Newsfeed

Find the original sources, practical reading and people behind the collective. These links open the full public pages.

1,808 pages · Page 3 of 19

• Amazon adds a performance-data choice to Sponsored Ads exports

  Amazon's developer release feed dated 9 September describes a performance-data checkbox for Sponsored Ads bulk exports. It is selected by default; omitting performance data can make exports faster…

• September reminder: prepare Sponsored Brands Collections migrations

  Amazon’s guidance states that migration recommendations for legacy Product Collections will arrive in Campaign Manager in September, ahead of the format’s February 2027 deprecation.

• Amazon Sponsored Brands now supports 3 to 10 products per ad with AI curation

  Amazon updated Sponsored Brands product collections to allow 3–10 products per ad, up from the previous limit of 3. AI dynamically selects and highlights the most relevant products based on customer intent, though manual selection is still available.

• Amazon Sponsored Products gains bid adjustment for video placements

  Prem Gupta has flagged that Amazon has introduced the ability to adjust bids specifically for Sponsored Products video placements, enabling granular spend allocation by placement type within SP campaigns.

• Amazon Sponsored Products tests automatic ROAS optimisation in rule-based bidding

  Amazon is beta testing an "Automatically optimising ROAS within the campaign budget" option under rule-based bidding for Sponsored Products. You set the daily budget, and Amazon adjusts bids to maximise ROAS without a target floor.

• Amazon extends Sponsored Products to third-party sites and apps

  Amazon Sponsored Products expands beyond its marketplace to third-party websites and apps, according to the newsletter.

• Amazon stops accepting credit cards from select sellers starting 1 August

  Jon Elder has surfaced a selective Amazon policy change appearing on some sellers' accounts: Amazon will stop accepting credit cards from selected sellers starting 1 August 2026, with no broad announcement. As compensation, Amazon is offering a $2,500/month credit for 5 months to affected sellers: a transitional cushion rather than a structural fix.

• Amazon STV introduces Dynamic and Manual audience types with required CPM caps.

  Amazon Streaming TV ads now offer two audience selection options: Dynamic and Manual. Dynamic targeting uses AI to optimise reach, while Manual gives you direct control over audience segments.

• Amazon tells 1P vendors it will reverse tariff cost increases.

  Amazon has begun notifying 1P vendors that it will reverse tariff-related cost price increases, effective as of August 2026, presenting this as a change rather than a negotiation. Martin Heubel notes that a cost shown on a purchase order form is part of Amazon's contract terms once confirmed.

• Amazon tests AI-generated product summary on the PDP

  Prem Gupta has identified that Amazon is testing an AI-generated product summary appearing below the listing on the product detail page, automatically generated from existing listing content, including title, bullet points, description, and structured attributes.

• Amazon tests an AI-powered Full Funnel campaign type

  Amazon appears to be rolling out a new AI-powered campaign type called Full Funnel Campaigns. The format lets Amazon's AI coordinate Sponsored Products, Sponsored Brands, and Sponsored Display under a single strategy that spans discovery through purchase, rather than requiring advertisers to run and budget for each campaign type separately.

• Amazon tests an AI Targeting Strategy for Sponsored Products

  A new Targeting strategy beta appeared at campaign level in manual Sponsored Products campaigns, off by default. Enabled, AI adds up to four new keyword and product targets.

• Amazon tests brand logos above the product title on detail pages

  Vanessa Hung has flagged that Amazon is testing brand logos above the product title on select detail pages, replacing the small "Visit the Store" text link most buyers ignored. Placed before a shopper reads the title or price, the logo becomes an early trust cue: tapping it lets the buyer verify whether the current seller is authorised.

• Amazon tests hiding the Other Sellers section.

  Some Amazon product pages no longer show the "Other Sellers on Amazon" section beneath the Buy Box, even where multiple offers are active on the SKU. Amazon has not confirmed the change, and results vary across accounts and locations, suggesting a test that follows the Featured Offer eligibility change covered in Week 28.

• Amazon tests Rufus-first search experience on mobile

  Ivan Marynych has flagged Amazon testing a new mobile search flow that redirects some users away from traditional product results to Rufus, the AI shopping assistant. The test is currently limited to a small percentage of mobile searches, but the direction of travel is clear: Amazon is moving discovery from keyword-matched results toward conversational, AI-mediated shopping journeys.

