August 2025

🚀 AI Evolution Accelerates

Marketing is transforming faster than ever. And the smartest leaders are already ahead of the curve.

Google now shows AI summaries to two billion users, cutting publisher traffic by twenty-five percent. Traditional SEO is evolving. More and more buyers skip your website entirely. They ask AI tools for recommendations and often make decisions before you even know they exist.

But here's what smart marketers are doing differently. They're turning disruption into opportunity. They're making this shift their competitive advantage.

Companies like n8n just hit two billion dollar valuations by automating workflows. Why? Because AI without proper process management is just expensive chaos. Fifty percent of executives think AI speeds up marketing, but only twenty-eight percent of marketers agree. The gap? Wrong focus. Winners fix operational bottlenecks first, then add AI features.

The new playbook is simple. Create content AI wants to cite. Build trust through transparency, not tracking. Let robots handle routine tasks while humans focus on strategy and relationships.

The companies winning today aren't just adding more AI tools. They're redesigning their entire business for an AI-first world.

Marketing is evolving, becoming more powerful and precise. The future belongs to those who embrace this transformation now.

🎥 Watch the video summary of this newsletter

Show All AI Agent Marketing Leader Advertising Video Search Organisation Tools

• Google has introduced Gemini 2.5 Flash Image (aka "Nano-Banana"), a next-gen model focused on character consistency and precise natural language edits. The tool helps mix several images together and gives creators more control, while making it easy to use at a bigger scale. My take: It's worth trying — an intuitive image editor that finally lets us experiment with photos without needing Photoshop. With automation, we're moving closer to hyper-personalization, tailoring each photo to a customer's individual preferences.
  | #AI #Marketing #Video | Google | 2025/08/26

  View on TikTok

• Just Eat is testing a new robot delivery service in Oerlikon, near Zurich. For the first month, a safety expert will accompany the robot; afterwards, it will be controlled remotely. The robot can carry up to 40 liters, making it suitable for larger orders. Just Eat plans to expand this service to other European cities by the end of the year. My take: last-mile delivery is growing fast, and as more of these robots appear, people will gradually get used to them. But what about robots in shops and boutiques—could this trend help popularize robotic assistance in physical retail spaces?
  | #AI #Marketing #Supply | 20minutes | 2025/08/22

  View on TikTok

• HubSpot and SurveyMonkey found most customers still prefer human service over AI, with 82% favoring personal support and 86% prioritizing empathy over speed. While younger consumers are more open to AI, many distrust its use. Klarna reverted to hiring staff after chatbot backlash. Experts suggest AI handle routine tasks, leaving complex, sensitive issues to humans to preserve trust. My take: AI should be integrated step by step, with cascading routing between human and AI based on performance. This requires a scaffolded monitoring system to ensure trust, adaptability, and continuous improvement.
  | #AI #ContactCenter #Marketing | CX DIVE | 2025/08/15
• Google's AI Overviews cut publisher traffic by up to 25%, sparking concern among brands. While Google denies major impact, trends suggest declines. My take: To adapt, we need to change our strategies as AI changes how people discover things.
  | #AI #Marketing #Search | emarketer | 2025/08/15
• MarTech interesting article about how AI is reshaping work and business. AI blurs role boundaries, letting generalists act like specialists and vice versa. Scaling AI needs curiosity, openness, and rethinking workflows. Mapping workflows together avoids duplication and reveals new value. Success starts small: focus on one process, set clear KPIs, and celebrate quick wins. Most successful deployments begin with pilot initiatives that prove value before scaling.
  | #AI #Organisation | MarTech | 2025/08/14
• Interesting insight from The Economist: AI is spreading slowly, much like electricity did long ago. At first, factories only swapped steam engines for electric motors, and little changed. Real progress came later, when factories were redesigned for electricity. Today, companies bolt AI onto old models. True gains will come only with full AI-ready business designs.
  | #AI #Organisation | The Economist | 2025/08/14
• Google's main profits come from ads on shopping-related searches, rather than purely informational queries. Losing low-value queries to AI doesn't hurt much, but commerce is at risk. Purchases differ: impulse buys still need ads, while AI can help with routines, lifestyle, or complex choices. To truly win, AI agents need better data (trustworthy reviews, real tests), unified APIs (buy anywhere, smoothly), identity and memory (know evolving preferences), and embedded capture (learn from user actions). Without these, AI stays weak in commerce.
  | #AI #Agent #Marketing #Search | ANDREESSEN HOROWITZ | 2025/08/12
  
