| |

This Week in AI Marketing (16th Mar 26)

ByAnn Stanley

AI Overviews Cut Clicks 42%, Google Opens Door to Gemini Ads, and Major Brands Back Agentic Commerce

This week brought hard data on AI’s impact on organic search, with AI Overviews slashing traditional clicks by 42%. Google hinted at ads coming to Gemini, while Meta overhauled its attribution model and hiked prices across six countries including the UK.

In ecommerce, L’Oreal, Unilever and Mars backed a new Agentic Merchant Protocol, and research suggests AI will drive over half of all ecommerce transactions by 2027. Meanwhile, LinkedIn emerged as the only major ad platform delivering positive ROAS, Anthropic shipped three major updates in a single week, and the CMA issued guidance on agentic AI with fines of up to 10% of global turnover for non-compliance.

Table of Contents

AI Search & SEO

Advertising Platforms

Ecommerce & Retail AI

Agentic AI & Automation

Social Media & Content

AI Tools & Models

Marketing Strategy

Key Takeaways

Frequently Asked Questions

Conclusion

AI Search & SEO

Google AI Overviews Cut Search Clicks 42%

Source: searchengineland.com | 12 March 2026

New data from Define Media Group confirms what many SEOs have feared: Google’s AI Overviews are reducing traditional organic search clicks by 42%. The losses are concentrated in informational and evergreen content — exactly the type of pages most brands have spent years building. However, the picture is more nuanced than the headline suggests. Breaking news traffic has surged 103% from November 2024 through early 2026, and Google Discover grew 30% across the publisher portfolio studied.

The most striking finding is that Discover and web search now drive roughly equal traffic for the first time. AI Overviews appear less often for news queries than for other topics, which explains why real-time content is actually gaining while evergreen pages decline. For publishers and brands, this represents a fundamental shift in content strategy: timely, original reporting is being rewarded, while generic informational content is losing ground to AI-generated summaries.

Why it matters
The 42% figure puts a hard number on the AI search disruption. If you’re still measuring SEO success purely by traditional organic clicks, you’re tracking a shrinking metric. Smart marketers should be monitoring AI Overview appearances, Discover traffic, and brand mentions across LLMs as new KPIs. The growth in breaking news traffic also suggests that brands publishing timely, original content — rather than rehashing existing information — will be the ones that maintain organic visibility.

Source: modernretail.co | 11 March 2026

Furniture.com, a brand literally built around SEO-first strategy, is now pivoting to optimise for AI search engines. The company is restructuring its content, product data and site architecture to ensure visibility not just in Google but across ChatGPT, Claude, Perplexity and other AI-powered search tools. It’s investing in structured data, conversational content formats and product information that AI agents can parse and recommend — a significant departure from the keyword-density approach that built the brand.

The pivot is particularly telling because Furniture.com’s entire business model was designed around capturing search traffic. When a company named after its target keyword concludes that traditional SEO alone isn’t enough, it signals a market-wide shift. The company’s approach includes creating rich product narratives that LLMs can reference, building FAQ-style content that matches conversational queries, and ensuring product specifications are structured for machine readability.

Why it matters
GEO (Generative Engine Optimisation) isn’t optional for ecommerce brands anymore — it’s becoming as fundamental as traditional SEO was a decade ago. If your product pages are optimised only for Google’s crawler, you’re missing the growing cohort of consumers who discover products through AI assistants.

Semrush Refreshes Its Identity for the AI Search Era

Source: adage.com | 12 March 2026

Semrush has repositioned its entire brand identity around AI search visibility, acknowledging that brand authority now extends well beyond owned domains. The platform is pivoting from its traditional SEO toolset to encompass visibility tracking across AI Overviews, LLM citations and generative search results. The rebrand reflects a broader recognition across the marketing tools industry that the metrics and methods of the past decade are no longer sufficient.

Why it matters
When the tools themselves rebrand around AI search, it signals a permanent market shift. Marketers should be evaluating whether their current toolstack can track AI visibility — not just traditional rankings. If your SEO platform can’t show you where your brand appears in AI-generated answers, it’s time to look at alternatives.

