness to test AI-generated ones (for Meta), and patience to wait for measurement to catch up before committing serious budget to ChatGPT. The platforms are building the pipes — the measurement infrastructure is lagging behind.

Older briefings

April 3, 2026

Meta goes big at NewFronts: Reels Trending Ads get category lineups, Creator Marketplace hits 1.5M

Meta used its Thursday NewFront to ship three updates that matter for paid social planners. Reels Trending Ads now include category-specific lineups tied to tentpole events (Fashion Week, NFL games), giving advertisers contextual placement options beyond broad algorithmic feed. The Partnership Ads Hub got a full redesign to streamline creator-brand collaboration workflows. And Instagram's Creator Marketplace expanded its matching filters with 1.5 million+ listed creators — making it meaningfully easier to find and activate creator partnerships at scale. If you run Meta campaigns with any creator or Reels component, explore the new lineup options before Q2 planning locks.

Source: Search Engine Journal

Google launches Commerce Media Suite — retailer first-party data meets YouTube targeting

Google's new Commerce Media Suite connects retailer purchase and loyalty data directly into ad targeting across YouTube and Display & Video 360. Kroger Precision Marketing is the launch partner, letting CPG brands target verified Kroger shoppers on YouTube using actual purchase history — not modeled audiences. This is Google's play to compete with Amazon and Walmart's retail media networks by offering retailer data + Google's reach. If you manage CPG or retail accounts, this is worth a test — deterministic purchase data targeting on YouTube is a step change from affinity audiences.

Source: SwipeInsight

Microsoft PMax gets customer acquisition goals and better measurement

Microsoft Advertising added customer acquisition goals to Performance Max, plus deeper visibility into where spend is going and improved conversion measurement. This closes a gap with Google's PMax — you can now explicitly optimize Microsoft PMax for new customers rather than letting the algorithm chase whoever converts cheapest. If you're running Microsoft PMax (or have been avoiding it because of limited controls), revisit it. The acquisition goal alone makes it viable for prospecting budgets.

Source: SwipeInsight / Microsoft Advertising

The bigger picture: The theme this week is platforms racing to prove they can deliver precision without sacrificing scale. Google is plugging retailer purchase data into YouTube. Meta is making creator-driven placements more structured and buyable. Microsoft is catching up on the controls that kept serious advertisers away from its PMax. The common thread: walled gardens are getting better at combining first-party data signals with automated delivery — which means the advertisers who invest in clean first-party data (customer lists, purchase feeds, creator relationships) will increasingly outperform those relying on platform defaults.

April 2, 2026

Google Ads experiments now auto-apply winning variants — check your settings today

As of April 1, Google flipped the default for Ads experiments: winning variants now auto-apply automatically instead of requiring manual approval. The upside is faster iteration. The risk: if your conversion tracking isn't airtight, Google might declare the wrong variant a winner and push it live without you noticing. Long sales cycles, brand campaigns, and any account where downstream conversions matter most are most exposed. Go to Experiments → review your auto-apply toggle now.

Source: Search Engine Land

Google Analytics Scenario Planner is live — model budget outcomes before you commit

Google launched Scenario Planner inside Google Analytics as part of a cross-channel budgeting suite, paired with a Projections tool that tracks live campaigns against revenue and conversion goals in real time. Scenario Planner uses at least 12 months of cross-channel historical data to build forward-looking models — genuinely useful for pitching budget increases or planning Q3 allocation. Requires beta access and 12+ months of multi-channel data. Check eligibility in your GA4 account under Advertising → Budget Planner.

Source: Search Engine Journal

Meta is feeding anonymized AI chat signals into ad targeting

Meta's April 2026 privacy update allows anonymized signals from AI product interactions to inform ad targeting — without exposing individual conversations. This expands the behavioral signal set Meta has access to beyond traditional app engagement. No immediate action required, but this matters for clients in privacy-sensitive sectors (healthcare, finance, legal) where the source of targeting signals is a compliance question. Worth flagging before clients ask.

Source: Digital Marketing Updates, April 2026

Google Shopping ads are now live inside AI Mode — feed quality just got more important

Google has graduated Shopping ads in AI Mode from experimental to live. Sponsored product listings now appear alongside AI-generated recommendations in AI Mode search results. This is a structural shift: your products can surface inside a conversational interface, not just a traditional SERP. AI Mode pulls heavily from Merchant Center structured data, so feed accuracy, product titles, and pricing freshness directly impact whether you show up. If your Merchant Center hasn't been audited recently, now's the time.