• Sam Altman is launching Merge Labs to rival Elon Musk's Neuralink, valued at $850M and seeking $250M funding. Backed by OpenAI Ventures, it develops brain-computer interfaces. Altman co-founds with Alex Blania, aiming to leverage AI advances against Neuralink's $9B lead. My take: today ChatGPT-like tools are limited by typing and reading speed—imagine instant interaction via Neuralink or OpenAI. Definitely worth watching.
  | #AI #Interface | Financial Times | 2025/08/12
• According to the VP of Product at Shopify, engineers often question the need for PMs, dismissing them as "just writing docs." They sometimes try running product themselves, only to find it overwhelming—handling meeting notes, tickets, roadmaps, go-to-market and testing plans. Product managers bring order, strategy, and customer focus. With the right education and AI integration, paired with strong communication and collaboration, they remain essential to product success. My take: with AI tools like ChatGPT making documentation easier, more engineers challenge the value of PMs and other business-oriented roles. But while writing docs is simpler now, the real business side goes far beyond documentation—it's about ensuring the product stays valuable for customers, not just technically flawless, which is where engineers already excel. This highlights the critical importance of collaboration between business and tech.
  | #Organisation | Shopify | 2025/08/08

  View on TikTok

• Interesting article showing why Marketing and IT must team up in today's digital world. The way value is created and delivered is evolving quickly, and neither function can succeed alone. CMOs and CIOs now share responsibility for growth and customer experience. They integrate skills, co-manage data and platform ownership, and ensure a unified view across business and technology. With customer journeys spanning many channels, this alignment drives clearer vision, smarter decisions, and stronger outcomes. My take: Marketing and IT initiatives are essential in large organizations with millions of customers, complex marketing functions, and heavy legacy IT systems. In these conditions, and with new AI tools, full alignment to avoid roadmaps that Marketing doesn't need or struggling with unrealistic IT deployments. What helps is an umbrella approach, like Customer Value Management, that provides shared focus end-to-end.
  | #Marketing | MarTech.org | 2025/08/07
• German workflow automation startup, n8n, is raising hundreds of millions in a funding round led by Accel, boosting its valuation to $2.3 billion from $350 million just four months ago—a jump on par with hyperscaling peers like Mistral AI. The company, serving clients like Vodafone and Delivery Hero, topped $40 million in recurring revenue amid Europe's booming AI investment wave. My take: in the world of AI, workflow automation is becoming even more important to orchestrate AI and build tailored workflows for each organization.
  | #Leader | Ainvest | 2025/08/07
• Elon Musk said Grok's AI responses will soon include ads, a move to increase revenue after former CEO Linda Yaccarino's departure. Musk claimed the effort to make Grok the "smartest AI" is complete. The strategy will show ads alongside user queries, with xAI's technology enhancing targeting on X, marking a shift in the company's ad approach. My take: Google and xAI (Grok) are pushing toward ad-based models, while OpenAI stays subscription or usage-based. Interesting that Musk calls Grok the smartest yet still needs ads to monetize it.
  | #AI #Advertising #Leader | TechCrunch | 2025/08/07