Source: raconteur.net | 11 March 2026

Raconteur explores how brands must orchestrate their digital presence across an increasingly fragmented AI search landscape, featuring insights from Google’s head of Search. With users now discovering products and services through multiple AI-powered channels — from Google AI Overviews to ChatGPT and Perplexity — the traditional funnel is breaking apart. The piece argues that richer intent signals and smarter orchestration across platforms are becoming essential for growth.

Why it matters
AI search isn’t a single channel — it’s a constellation of platforms. Brands that only optimise for Google risk becoming invisible to a growing segment of their audience who discover products through AI assistants, social search, and conversational interfaces.

Advertising Platforms

Google Is Not Ruling Out Ads in Gemini

Source: wired.com and searchengineland.com | 12 March 2026

Google’s senior vice president Nick Fox told WIRED that the company is “not ruling out” ads in its Gemini AI app — a notable shift from months of denials. Just weeks ago at Davos, DeepMind CEO Demis Hassabis told reporters Google had “no plans” for Gemini ads. Now Fox is saying learnings from ads in AI Mode will “likely carry over” to Gemini. His comment that “our research shows that users actually like ads within the context of Search” suggests Google sees AI-powered ad formats as a natural evolution rather than a compromise.

The commercial pressure is clear. Gemini now has more than 750 million monthly active users, up from 350 million a year ago. OpenAI began testing ads on ChatGPT’s free tier in January. Rather than rushing into Gemini directly, Google is using AI Mode — its Gemini-powered Search product — as a testing ground. The brands that figure out how to show up relevantly in conversational AI environments before the auction gets competitive will have a significant first-mover advantage.

Why it matters
When a senior executive goes on record saying “not ruling them out,” it’s corporate-speak for “we’re building it.” With 750 million users, Gemini represents a massive new advertising surface. Advertisers should start thinking about what AI-native ad formats look like — conversational product recommendations, contextual suggestions within AI responses, and sponsored answers will likely look very different from traditional search ads.

Source: searchengineland.com | 9 March 2026

Google is automatically enrolling advertisers in an AI voice-over programme that will narrate their Performance Max video ads — unless they opt out by 20 March. The feature uses AI-generated voices to add narration to video creative, with Google framing it as an “asset optimisation feature” that enhances viewer experience. Advertisers who don’t want AI narrating their brand message have a narrow window to disable it before it goes live.

This follows Google’s pattern of launching AI features as opt-out rather than opt-in, which has frustrated advertisers who prefer to maintain control over their creative output. Previous examples include automatically generated ad copy and AI-expanded audience targeting — both launched as defaults that required manual intervention to disable.

Why it matters
The opt-out deadline is 20 March — just days away. If you’re running PMax video campaigns and haven’t reviewed this, check your account now. More broadly, Google’s default-on approach to AI features means advertisers need to actively monitor new rollouts. Set a calendar reminder to review Google Ads announcements weekly — otherwise your brand’s voice might literally become an AI-generated one.

Meta Overhauls Attribution and Hikes Ad Prices Across Six Countries

Source: socialmediatoday.com | 3 and 10 March 2026

Meta made two significant moves this week. First, it’s introducing “location fees” of 2-5% on ads delivered in Austria, France, Italy, Spain, Turkey and the UK to pass on Digital Service Tax costs to advertisers. The fees are based on audience location, not advertiser location, and will appear as separate line items on invoices. Advertisers targeting UK audiences will see costs rise regardless of where their business is based.

Second, Meta is overhauling its attribution model in a move that should actually improve data quality. Click-through attribution will now only count link clicks — not likes, shares or saves — while other social interactions fall under a new “engage-through attribution” category. The video engaged-view window has also been shortened from 10 to 5 seconds, aligning more closely with how Google Analytics measures video engagement. These changes will likely cause an initial drop in reported conversions, but the resulting data will be more accurate and more comparable across platforms.

Why it matters
UK advertisers will see direct cost increases. But the attribution change is arguably more significant long-term — it aligns Meta’s measurement with Google Analytics, giving marketers cleaner cross-platform data on what’s actually driving conversions. Review your attribution windows, update your benchmarks before the numbers shift, and communicate the change to stakeholders so a temporary drop in reported conversions doesn’t trigger budget cuts.