• Amazon tests "Shop by feature" sponsored placements in search results

  Destaney Wishon has spotted a new "Shop by feature" sponsored ad placement appearing in Amazon search results. The format surfaces sponsored products organised by feature rather than strictly by the search term entered, expanding where sponsored products can appear beyond the direct intent of a query.

• Amazon tests the Hear the Highlights AI podcast feature on product pages

  Mansour Norouzi has flagged that Amazon is testing "Hear the Highlights": a feature that auto-generates short audio conversations between two AI hosts about a product listing. The hosts discuss product features, pull review themes, and answer shopper questions in real time.

• Amazon tightens seller-account transfer and financing terms

  Amazon's updated Business Solutions Agreement, effective 24 August, restricts sellers from transferring contractual rights or obligations without Amazon's consent and from pledging them as collateral. The wording could affect account sales, aggregators, and financing structures tied to future Amazon disbursements.

• Amazon Unmet Demand report surfaces low-conversion search clusters

  Isaac Gross has flagged Amazon's Unmet Demand section inside the Product Opportunity Explorer report, which highlights keywords where search volume is surging but conversion rates are low: indicating shoppers are actively searching but not finding the right product. The data is first-party Amazon shopper behaviour, not scraped or estimated, which makes it materially more reliable than third-party tooling for this kind of opportunity discovery.

• Amazon updates FBM shipping settings for multi-warehouse fulfilment

  Amazon has updated FBM shipping settings to better support multi-warehouse operations. The changes fix delivery promise accuracy when fulfilling from multiple locations.

• Amazon updates SP-API software pricing model

  Amazon has changed how software costs work for SP-API integrations. Most sellers see "$1,400/year plus usage fees" and assume their tools handle it, but the pricing shift affects how third-party software providers pass costs to sellers.

• Amazon Upfront 2026 runs alongside GML and IAB Upfronts week

  Amazon Ads ran its annual Amazon Upfront on Tuesday, 12 May 2026, in New York, unveiling its 2026–27 content slate, sports rights, ad-product roadmap, and a major measurement and creative push for Prime Video.

• Amazon Vendor Central adds bulk invoice-discrepancy reporting

  Amazon has updated the Financial Reports section of Vendor Central so 1P vendors can download financial data in bulk for a selected time range, flagged by Martin Heubel. The export includes purchase order IDs, invoice IDs, invoice status, identified invoice defects, and total deduction amounts, and separates price discrepancies from shortages so vendors can identify different root causes.

• Amazon walks back FBA fee increases after seller pushback

  Amazon has walked back a portion of the FBA fee increases announced last year, with Jason Mclellan flagging the change as "a big win for everyone in the Amazon community." Last year, Amazon announced plans for material FBA fee increases that would have meaningfully tightened seller margins through 2026; the walk-back reverses or softens specific fee categories in response to community feedback and seller pushback.

• Amazon Warehousing and Distribution launches in Europe

  Amazon introduced AWD in Germany, France, Italy, Spain, and the UK on 20 August. The programme offers bulk inventory storage and automated replenishment into FBA, reducing the need to manage every restock manually.

• Amazon warns sellers about "Frequently Returned" badge

  Amazon is now surfacing warnings about products at risk of earning a "Frequently Returned" badge. The badge can negatively impact conversion rates and visibility.

• An Ad Preview Hub arrives for Performance Max

  The new Ad Preview Hub lets advertisers preview how eligible Performance Max ads may appear across devices and placements before approval. For you, this means creative sign-off gets easier on Microsoft, so use the hub to catch rendering issues pre-launch.

• An Insights carousel replaces single recommendations

  Google Ads now shows a swipeable carousel of up to seven insight cards instead of a single recommendation. Cards cover budget issues, performance shifts, and campaign attention flags.

• An organic reporting dip appeared in Merchant Centre analytics around 11 August

  Merchant Centre analytics showed a dip in organic figures across multiple accounts from around 11 August, suggesting a reporting issue rather than a real change in performance. For you, this means the drop is probably not due to your listings, so hold off on corrective action until the data is confirmed to be clean.

• Andy Jassy confirms Alexa for Shopping rollout

  Andy Jassy confirmed at CEO level this week that Alexa for Shopping has rolled out broadly: bringing Rufus and Alexa+ together as a unified AI shopping assistant across phone, laptop, and Echo devices. In his words: "Alexa for Shopping is like having an expert personal shopper who already knows you and remembers your preferences, past purchases, and conversations, and carries that knowledge and understanding across your phone, laptop, and Echo devices."