• OpenAI has launched GPT-5, describing it as a major step forward for ChatGPT with stronger reasoning, reliability, and advanced support for coding and healthcare. For the first time, the system is open to free users, while paid subscribers receive extended features and model options. The rollout followed backlash over the quiet removal of GPT-4o, which drew viral criticism on Reddit and X. Users accused OpenAI of a bait-and-switch, with some praising 4o's empathy and others debating whether it was overrated or close to AGI. CEO Sam Altman admitted underestimating 4o's appeal and promised GPT-5 would adopt a warmer personality with more customization. Another source of confusion came from GPT-5's initial lack of mode controls. At launch, users could not manually switch between "Auto," "Fast," and "Thinking," and it was unclear which mode was active by default. Altman later clarified the settings and added manual selection, giving users more visibility and flexibility in choosing how the model responds. Reports also noted that GPT-5 introduces a "Universal Verifier," a new technology designed to reduce hallucinations and factual errors. By using an internal adversarial verification process, it aims to refine output quality and strengthen market competitiveness. My take: In coding and complex reasoning, GPT-5 is noticeably stronger than 4o, delivering real improvements in workflows. Yet in Auto mode it often requires tailored prompts, underscoring the need for better prompt-engineering methodology. The debate over warmth versus performance reveals a deeper philosophical fork: should AI evolve toward relational intelligence—valued for empathy and companionship—or remain a sterile, performance-driven tool? Finally, assessing any model requires testing it against repetitive workflows to measure time and quality. That is why, in certain coding scenarios, I still favor Claude Opus 4.1.
  | #AI #Leader | YouTube | 2025/08/07
• Salesforce and ServiceNow invest $1.5B in Genesys to boost AI-driven customer experience. The partnership enhances Genesys CX Cloud, sustainability, and growth, fostering seamless data flow, agentic AI, and collaboration. My take: interesting to see Salesforce (CRM giant leader), ServiceNow (workflow automation powerhouse), investing in Genesys (customer experience innovator). Is it another move to truly integrate sales and marketing workflows?
  | #CRM #ContactCenter | Technology Magazine | 2025/08/06
• Interesting article on how AI is rapidly reshaping SEO. As David Bell, SEO consultant and co-founder of Previsible, points out, unlike Google's traditional indexing cycles, large language models (LLMs) now prioritize content usefulness and direct answers. This shift creates faster visibility and higher engagement, rewarding content that immediately solves user needs. Traditional SEO strategies, focused mainly on rankings and tags, might lose relevance. The key takeaway: early adopters of AI-driven search dynamics will gain a strong competitive edge as discoverability shifts.
  | #Search | Search Engine Land | 2025/08/05
  
• Google DeepMind has announced Genie 3, a new AI that can create interactive worlds in real time. The model keeps these worlds consistent for minutes and runs at 24 frames per second. DeepMind says it improves on Genie 1 and 2, giving AI endless training spaces, but it's still a research claim, not publicly available. Games are at the center, acting as safe ways to test and grow AI skills. DeepMind is also working with game makers to build new tools, following earlier successes like AlphaGo. Short note: For now, you can't actually "play" Genie 3. DeepMind announced the model and showed demos, but it is only available in a limited research preview for selected researchers and creators—no public access or test link.
  | #Leader | Google | 2025/08/05
• OpenAI released two open-source models, OSS-20b and OSS-120b. They can run on personal computers, with OSS-20b working on just 16GB of RAM. Their strength is comparable to O3/O4 mini. My take: This is an important move by OpenAI and something many companies were waiting for. OpenAI models are highly desirable, but big companies often turned to LLaMA or other open-source options to build their own stacks without sending data to OpenAI's servers. Now, they can use OpenAI models directly in the same way.
  | #Leader | OpenAI | 2025/08/05