LinkedIn Delivers 121% ROAS — Outperforms All Major Platforms

Source: prnewswire.com (Dreamdata) | 10 March 2026

Dreamdata’s annual LinkedIn Ads Benchmarks Report reveals that LinkedIn is the only major advertising platform delivering positive return on ad spend at 121%, compared to Google Search at 67% and Meta at just 51%. The report also finds that B2B buyer journeys now span 272 days with 88 touchpoints on average, and LinkedIn captures 41% of all B2B paid social budgets. These figures are based on multi-touch attribution modelling across Dreamdata’s client base, giving a more complete picture than last-click metrics typically provide.

The 272-day buyer journey figure is particularly significant because it explains why LinkedIn often looks expensive on short-term attribution windows. When the average B2B sale takes nine months and involves 88 interactions, measuring any single platform on a 30-day last-click basis will systematically undervalue upper-funnel channels like LinkedIn.

Why it matters
For B2B marketers, this data makes a compelling case for shifting budget towards LinkedIn. But it’s equally an argument for overhauling your attribution model. If you’re measuring LinkedIn on a 30-day window, you’re missing most of the picture — and likely underinvesting in the one platform that’s delivering positive returns.

Ecommerce & Retail AI

AI Set to Drive Over Half of Ecommerce Transactions by 2027

Source: channelx.world | March 2026

New research reported by ChannelX suggests that AI will drive over 50% of all ecommerce transactions by 2027. This encompasses AI-powered product discovery, personalised recommendations, automated purchasing through AI agents, and conversational commerce. The projection represents a fundamental shift from human-browsed to AI-mediated shopping experiences, where the “customer” interacting with your product page may increasingly be an AI agent rather than a person.

The timeline is aggressive but backed by current trajectory data. AI shopping assistants like Amazon’s Rufus, Google’s Shopping AI, and standalone agents from companies like Perplexity and OpenAI are already influencing purchase decisions. Combined with the Agentic Merchant Protocol story below and Morgan Stanley’s prediction that 10-20% of US commerce spend could be agent-mediated in the near term, the direction is clear even if the exact timing is debatable.

Why it matters
If even a fraction of this prediction holds, brands that haven’t optimised their product data for AI consumption are already behind. The infrastructure for AI-driven commerce is being built right now — and the winners will be the brands whose product information is structured, comprehensive and machine-readable.

How to Make Your E-Commerce Product Visible to AI Agents

Source: venturebeat.com | 10 March 2026

VentureBeat reports on a new system for making ecommerce products visible to AI shopping agents, backed by L’Oréal, Unilever, Mars and Beiersdorf. The system creates structured product data that AI agents can parse, compare and act on — effectively building the “API layer” between products and AI shoppers. Morgan Stanley research cited in the article suggests 10-20% of entire US commerce spend could be agent-mediated in the near term.

The approach goes beyond traditional product feeds. Where Google Shopping feeds provide structured data for search results, this system is designed for AI agents that need to understand product benefits, compare alternatives, assess compatibility, and make purchase recommendations autonomously. It’s the difference between listing specifications and enabling a conversation about why one product is better than another for a specific use case.

Why it matters
Your product feed is no longer just for Google Shopping. AI agents need rich, structured, machine-readable product data to recommend and purchase on behalf of consumers. Brands that make their catalogues AI-agent-friendly — with detailed product narratives, structured specifications, and clear comparison data — will have a significant first-mover advantage as agent-mediated commerce scales.

Amazon Expands Multi-Retailer Shopping and Brings Audiences to Netflix

Source: techcrunch.com and adexchanger.com | 11 March and 4 March 2026

Amazon made two significant moves this week. First, it’s expanding Shop Direct, a programme that lets customers discover and buy products from other retailers’ websites through Amazon’s ecosystem. The expansion now supports third-party product feeds from Feedonomics, Salsify, and CedCommerce, giving Amazon real-time access to merchants’ inventory and product information. Crucially, Amazon’s AI shopping assistant Rufus can now recommend and facilitate purchases from these third-party merchants.

