AI search visibility
Vanaxity insights tagged AI search visibility - notes on AI content, SEO, GEO and AEO from Van Data Team.
Muse Spark Open for Marketers: License, Cost and Uses
Meta released the weights for Muse Spark Open on October 15, 2026, the first open version of its Muse Spark line. It has 816 billion parameters, 42 billion active per token, a 1 million token context window and native support for text, images, video and audio. Meta calls it the most capable US open-weight model and says it can run on two Apple M5 Ultra Mac Studios. In Meta's own tests it leads on agent and tool-use benchmarks but trails Chinese open models on overall scores and coding. The license is free for most uses but requires a Built with Muse Spark Open label, and hosted access starts at $1.25 and $4.25 per million input and output tokens.
B2B Marketing Board Credibility: What Boards Want to See
At Marketing Week's Festival of Marketing on October 14, 2026, three B2B marketing leaders said marketers earn board credibility by understanding business economics, measuring what matters and showing how brand investment drives growth. The gap is real: McKinsey found 70% of CEOs judge marketing on revenue growth and margin, while only 35% of CMOs track that as a top metric. This guide turns the panel's advice into steps: learn the profit and loss, swap activity metrics for business ones, make the case for brand as future demand, apply deal rigour to spend, and add AI search visibility as an early signal of demand.
Gemini 4 Argon: What It Means for Marketers and AI Search
Google made Gemini 4 Argon, its new frontier model, broadly available on October 13, 2026, after two weeks of access limited to cybersecurity partners. It runs in the Gemini app and AI Mode in Search for Google AI Pro and Ultra subscribers and in the paid Gemini API at an introductory $2 and $10 per million input and output tokens, rising to $4 and $20 from December 1. Google says it leads 12 of 18 benchmarks it published. Artificial Analysis scores it 53, level with GPT-6 Astra, with a 15% hallucination rate against Astra's 51%, though it answers fewer questions correctly. For marketers, the bigger story is AI Mode, tasks mode and Gemini's 40-plus connected apps.
Small Agencies: What Wins When Independence Isn't Enough
Small agencies can no longer win on independence alone, Ad Age reported in July 2026, as holding companies consolidate and clients ask for proof, not positioning. Ad Age's first Small Agency Network benchmark survey found 71% of AI-mature agencies grew revenue last year against 33% of lighter AI users, a link rather than proof that AI caused the growth. Publicis is winning big accounts without pitches, more than half of independent owners say they'd consider selling, and business buyers now start much of their research in AI tools. The small agencies that win pick a clear focus, build AI into delivery, price on outcomes and make sure buyers can find them, including in AI answers.
Meta Ad Costs Are Up 12%: How Media Buyers Are Adapting
Meta's average price per ad rose 12% year over year in both the first and second quarters of 2026, according to its earnings reports, and media buyers told Marketing Brew that CPMs have climbed across ad formats. Meta says results have improved too, citing an internal study that found $4.13 in revenue per ad dollar, while Metricool's independent data shows Meta Sales campaign ROAS falling from 5.48x to 4.57x. Some advertisers are moving budget from the holidays to the cheaper 'Q5' window after Christmas, leaning on Advantage+ and testing more creative. Higher paid costs also make owned visibility, including how AI tools describe your brand, worth more.
Claude Sonnet 5.5 vs Opus 5.5: Which Fits Marketing Work?
Anthropic released Claude Opus 5.5 on September 22, 2026 and Claude Sonnet 5.5 on September 28. Sonnet 5.5 costs half as much per token and runs faster. On independent Artificial Analysis tests at top settings, the two are nearly tied on documents, spreadsheets and automation, but Opus 5.5 is far more reliable on facts and costs less per task. At matched lower settings, Opus 5.5 wins every test for a little more money, while Sonnet 5.5 starts answering much sooner. For marketing teams, the right choice depends on the task: accuracy-critical content favors Opus, and fast, high-volume drafting favors Sonnet.
Global Brands Going Desi: What Cultural Marketing Pays
Global luxury and sportswear brands are drawing on Indian cultural codes, from Louis Vuitton's auto-rickshaw bag to Prada's Kolhapuri-style sandals, Fendi's embroidered Baguette and an Adidas Samba mule inspired by the jutti. Experts quoted by exchange4media say the payoff is mostly conversation and differentiation, not direct sales. The cases also show the risk: Prada's uncredited design drew a backlash that ended in a partnership with Indian artisans, and Fendi faced criticism over credit. Because AI answers draw heavily on news and social coverage, how a brand credits the culture it borrows from now shapes its reputation in AI search too.
Experiential Marketing in the AI Era: Make Events Citable
Experiential marketing is growing again, with PQ Media putting global spend at $128.35 billion in 2024 and 84% of consumer event marketers in EventTrack 2026 planning to spend more. The case for it is getting stronger as AI floods feeds with similar content. But AI search engines mostly cite Reddit, YouTube, LinkedIn, review sites and editorial pages, while attendees mostly post to Facebook and Instagram or tell friends in private, so most event buzz never reaches the answer layer. This guide shows how to design experiences that leave a public, citable trace, where AI can help run them, how to measure them, and which popular statistics not to repeat.
Marketing Agency Acquisition: What Clients Should Protect
Marketing M&A is accelerating, with LUMA Partners counting 469 ad tech and marketing deals in 2025, up 13%, alongside giant moves like Omnicom's $13 billion purchase of Interpublic and Adobe's $1.9 billion deal for Semrush, and while a new owner can bring better tools and specialists, clients feel an acquisition through changed teams, reporting, and priorities, so the practical response is to protect account knowledge, keep your own copies of data and history, compare results before and after, and write change-of-control protections into agency contracts.
Technical SEO Debt Is Now an AI Search Visibility Problem
Technical SEO debt used to cost you crawl budget and rankings, but it now costs AI search visibility too, because testing by Vercel and MERJ found that none of the major AI crawlers execute JavaScript, so anything rendered in the browser is invisible to ChatGPT, Claude, and Perplexity, and since those crawlers already fetch at roughly a quarter of Googlebot's volume, the practical answer is to reserve a fixed slice of every sprint for debt paydown, prioritize by revenue impact rather than audit severity, and check what bots actually see rather than what your browser shows.
HubSpot AEO at Unbound: What CRM-Native AI Search Means
At its Unbound conference on September 16, 2026, HubSpot tied answer engine optimization, ChatGPT Ads, and AI agents into its CRM through a shared layer it calls growth context, building on the HubSpot AEO product it launched in April, and while the company reports large gains from early customers, an independent analyst calls the release an evolution in better packaging, so the practical lesson for marketers is that CRM-native AEO is most valuable for linking AI search visibility to pipeline, and least reliable when a single visibility score is treated as proof of results.