  View on TikTok

• Perplexity launched its AI browser Comet for US Pro subscribers, with global rollout coming soon. The tool reads pages, fills forms, and makes purchases, excelling at repetitive tasks. Early reviews are mixed on speed but highlight efficiency. With OpenAI pursuing its own browser, AI-driven browsing could reshape web interactions. My take: Definitely something to keep an eye on—browsers may become the next big enabler for AI adoption. One current limit of AI agents is their inability to act directly on the websites you use. With AI browsers, you'll be able to ask the agent to act as you browse, not just in isolation.
  | #AI #Agent #Leader | Perplexity | 2025/08/01
• A recent survey reveals a misalignment between executive optimism and marketing team realities in AI adoption. While 51% of executives believe AI speeds up marketing cycles, only 28% of marketers agree. This gap stems from information asymmetry, causing misplaced AI investments focused on customer-facing personalization rather than solving operational bottlenecks. Companies with smoother operations achieve stronger AI ROI, linking efficiency to success. My take: I've seen this especially in large organizations. In one multi-country tech company, one market used AI to support operations, while another added personalization features without changing legacy workflows. The result was striking—the first improved efficiency, while the second worsened performance and frustration. Addressing bottlenecks and aligning executive expectations with operational realities can unlock AI's true potential, despite survey limitations on maturity and sector differences.
  | #AI #Organisation | MarTech.org | 2025/08/01
• OpenAI now generates $12–13 billion in annual revenue, holds a $300 billion valuation, and surpassed 700 million weekly ChatGPT users—an unprecedented milestone reached at record speed in internet history. In its latest round, the company secured $8.3 billion and aims for $40 billion total funding by year-end. Revenue is projected to hit $20 billion, with strategies shaped by Trump's AI Action Plan and ongoing Microsoft negotiations, underscoring both rapid growth and geopolitical influence.
  | #Leader | TechCrunch | 2025/08/01
• Two-thirds of retailers now offer loyalty programs, but more than one-third of enrolled shoppers never use them — mainly due to siloed data and generic offers. The most successful brands win by focusing on five key elements. Simplicity makes joining and rewards easy (Lowe's phone-number sign-up; Wayfair's flat $29/year model). Personalization goes beyond discounts with AI-powered recommendations (Amazon, Etsy). Integration connects every touchpoint through CDPs and CEPs for real-time, consistent offers. Trust builds with transparent data use, instant benefits, and clear value. Finally, Engagement features like Foot Locker's drops and gamified incentives keep programs exciting.
  | #Marketing | Retail Dive | 2025/08/01
• Google is adding ads to AI Mode and AI Overviews on desktop and mobile. This helps pay for its big AI investments. Right now, ads come from normal search campaigns, not special AI targeting. Longer AI questions give brands new chances, but top search results get fewer clicks. Google is also using AI tools like Veo and Imagen to make ads automatically. This shift could reduce the role of ad agencies. Marketers are unsure—automation saves time but may reduce control of budgets. Google's big test: earn money from AI while keeping search easy to use.
  | #AI #Advertising #Leader #Marketing | Business Insider | 2025/08/01
  
• Google added AI Overviews to search results in May 2023. A Pew Research study shows they cut website clicks—dropping from 15% to just 8%. Many users stop browsing after reading summaries, with only 1% clicking sources. Wikipedia, YouTube, and Reddit make up 15% of cited references. AI Overviews now appear in about 20% of searches, mostly question-based. While publishers fear traffic loss, Google reports record profits and insists AI features improve engagement. The company also disputes Pew's methods, calling the study unrepresentative. The debate continues: are AI Overviews helping users—or hurting the open web?
  | #AI #Marketing #Search | Pew Research Center | 2025/08/01
  