Second, Amazon Ads announced that US media buyers will soon be able to use Amazon Audiences for targeting Netflix campaigns through Amazon DSP. Prime Video now reaches roughly 130 million people on its ad-supported plan since defaulting subscribers in 2024. By tying commerce and shopping data to streaming impressions on Netflix, Amazon is giving buyers what they want most: performance data linking ad exposure to actual purchases. Amazon’s director of agency partnerships Sarah Iooss told AdExchanger this year’s pitch to advertisers focuses on “improvements to its ad tech stack” beyond just content.

Why it matters
Amazon is building a commerce and advertising layer that extends far beyond amazon.co.uk. The Netflix targeting integration means Amazon’s first-party shopping data — arguably the richest commerce dataset in the world — is now available across premium streaming inventory. For ecommerce brands, this opens entirely new audience targeting possibilities where you can reach people based on what they actually buy, not just what they browse.

The AI Attribution Blind Spot

Source: practicalecommerce.com | 8 March 2026

Practical Ecommerce explores a growing challenge: AI-assisted purchases are creating attribution blind spots that existing analytics tools can’t track. When a customer asks ChatGPT or Claude to research a product, compares options in Perplexity, then visits your site directly, the entire AI-mediated research phase is invisible to your attribution model. The article argues that this untracked AI influence is already distorting marketing budget decisions across the ecommerce industry.

The problem is compounding as AI usage grows. If 10% of your customers are using AI to research purchases before arriving at your site, your attribution data is systematically undervaluing the channels and content that AI assistants are referencing. This creates a vicious cycle: channels that AI amplifies appear less effective in analytics, leading to budget cuts that reduce the content AI assistants can reference.

Why it matters
Traditional attribution is becoming increasingly unreliable for understanding the full customer journey. Consider running brand lift studies, post-purchase surveys asking how customers discovered your products, and monitoring your brand’s visibility in AI search tools to capture this hidden influence layer.

Pinterest Launches Shoppable TV Series on Roku

Source: pymnts.com | March 2026

Pinterest is launching “Bring My Pinterest to Life,” a six-episode shoppable TV series on Roku. Hosted by creators including Drew Michael Scott and Caroline Vazzana, the show lets viewers shop products directly as they appear on screen, with brand partners including Wayfair, eos and Michaels. Instead of traditional ad breaks, products are woven into the content with shoppable purchase options built into the viewing experience.

The series is part of a broader CTV push by Pinterest. The company recently acquired tvScientific, a CTV performance advertising platform that allows advertisers to run their own connected TV campaigns, pay by outcome, and measure the actual impact of TV advertising. Together, the shoppable series and the tvScientific acquisition position Pinterest as a bridge between social discovery and performance-driven television commerce.

Why it matters
This blurs the line between content and commerce more than any previous format. If shoppable CTV works for Pinterest, expect every social platform to follow. For brands, this means creative that works as entertainment — not just as ads — becomes the differentiator. The tvScientific acquisition also signals that Pinterest is serious about proving CTV ROI, which could make it a compelling alternative to YouTube and Meta for video-first advertisers.

Agentic AI & Automation

L’Oréal, Unilever & Mars Back New Agentic Merchant Protocol

Source: channelx.world | March 2026

Three of the world’s largest consumer goods companies — L’Oréal, Unilever and Mars — are backing a new Agentic Merchant Protocol designed to standardise how AI shopping agents interact with ecommerce platforms. The protocol creates a common framework for AI agents to discover, compare and purchase products across different retailers, effectively building the infrastructure layer for autonomous AI-driven commerce. Beiersdorf is also among the early backers.

The protocol addresses a fundamental problem: AI shopping agents currently have no standard way to interact with product catalogues across different platforms. Each retailer’s data is structured differently, priced differently, and presented differently. The Agentic Merchant Protocol creates a shared language for AI agents to navigate this fragmented landscape — similar to how schemas.org standardised structured data for search engines a decade ago.

Why it matters
When L’Oréal, Unilever and Mars collectively back a standard, it becomes the standard. This protocol will define how AI agents shop, and brands that adopt it early will be the ones AI agents can find and recommend. Think of it as structured data markup for the agentic commerce era — ignore it now and you risk becoming invisible to AI shoppers within the next 18 months.