• Anthropic CEO warns of AI "tsunami on horizon"

  Dario Amodei, CEO of Anthropic, warned that AI models are approaching human-level intelligence sooner than society recognises. In an interview on the WTF Is podcast, Amodei said: "It's as if this tsunami is coming at us, and it's so close we can see it on the horizon.

• Anthropic introduces Claude Fable 5.1 and Mythos 5.1

  Anthropic announced Claude Fable 5.1 and Claude Mythos 5.1, describing advances in coding, research and extended knowledge-work tasks.

• Anthropic publishes blueprints for shopping and merchant agents

  Anthropic released reference implementations for shopping and merchant agents across retail, travel, telecoms and ticketing.

• Anthropic discloses a US export-control directive affecting Fable 5 and Mythos 5

  Anthropic has disclosed a US government export-control directive, citing national-security authorities, that suspends access to its Fable 5 and Mythos 5 models for any foreign national, whether inside or outside the United States, including foreign-national Anthropic employees. To comply, Anthropic said it had to abruptly disable both models for all customers as of the evening of 12 June; access to all other Anthropic models remains unaffected.

• Anthropic outlines customer-controlled enterprise safeguards

  Anthropic announced Enterprise Frontier Safeguards, designed to let customers retain monitoring data in their own cloud infrastructure and review flagged activity through their own teams.

• Anthropic introduces Claude Science

  Anthropic has launched Claude Science, an AI workbench for research and analysis in beta. It ships pre-configured with access to more than 60 optional scientific databases.

• Anthropic launches Claude Fable 5 as a new frontier model

  Anthropic has released Claude Fable 5, its latest frontier model, now generally available on Google Cloud's Agent Platform, as noted by Google Cloud and discussed by Charly Wargnier in a comparison with Claude Opus 4.8. Fable 5 is part of a higher-capability tier and is positioned for enterprise teams building at scale, and is available across the same surfaces as other Claude models.

• Anthropic launches Claude Financial Services Suite

  Anthropic has released the Claude Financial Services Suite: a set of pre-built, deployable Claude agents for financial services workflows. Charly Wargnier has flagged the launch with its practical scope: ready-to-run agents that navigate analytical workflows from signal to decision, integrated natively with Microsoft 365 so finance teams can delegate manual work through Claude Cowork.

• Anthropic launches Claude Opus 5

  Claude Opus 5 approaches Fable 5's frontier intelligence at half the price, with state-of-the-art results on several coding and knowledge-work evaluations. It is positioned as thoughtful, proactive, and efficient.

• Anthropic launches Claude Sonnet 5

  Anthropic has introduced Claude Sonnet 5, its most agentic Sonnet model. It plans, uses tools such as browsers and terminals, and runs autonomously at a level previously requiring larger models.

• Anthropic launches Cowork for non-technical task automation

  Anthropic has released Cowork, a tool that lets users complete non-technical tasks similar to how developers use Claude Code. Cowork accesses a folder on your computer to read, edit, or create files.

• Anthropic publishes AI Fluency Index research

  Anthropic has published the AI Fluency Index, its first empirical measurement of how people collaborate with AI. The research tracked 11 behaviours across thousands of Claude.ai conversations, including how often people refine and iterate on work with Claude, question Claude's reasoning, and fact-check outputs.

• Anthropic releases free curriculum covering the AI ecosystem

  Anthropic has released a comprehensive curriculum of free courses covering Claude Code for CLI automation, MCP Mastery for building custom tools in Python, a complete API guide, AI Fluency frameworks for safe collaboration, and Claude 101 for core features. For you, this means structured learning paths are now available to maximise your Claude implementation.

• Anthropic restores Fable 5 and Mythos 5 with reduced capability

  Anthropic has restored global access to Fable 5 and Mythos 5 after the June export-control suspension. Fable returns with its capabilities reduced by stricter safety filters, and early benchmark tests report notable drops in performance on coding tasks.

• Anthropic's export-control suspension reframes the build-versus-rent calculation

  App Economy Insights has framed last week's events as a turning point for the AI market in its "Trillion-Dollar Off Switch" analysis. SpaceX's record IPO raised around $75bn at a valuation near $1.8tn, with its market cap surging past $2.5tn in its first two days.