• Amazon reported strong ad growth on Prime Video, as U.S. ad-supported viewership rose to 300 million monthly, up from 275 million. Amazon unveiled new ad formats powered by generative AI, including scene-aware placements, interactive campaigns, and real-time shopping ads that let viewers buy directly with their remotes. Building on last year's 30% lift in brand awareness, Amazon Ads says these formats will drive measurable results while blending seamlessly into entertainment.
  | #AI #Advertising #Marketing | Marketing Dive | 2025/08/01
• According to Robby Stein, VP of Product for Google Search, the platform is entering a new AI era. Google Search is transforming into a full AI product, aiming to deliver smarter, more natural user experiences. Each month, around 1.5 billion people use AI-powered search features worldwide. New tools let users ask longer, complex questions, hold natural conversations, and get deep results. Visual search is booming—Google Lens usage is up 70%, especially among youth. Future updates will bring multimodal features (voice, image, video), stronger personalization, and agentic functions that perform tasks for users. My take: it will be interesting to see how Google manages to grow revenue and profit during this transition. Today, Google's ad empire relies on keyword search. Putting ads directly into AI responses is still untested. History shows how risky this can be—think of Nokia's fall during the smartphone shift when the business model failed to adapt.
  | #AI #Leader #Marketing | Google | 2025/08/01
• According to MarTech.org, go-to-market teams are overloaded with AI tools but lack real strategy. Many build "Frankenstein stacks" that look powerful but break easily. The real problem is weak signals—teams don't know how to spot true buyer intent. Signals need smart thinking and custom rules, not just generic dashboards. As MarTech experts say, marketing leaders must take back control of GTM design. The winners will be those who align sales, marketing, and customer success around signal-first systems that build trust and drive growth. My take: I've seen this with many clients. The most effective approach was strong cooperation between Marketing and Tech teams under umbrella programs such as Customer Value Management, supported by a solid make-or-buy analysis: make for core activities (differentiation points) and buy for system parts that are not part of the company's unique value. For example, in loyalty programs, all customer information analysis was developed in-house, using open and edge technologies, with off-the-shelf tools layered on top to fully leverage the high-quality data provided by the core stack.
  | #Marketing #Organisation #Tools | MarTech.org | 2025/08/01
• Amazon's Q2 2025 ad revenue rose 22% to $15.7B, outpacing retail's 13% growth to $167.7B. The key drivers: its DSP harnesses shopping, browsing, and streaming data with clean rooms for unmatched precision, reinforced by integration with Disney's DRAX. A Roku partnership expanded the largest authenticated CTV footprint, while Amazon's full-funnel reach spans marketplace, Prime Video, and Twitch, giving advertisers scale across 300M U.S. consumers. Combined with AI-powered contextual ads, creative tools, and data collaborations, these innovations explain why advertising grows faster than Amazon's retail base.
  | #Advertising #Leader | Marketing Dive | 2025/08/01
• According to Menlo Ventures, a San-Francisco–based firm, LLMs are moving beyond text generation into broader enterprise applications. Enterprises are shifting from model development to production inference, with code generation emerging as a standout success. Anthropic overtook OpenAI in enterprise share (32% vs. 25%), boosted by Claude 3.5 and 3.7, while Google also gained momentum. Open-source adoption declined (19%→13%) due to performance gaps, reinforcing closed-source dominance. Surveys highlight vendor loyalty, yet performance drives rapid switching—Claude 4 adoption jumped to 45%. Spending on model APIs surged from $3.5B to $8.4B in just six months, underscoring the market's explosive growth.
  | #AI #Leader #Organisation | Menlo Ventures | 2025/08/01
  