CMA Issues Guidance on Agentic AI with Fines of Up to 10% of Global Turnover

Source: lewissilkin.com | 13 March 2026

The Competition and Markets Authority (CMA) has published detailed guidance on how businesses can use agentic AI whilst complying with consumer protection law. The guidance draws a clear distinction between traditional AI tools that assist human decisions and AI agents that sense their environment, plan workflows, retrieve real-time data, execute actions autonomously (including making payments), and store memory of past interactions. A business is responsible for what an AI agent does in exactly the same way as it would be for what an employee does.

The CMA sets out four practical steps: tell customers if they’re interacting with an AI agent; don’t allow AI agents to mislead consumers; ensure easy access to a human when needed; and take full responsibility for the agent’s actions, even when a third party designed or provides it. Non-compliance carries fines of up to 10% of annual global turnover under the CMA’s new enforcement powers.

Why it matters
This is the first major UK regulatory framework specifically targeting agentic AI in commerce. If your business deploys AI agents that interact with customers — chatbots that make recommendations, automated pricing tools, AI-powered customer service, or autonomous purchasing agents — you need to review this guidance immediately. The 10% turnover fine is not a slap on the wrist, and the CMA has explicitly stated it intends to use its enforcement powers.

UiPath Secures Landmark AIUC-1 Certification for AI Agents

Source: itbrief.co.uk | 12 March 2026

UiPath has become the first enterprise automation platform to receive AIUC-1 certification, a security and reliability standard for AI agents used in production environments. The certification, run by the Artificial Intelligence Underwriting Company and audited by Schellman, evaluates AI agents against more than 2,000 enterprise risk scenarios including data protection, operational boundary controls, resistance to adversarial attacks, and error prevention. UiPath’s Intelligent Extraction Processing, Agents, and Autopilot products all passed the assessment.

Importantly, AIUC-1 requires quarterly re-evaluations, designed to track safeguards over time as AI systems evolve and threat patterns shift. This continuous assessment model sets it apart from one-off compliance certifications and provides ongoing assurance that AI agents remain secure as they’re updated.

Why it matters
As AI agents move from experiments to production, certification standards will become table stakes for enterprise adoption. If you’re evaluating AI agent platforms for marketing automation, customer service, or operational workflows, AIUC-1 compliance gives you a benchmark for comparing security and reliability across vendors. Expect procurement teams to start requiring this type of certification.

AI Agents Take Aim at the Local Marketing Agency

Source: forbes.com | Charlie Fink | 9 March 2026

Forbes reports on Mega, an AI startup that raised $11.5 million to build AI agents that directly compete with local marketing agencies. The platform handles tasks typically outsourced to small agencies — ad creation, campaign management, social media posting, and performance reporting — at a fraction of the cost. It’s specifically targeting the local business market where budgets are too small for traditional agency retainers but large enough to pay for AI-powered tools.

Mega isn’t the only player in this space, but the $11.5 million raise signals investor confidence that AI can genuinely replace the services provided by small, generalist agencies. The platform’s pitch is simple: why pay £2,000 a month for a junior account manager when an AI agent can handle the same tasks for a fraction of the cost, 24 hours a day, with consistent execution?

Why it matters
This is the agency disruption story playing out in real time. Local agencies handling basic campaign management are most at risk. The survival strategy is clear: move up the value chain into strategic consulting, complex multi-channel campaigns, and the kind of nuanced client relationships that AI agents can’t replicate. Agencies that are still selling execution hours rather than strategic thinking should see this as an urgent wake-up call.

Social Media & Content

TikTok Launches Radio and Podcast Network with iHeartMedia

Source: variety.com | 13 March 2026

TikTok Radio has relaunched on iHeartMedia’s app and 28 US broadcast radio stations across major markets including New York, Los Angeles, Atlanta and Chicago, ending a four-year SiriusXM partnership. The channel blends creator content with music, described as “guiding listeners through what’s trending before it trends.” Alongside the radio deal, a new TikTok Podcast Network is launching with creator-led shows distributed by iHeart, including “Suite 305 with Lele Pons,” “Sports Slice” with Tim Martin, and “Caroline’s Closet” with Caroline Vazzana.