Source: Seafoam Media / AI Adtech News, April 2026

The bigger picture: Google is systematically transferring campaign decision-making from advertisers to its own systems — experiments auto-apply, voice-overs auto-apply, AI Mode surfaces products algorithmically. Each change in isolation looks like a convenience feature. Collectively, they represent a meaningful reduction in manual control. The advertisers who win in this environment have tight conversion tracking, clean data infrastructure, and strong exclusion controls. The ones who get hurt are running on defaults and hoping Google optimises toward their actual business goals.

April 1, 2026

Google launches non-skippable Video Reach Campaigns for Connected TV — globally

Google is now offering non-skippable Video Reach Campaigns for YouTube on Connected TV screens worldwide. The format uses automation to optimize ad delivery and format selection across CTV inventory. If you run video prospecting or brand awareness campaigns, this unlocks YouTube's fastest-growing screen with a forced-view format. Test it against your existing skippable CTV mix — the reach/frequency trade-off just shifted.

Source: Google Ads

TikTok expands Pulse suite with AI-curated contextual placements for US and Canada

TikTok detailed its full Pulse advertising suite — four contextual placement products — and revealed two new additions: Pulse Tastemakers (creator-driven placements) and Pulse Mentions (brand mention targeting), both rolling out to US and Canadian advertisers through Q2–Q3 2026. Custom Lineups use generative AI to curate trending content collections matched to your brief. For B2C advertisers already on TikTok, this is a meaningful step toward brand-safe, contextually relevant placements without relying purely on algorithmic feed positioning.

Source: ppc.land

Microsoft Ads adds self-service negative keyword lists — finally

Microsoft Advertising now lets you create and manage negative keyword lists directly in the platform. Previously, this required filing support tickets. It's a basic hygiene feature that should have existed years ago, but if you manage Microsoft Ads campaigns, go build your shared negative keyword lists now rather than maintaining them at the campaign level. Less friction = fewer wasted clicks.

Source: SwipeInsight

Google enforces Lookalike user list uniqueness in Demand Gen — audit by April 30

Starting April 30, Google Ads API will enforce a uniqueness check on Lookalike user lists in Demand Gen campaigns. Duplicate lists will be rejected. If you've been creating multiple lookalike segments from the same seed list (common when testing expansion rates), audit your account now and consolidate before the deadline hits. This is an API-level enforcement — it'll break automated workflows that generate duplicates.

Source: Google Ads API Changelog

The bigger picture: Today's enforcements are live — Customer Match API uploads via the old method stop working and Demand Gen's $5 daily minimum kicks in (both flagged yesterday, act now if you haven't). Beyond that, the pattern this week is platforms competing hard on video and CTV inventory: Google opens non-skippable CTV globally, TikTok counters with AI-curated contextual video placements. The battle for video ad dollars is shifting from "do you have video creative?" to "which screen and context do you want it on?" Meanwhile, Microsoft quietly ships a feature that should have existed at launch — a reminder that the unsexy platform improvements often save the most money.

March 31, 2026

Meta starts charging "Location Fees" in 6 countries starting April — your CPMs just went up

Meta will add Digital Service Tax surcharges to ad spend in Austria (5%), France (3%), Italy (3%), Spain (3%), Turkey (5%), and the UK (2%) starting in April 2026. These aren't absorbed by Meta — they're passed directly to advertisers as line-item fees on top of existing costs. If you run campaigns targeting any of these markets, update your budget forecasts and flag this to clients before they see invoices jump. Small percentages, but at scale they add up fast.

Source: Search Engine Land

Google PMax finally gets customer list exclusions and real reporting

Google shipped a batch of Performance Max upgrades: first-party customer list exclusions (you can now block existing customers from prospecting campaigns), budget pacing reports, demographic breakdowns in audience insights, and placement reporting by network. This is the transparency PMax has been missing since launch. The customer list exclusion alone changes how you structure prospecting vs. retention — go set it up.