• Interesting point from a MarTech article: before buying AI tools, make sure they solve real marketing problems, not just look shiny. Involve your team to spot daily pain points. There's a difference between AI-native tools like Jasper (built for AI from the start, easier to scale) and AI-wrapped tools like MailChimp (added AI later but still useful). Native tools grow faster, but wrapped ones can also bring value. The key is clarity, teamwork, and choosing tools that really fix your needs. My take: since we have a new AI paradigm changing drastically possibilities, the question about what problem we want to solve is becoming crucial. Adding new tools without knowing their possibilities—and how the team must evolve with the technology—can be a very risky bet.
  | #Marketing #Organisation #Tools | MarTech.org | 2025/08/01
• According to MarTech, buyers have changed the way they act. In the past, they visited company websites, filled out forms, and asked for demos. Today, they first turn to AI tools, ask questions in private online groups, and check review sites before talking to a vendor. A company can lose a prospect before even realizing it was being considered. The old model no longer works: marketing, sales, and customer support spoke with different voices, sales teams waited for form sign-ups, and product built features no one noticed. AI disrupted this by giving buyers everything in one place. The new path requires clarity, proof, and speed.
  | #AI #Marketing #Sales | MarTech.org | 2025/08/01
• Amazon has blocked AI agents like ChatGPT Agent and Google Gemini from shopping on its platform, updating its robots.txt file to disallow automated purchases. The move caused ChatGPT Agent to hit "503 errors" when attempting to buy on Amazon, though it succeeded elsewhere, such as PetSmart. By restricting AI access, Amazon seeks to preserve control over customer experience and advertising. For AI developers, this limits agent usefulness, while smaller retailers may see opportunity by embracing AI optimization—potentially shifting e-commerce toward fragmented, proprietary shopping ecosystems. My take: Technically, AI agents could already be good enough to handle simple shopping tasks— definitely not "buy me a new car", but everyday essentials like a white 2-meter extension cord. The moment you can just say "get this" and it arrives at your door a few days later, for some users, that's a true assistant. With a price cap—say $10—many users won't bother if it's not the perfect choice; they'll value the saved time more. At that point, people stop caring where the purchase happens. E-commerce becomes just a delivery pipe. That's why big platforms want to block independent agents and instead build their own, keeping purchases inside their ecosystem.
  | #Agent #Marketing | TechRadar | 2025/08/01
• According to MarTech, AI could transform how companies deal with messy data. The old rule of cleaning everything before analysis is fading, as language models can extract meaning from unstructured sources. Advanced ETL workflows support this shift, but the focus now lies on intent rather than syntax. Untapped "digital droppings" like hidden website logs carry hidden value. Businesses that embrace dirty data can uncover insights, gain a competitive edge, and drive transformation. My take: It's hard to imagine running LLM-driven ETL on the core, mission-critical data volumes of a company. But for the messy sources that were never touched—because cleaning them was too costly—like hidden website logs or raw customer comments, this approach is ideal. Once an LLM surfaces meaningful signals, those can then be reinforced and integrated through proper ETL to make them enterprise-ready
  | #AI #Analytics #Marketing #Tools | MarTech.org | 2025/08/01
• In video games like StarCraft, actions per minute give players a clear edge in speed and strategy. Some experts see a similar pattern in AI usage: the more iterations you run, the faster you progress through tasks like transcripts or documents. But speed alone isn't enough—the real advantage comes from refining results, juggling multiple threads, and keeping perspective.
  | #AI #Organisation | nate.b.jones | 2025/08/01

  View on TikTok

• In digital marketing, privacy is becoming the new currency. The decline of cookies and stronger regulations reshape consumer expectations: brands must build trust through transparency and real value. The focus shifts to willingly shared data: first-party (interactions — purchases, logins, newsletters) and zero-party (data customers consciously and voluntarily declare — preferences, surveys, quizzes). Companies that replace hidden tracking with open value exchange can turn privacy from a legal obligation into a lasting competitive advantage.
  | #Analytics #Marketing | MarTech.org | 2025/08/01
• OpenAI added a checkbox in the ChatGPT app allowing users to share conversations that search engines like Google could index. Soon after launch, Fast Company uncovered numerous private chats—some containing sensitive personal and proprietary information—showing up in search results. In response to user backlash and privacy concerns, CISO Dane Stuckey called the experiment "short-lived" and disabled the feature for all users, then worked with search engines to remove any already-indexed content. My take: OpenAI tried a feature that let people make their private ChatGPT talks show up in Google searches. Many users didn't realize their chats could become public, and some private information appeared online. After only a few days, OpenAI turned off the option and began removing exposed chats from search engines. This event shows how easy it is to share more than intended when sharing rules are unclear. In the future, AI companies should use clear labels and strong defaults so people don't accidentally share private details.
  | #Leader | Fast Company | 2025/08/01
  