The move represents TikTok’s first serious expansion beyond short-form video into audio content. While the platform has experimented with longer videos and live streaming, a radio and podcast network creates entirely new advertising inventory and creator monetisation opportunities in a format that reaches audiences during commutes, workouts and other contexts where video isn’t practical.

Why it matters
For brands already investing in TikTok, the podcast network opens longer-form storytelling formats — and potentially more engaged audiences than 60-second clips. The 28-station radio presence also creates a bridge between social media audiences and traditional broadcast reach, which could be particularly valuable for brands targeting demographics that straddle both channels.

Webflow Buys AI Content-Generation Platform Vidoso

Source: techcrunch.com | 12 March 2026

Webflow has acquired Vidoso, an AI content-generation platform, to bolster its marketing capabilities. The acquisition integrates AI-powered content creation directly into Webflow’s website building platform, enabling marketers to generate, test and publish content without leaving the CMS. It’s part of a broader trend of website platforms absorbing AI content tools rather than relying on third-party integrations.

Why it matters
The CMS-plus-AI-content trend is accelerating. Expect WordPress, Shopify and other platforms to follow with similar acquisitions or native AI content features. For marketers, the days of switching between multiple tools for content creation and publishing are numbered — and the platforms that integrate AI content generation most seamlessly will win market share.

Source: sproutsocial.com | Sinéad McNamee | 10 March 2026

Sprout Social’s comprehensive UK trends report argues that social media has become the UK’s “primary answer engine” — the place audiences go to diagnose problems, validate purchasing decisions and hold businesses accountable. The report’s most provocative claim is that “the exhausting scramble for a fleeting viral moment is officially over.” Instead, today’s market leaders are trading one-off viral hits for consistent, community-driven content that builds trust over time.

The report also explores the “human-AI editor” era, where social teams must find the right balance between AI-generated efficiency and human authenticity. The guidance is practical and UK-specific, covering platform trends, content format shifts and the growing role of AI in social media management, with input from Kikora Mason, VP of Social Media at Chase, who emphasises that “brand-consumer communication these days is more instant, transparent and community-driven than ever before.”

Why it matters
The “social as answer engine” framing aligns with the broader AI search disruption theme: consumers aren’t just searching Google anymore, they’re asking TikTok, Instagram and YouTube. If your social content strategy is still focused on broadcasting brand messages rather than answering questions and building community, you’re optimising for a model that’s already fading.

AI Tools & Models

Anthropic Ships Three Major Updates: Code Review, Visual Responses, and 1M Context

Source: techcrunch.com and theverge.com | 9-13 March 2026

Anthropic had a busy week with three significant releases. First, it launched a dedicated code review tool designed to check the flood of AI-generated code now entering production systems — essentially building quality control for its own output, acknowledging that AI-generated code needs a different type of review than human-written code. Second, Claude can now respond with charts, diagrams and other visual content, moving beyond text-only responses into data visualisation and technical illustration.

Third, and perhaps most significantly for enterprise users, the 1 million token context window is now generally available for both Opus 4.6 and Sonnet 4.6 at standard pricing. A million tokens translates to roughly 750,000 words — enough to analyse an entire codebase, process a complete marketing report library, or review lengthy legal documents in a single conversation. This removes the “document too long” limitation that has prevented many enterprise use cases from being practical.

Why it matters
Anthropic is building the full stack: code generation, code quality checking, visual output and massive context windows. The 1M context window is a game-changer for complex business tasks — analysing entire campaign performance datasets, processing full audit reports, or reviewing lengthy proposals in one pass. If you’re evaluating AI tools for your team, these updates make Claude significantly more capable for the kind of multi-document, data-heavy work that marketing and agency teams do daily.

The Need for “Grounded AI” — Not Hype

Source: channelx.world | March 2026

ChannelX makes the case for “grounded AI” — practical, measurable AI implementations that solve real business problems rather than chasing hype cycles. The piece argues that too many businesses are investing in AI for the sake of appearing innovative, without clear ROI frameworks or realistic expectations of what current AI can deliver. It advocates for starting with specific, bounded use cases that generate measurable returns before scaling to more ambitious applications.