Source: Search Engine Journal

Veo AI video generation is now built into Google Ads — 3 images → 10-second video

Google integrated its Veo generative video model directly into Asset Studio. Upload three static images and it generates a 10-second video ad, ready to deploy in Demand Gen and other formats. This dramatically lowers the creative production barrier for video testing. The quality won't replace proper production for hero assets, but for rapid variation testing and filling asset group gaps, it's a real workflow shift.

Source: Search Engine Journal

Google Ads Editor bug silently links structured snippets across accounts

A bug in Ads Editor is causing copied structured snippet extensions to remain linked between source and destination accounts. If you change the language in one account, it changes in the other. This is the kind of thing that goes unnoticed until a client's ads suddenly show snippets in the wrong language. If you've copied structured snippets between accounts recently, check them manually.

Source: Search Engine Land

The bigger picture: Google is finally addressing the two biggest PMax complaints — lack of exclusion controls and opaque reporting — while simultaneously making AI-generated creative trivially easy to produce. The combination means PMax is becoming genuinely usable for sophisticated advertisers, not just a black box you throw budget at. Meanwhile, Meta passing DST costs directly to advertisers is a reminder that your effective CPM is increasingly a function of geography and regulation, not just auction dynamics.

March 30, 2026

URGENT: Google Customer Match API stops working April 1 — migrate to Data Manager today

Starting tomorrow, Customer Match uploads via the Google Ads API will fail silently. Google is forcing a full migration to the Data Manager API. If any of your accounts or scripts push audience lists via the old API, those lists stop refreshing as of midnight tonight. Check every account that uses Customer Match and update your upload connectors or scripts before EOD. This one will bite you if you sleep on it.

Source: Google Ads Help / ALM Corp

Google auto-applied AI voice-over to PMax videos — opt-out window already closed

Google rolled out AI-generated voice-over to Performance Max video assets that lack an existing voice track, layering a synthesized voice automatically. The opt-out deadline was March 20 — if you didn't disable it under Video Enhancement Controls, your ads may already be running with AI-generated audio. Go check PMax asset groups now, especially on brand-sensitive or regulated accounts. To disable: Campaign settings → Video enhancements → uncheck AI voice-over.

Source: Google Ads / Profitable Digital

Meta quietly redefined what counts as a "link click" — expect metric drops

Meta updated its link click attribution to only count genuine link clicks, stripping out engagements that were previously lumped in and inflating numbers. Separately, the "engaged view" threshold for video ads dropped from 10 seconds to 5 seconds, which will increase video engagement counts. Net effect: link click metrics in Ads Manager will look lower (that's cleanup, not performance decline), while video engagement numbers will rise. Update your reporting baselines — and if you use link clicks as a KPI for client reporting, flag this before it causes confusion.

Source: Social Media Today

ChatGPT ads live via Criteo at $60 CPM — compelling channel, terrible measurement

Criteo became the first programmatic platform to integrate with OpenAI's ad inventory, giving ~17,000 advertisers access to ChatGPT ad placements through standard programmatic workflows. The rate is ~$60 CPM — roughly 3× Meta, comparable to Netflix's ad tier. Catch: advertisers report minimal performance data visibility, and OpenAI isn't yet providing standard attribution. Interesting to watch, but hard to justify meaningful spend until you can actually measure it. Put it on the radar, not the media plan.

Source: ALM Corp / Paid Search Association

LinkedIn beats Google and Meta on ROAS for B2B — Dreamdata benchmark report

Dreamdata's 2026 LinkedIn Ads Benchmarks Report found LinkedIn is the only major ad platform delivering positive ROAS for B2B: 121%, vs. Google Search at 67% and Meta at 51%. LinkedIn now captures 41% of B2B paid social budgets. If you've been fighting client skepticism about LinkedIn's premium CPCs, this is the data point to bookmark. It's the most rigorous B2B performance benchmark published this year.

Source: Dreamdata / PR Newswire

The bigger picture: Google and Meta are both defaulting advertisers into AI-driven automation — PMax voice-overs auto-applied, attribution definitions quietly shifted — with tight opt-out windows that reward the inattentive. Meanwhile, ChatGPT opens as a new ad channel but without measurement infrastructure, it's still faith-based marketing. The through-line: platforms are expanding control while reducing advertiser visibility. LinkedIn's ROAS data is the outlier here — a platform actually showing its work.