• AI is rapidly reshaping B2B buying. Chris Penn switched vendors in 15 minutes using Gemini, cutting costs in half and doubling output. The original provider never knew why they lost the deal. This new reality means shorter cycles, self-serve sign-ups, and no patience for gated pricing. Vendors hiding costs won't appear in AI shortlists, since algorithms don't call sales reps. Yet AI often prioritizes "helpful" over "truthful," so marketers and buyers must verify outputs, using AI as support, not autopilot.
  | #Marketing #Sales #Search | MarTech.org | 2025/08/01
• HubSpot has launched a CRM connector for Anthropic's Claude, following its earlier ChatGPT integration. The tool allows marketers, sales, and support teams to generate summaries, visualizations, and insights directly from HubSpot data, boosting productivity. Use cases include campaign optimization, deal prioritization, and ticket management. HubSpot stresses strong data permissions and privacy controls. The connector is available across paid Claude tiers, as Anthropic expands its ecosystem with tools like Canva and Google Workspace. My take: HubSpot's strategy is to surround its core product with as many LLM connectors as possible, aiming to make it plug-and-play with a company's existing stack. This contrasts with Salesforce and more traditional players focused on building proprietary LLM add-ons.
  | #Marketing #Sales #Tools | MarTech.org | 2025/08/01
• According to the latest report of We Are Social Agency, a global creative agency, search behavior is rapidly shifting as AI and video reshape digital habits. Google still attracts huge traffic, but referrals and outbound clicks fall with AI Overviews and zero-click results. ChatGPT and other AI platforms now capture billions of visits, disrupting traditional search. Video dominates attention: YouTube leads in engagement, ad reach, and traffic, while short-form content surpasses TV and streaming growth, especially among Gen Z. Streaming claims over half of TV watch time. Beyond media, online grocery rebounds globally, and mobile internet speeds rival broadband.
  | #AI #Marketing #Video | We Are Social | 2025/08/01
  
• Reddit plays a growing role in SEO and LLM training, shaping search results and AI outputs. Its authentic discussions drive organic traffic with conversion rates close to Google's. Success requires genuine, long-term participation, since overt promotions are quickly rejected. Instead of launching new subreddits, brands should join existing conversations. My take: moreover, Reddit is the largest network that LLMs can crawl freely.
  | #Authentic #Marketing #Search | DEAR STAGE 2 | 2025/08/01
• The new NotebookLM Video Overviews deliver narrated slide videos combining visuals, diagrams, quotes, and data for clearer understanding. Users can customize videos to specific learning needs. My take: Today, this is a state-of-the-art solution that transforms static PDFs into clear, engaging video presentations.
  | #Marketing #Tools #Video | Google | 2025/08/01
• Study of 17M citations shows AI assistants cite 25.7% fresher URLs, updated 13.1% more recently. ChatGPT, Perplexity, Gemini, and Copilot prioritize newer content, unlike Google AI Overviews. AI tools also order sources by freshness, whereas organic SERPs rank older articles higher. This highlights AI's bias toward recency. My take: for marketing, it means that being more reactive with new content could improve positioning in AI assistants.
  | #AI #Marketing #Search | Ahrefs | 2025/08/01
  
• Google reports massive AI adoption: AI Overviews now reach 2B users, Gemini counts 450M monthly, and AI Mode surpassed 100M with new features coming. Developers and creators accelerate growth, with 9M+ using Gemini and Veo 3 producing 70M videos. Productivity tools like AI notes in Meet engage 50M users. Yet, scaling to 980T tokens raises investor concerns about spending and uncertainty over true consumer-driven adoption. My take: Google is doubling down on AI—AI Overviews, Gemini, and AI Mode—to stay ahead of competitors and reshape the search interface. Yet around 80% of revenue still depends on Search and YouTube ads. AI Overviews risk changing search behavior, with some reports of reduced clicks, and Google hasn't outlined a clear monetization path. Balancing innovation with ad-driven revenue remains its biggest challenge.
  | #Leader | TechCrunch | 2025/08/01
• YouTube Shorts has launched new creative tools to make video production easier and more engaging. Users can now transform photos into videos, apply fun selfie effects, and experiment in the AI Playground to generate videos, images, and music. The rollout begins in the US, Canada, Australia, and New Zealand. My take: Another tool set that will fuel video content creation and accelerate influencer growth.
  | #Video | YouTube Blog | 2025/08/01