Why it matters
After months of breathless AI announcements, this reality check is welcome. The most successful AI adopters aren’t the ones deploying the most advanced models — they’re the ones that identified a specific business problem and applied AI to solve it with clear metrics. Start with the problem, not the technology.

AI Layoffs Spread Across Tech as Martech Evolves

Source: martech.org | Constantine von Hoffman | 12 March 2026

MarTech’s weekly roundup reports that AI-driven layoffs are spreading across the technology sector, with companies increasingly replacing roles with AI automation. The article provocatively suggests that if companies truly believe AI can replace workers, they should start with the C-suite. Alongside the layoff trend, the roundup covers new martech releases and AI-powered marketing tools entering the market at an accelerating pace.

Why it matters
The AI layoff trend is accelerating from prediction to reality. Marketing teams should be proactively upskilling — learning to use AI tools effectively — rather than waiting to see if their roles are affected. The agencies and teams that learn to work with AI tools will be the ones that survive the shake-out; those that resist adoption risk being replaced by smaller teams who embrace them.

Marketing Strategy

The Marketing Data Most Companies Still Fail to Measure

Source: martech.org | 10 March 2026

As privacy changes continue to weaken digital attribution models, MarTech argues that phone conversations have become one of the most valuable — and most overlooked — first-party data signals in marketing. With cookie deprecation and iOS privacy changes eroding online tracking, the humble phone call contains rich intent data that isn’t subject to the same tracking restrictions. Companies that capture and analyse call data are building a significant competitive advantage in attribution.

The article points out that while marketers have invested heavily in digital analytics, many have neglected the offline conversion signals that privacy changes have made more valuable, not less. Call tracking and conversation intelligence tools can fill the attribution gaps that digital tracking restrictions have created — and they provide first-party data that’s increasingly hard to get from digital channels alone.

Why it matters
In the rush to adopt AI and digital-first measurement, many marketers have overlooked the most direct signal of purchase intent: someone picking up the phone. If you’re not tracking calls as a conversion signal, you may be missing a significant portion of your marketing ROI — and making budget decisions based on incomplete data.

ROI Pressure Is Pushing CMOs’ AI Strategy in Five Directions

Source: thedrum.com | Graham Wilkinson, Acxiom | 11 March 2026

With boards demanding tangible returns on AI investments, CMOs are being pushed towards five strategic priorities: hyper-personalisation at scale, AI-powered customer journey orchestration, predictive analytics for budget allocation, automated content production, and real-time campaign optimisation. Graham Wilkinson, Acxiom’s global head of AI, argues that the days of AI experimentation without accountability are over — marketing leaders now need to demonstrate clear ROI from every AI initiative or risk having budgets cut.

The article highlights a growing tension in marketing leadership: the pressure to adopt AI quickly (driven by competitive anxiety) versus the need to prove measurable returns (driven by board-level scrutiny). The CMOs navigating this successfully are those who frame AI as a performance multiplier for existing strategies rather than a standalone innovation project.

Why it matters
The shift from “we’re experimenting with AI” to “show me the ROI” is the most important trend in enterprise AI adoption right now. CMOs who can connect AI initiatives to revenue outcomes will keep their budgets. Those who can’t will see AI projects defunded, regardless of how innovative they are. Frame every AI investment as a business case, not a technology project.

Key Takeaways

  • AI Overviews cut organic clicks 42% — but breaking news (+103%) and Discover (+30%) are growing. Track AI visibility alongside traditional SEO metrics and prioritise timely, original content.
  • Google ads in Gemini are coming. With 750M monthly users, Gemini is a massive new ad surface. Nick Fox’s “not ruling them out” is the clearest signal yet. Start planning for conversational ad formats.
  • Meta’s UK ad costs are rising 2-5% via Digital Service Tax location fees, and attribution is changing to count only link clicks. Review your Meta budgets and update your benchmarks before reported conversions shift.
  • LinkedIn delivers 121% ROAS versus Google (67%) and Meta (51%) for B2B. If you’re underweighting LinkedIn and measuring on a 30-day window, you’re missing a 272-day buyer journey.
  • The Agentic Merchant Protocol backed by L’Oréal, Unilever and Mars is building the infrastructure for AI-driven commerce. Get your product data AI-agent-ready now.
  • CMA agentic AI guidance comes with fines up to 10% of global turnover and four clear compliance steps. If you deploy customer-facing AI agents, review compliance immediately.
  • PMax AI voice-overs go live by default — opt-out deadline is 20 March. Check your Google Ads account before it’s too late.

Frequently Asked Questions

How should I prepare for Google ads appearing in Gemini?

Start by ensuring your Google Ads account is optimised for AI Mode in Search, which Google has confirmed will inform Gemini ad formats. Focus on conversational query targeting, strong brand signals, and creative that works in AI-generated contexts rather than traditional search results pages. Gemini has 750 million monthly users — this will be a significant advertising channel when it launches.

What is the Agentic Merchant Protocol and do I need to adopt it?

The Agentic Merchant Protocol is a new standard for how AI shopping agents interact with ecommerce platforms, backed by L’Oréal, Unilever and Mars. If you sell products online, you should monitor its development closely. Making your product data structured and machine-readable is the first step, regardless of the specific protocol — think of it as structured data markup for the AI shopping era.

How do Meta’s attribution changes affect my reporting?

Meta now only counts link clicks (not likes, shares or saves) for click-through attribution, and has shortened video engaged-view windows from 10 to 5 seconds. You’ll likely see reported conversions drop in the short term, but the data will be more accurate and more comparable with Google Analytics. Update your benchmarks and communicate the change to stakeholders before the numbers shift.

Should I shift SEO budget to AI search optimisation?

Don’t abandon traditional SEO, but start allocating budget for GEO (Generative Engine Optimisation). Monitor your brand’s visibility in AI Overviews, ChatGPT and Perplexity. The 42% click reduction from AI Overviews shows the direction of travel, but organic search still drives significant traffic for most businesses. The key is to add AI visibility tracking alongside your existing SEO metrics, not replace one with the other.

Conclusion

This week’s stories converge on a single theme: the infrastructure for AI-powered marketing and commerce is being built right now, and the window to prepare is closing fast. From Google’s inevitable Gemini ads to the Agentic Merchant Protocol creating standards for AI shopping, the market is moving from experimentation to implementation. The CMA’s regulatory guidance with 10% turnover fines adds urgency — businesses deploying AI agents need to do so responsibly and compliantly.

Three actions to take this week: first, check your Google Ads account for the PMax voice-over opt-out before 20 March. Second, review your product data readiness for AI agents — structured, comprehensive, machine-readable product information is no longer optional. Third, update your measurement framework to account for Meta’s attribution changes and the growing AI attribution blind spot that’s distorting budget decisions across the industry.

Need help adapting your AI marketing strategy? Contact the Anicca team for expert guidance.

This roundup is compiled from publicly available sources using AI-assisted research. While we review every article for accuracy, our analysis reflects our interpretation of the original reporting. We strongly encourage readers to click through to the original sources linked throughout this post for full context and detail. If you spot anything that needs correcting, please let us know.

FREE GUIDE

Free download

The Complete Claude Code Guide

67 pages for marketers, managers and business leaders. No coding required.

Ann Stanley
Ann StanleyFounder & CTO, Anicca Digital
Why it matters Understanding Claude Code Getting set up Working with Claude Skills Business use cases Rollout & governance
Download → anicca.co.uk/claude-code-guide
The Complete Claude Code Implementation Guide ebook cover
The Thursday AI Club logo

Thursday AI Club

A hands-on AI club for marketers and managers. Fortnightly 3-hour sessions: Hour 1 is an open Q&A on any AI question; Hours 2-3 split into a workshop track (for newbies) and an advanced track (live demos and deeper questions).

Led by Ann Stanley, James Allen.

Workshop recordings library
Skills and prompts to copy
Private community channel
Secret Agents community access
Full-day hackathon every quarter
Cancel any time

Membership: £40/month or £400/year

Join the Thursday AI Club →

Similar Posts