• Anthropic's revenue run rate is reported to have overtaken OpenAI's

  Anthropic is now reported to generate a substantially higher annualised revenue run rate than OpenAI despite having a fraction of ChatGPT's users, flagged by Endrit Restelica. Anthropic's run rate, around $9bn a few months ago, is now reported at roughly $47bn, compared with OpenAI's roughly $25bn.

• Google App product reports add cost and conversions

  Google’s App campaign product-reporting documentation now lists Cost and Conversions alongside impressions, clicks and click-through rate.

• App Connect unifies web-to-app workflows with agentic implementation

  App Connect (formerly Web to App Connect) becomes the unified suite of tools that transforms apps into high-performance growth engines connected to Google Ads. The 2026 expansion includes agentic implementation workflows: intelligent assistants that find opportunities to improve setup and guide advertisers through automated processes for tracking, bidding, and deep linking.

• Prize-led app acquisition: measure the users a raffle brings, not only the installs

  David Vargas shares a practitioner observation that raffles can stimulate app installs. A prize can change the mix of people installing, so headline install volume alone cannot establish whether the tactic improves a sustainable acquisition programme.

• Apple Ads launches Insights, replacing Custom Report Builder

  Apple Ads has launched Insights, a new data analytics engine that replaces the platform's Custom Report Builder. The tool allows advertisers to explore performance across every dimension in their account: campaigns, ad groups, keywords, ad placements, and search terms: with predefined report templates for both standard performance views and advanced metrics, including Impression Share, rank, and search term popularity by country.

• Apple Ads' second position data surfaces in Insights, but reporting gaps frustrate practitioners

  David Vargas has reported that two weeks after Apple Ads' second-position ad placement launched, the Insights section shows impression share only for the first position, not for all positions combined, with no visibility into second-position appearances or direct performance. Vargas's own data shows that only one keyword gained +10% in impression share from the second placement, with CPA there ($9.41) higher than in the first position ($5.50).

• Apple announces Apple Maps Ads launching in the US and Canada

  Apple has announced paid advertising placements inside Apple Maps, launching initially in the United States and Canada. Local businesses will be able to bid on search queries: such as "coffee near me" or "pharmacy": and surface directly in Maps search results, with users able to call, navigate, or order directly from the placement.

• Apple commits to neutral ATT consent prompts

  Germany's competition authority made Apple's ATT commitments binding, requiring more neutral wording, content, and layout, more room for publishers to explain personalised advertising, and greater freedom to combine ATT with data-protection consent flows. Apple has four months to implement the changes; the commitments remain in force for seven years and are expected to cover almost all EU countries.

• Apple Maps launches ads for local businesses

  Apple Maps ads give local businesses a way to reach people searching for places and drive actions such as calls and visits.

• Apple passes $5tn in market cap after sitting out the compute race

  Apple became the second company to surpass $5tn in market value, having avoided the $700bn-plus AI compute arms race. Days later, it reported Q3 revenue up 16% to $109.4bn: iPhone up 22%, Mac up 29%, China up 22%: though shares fell about 7% on the outlook.

• Apple proposes 5% to 15% commissions on external link purchases

  Apple proposed new commission rates for purchases made through external links: 5% for small businesses, 15% for standard transactions, and 10% for subscriptions and partner programmes: down sharply from the 27% rate a court blocked in 2025. For you, this means the app-economy toll on web checkout is being renegotiated downward, so re-run the maths on steering users to web payment.

• Apple rebuilds Siri on Google Gemini at WWDC

  Apple has rebuilt Siri on Google Gemini, unveiled at WWDC 2026, flagged by Endrit Restelica. The overhauled assistant gains a dedicated app, remembers past conversations, can read what is on screen, and can take actions inside apps, working across Apple devices as one continuous thread.

• Apple Search Ads expands to multiple ad positions in search results from March 2026

  Apple will make Search Results campaigns eligible for multiple ad positions starting in March 2026, with the rollout worldwide by month's end. Ads can appear at any position, but you can't choose or bid for specific spots.

• Apple Search Ads is removing the CPA cap at the campaign level

  Apple is discontinuing the ability to set CPA caps at the campaign level in App Store search campaigns. Advertisers must now use account-level caps instead.

• Ask Advisor in Google Ads: revamped homepage, intelligent reporting, 24/7 security monitoring

  Ask Advisor in Google Ads is upgraded with five 2026 additions: Gemini 3 model upgrade, mobile app availability, instant certifications that replace weeks of manual paperwork with instant approvals (Anthony Higman flagged the certification surface this week appearing inside the Google Ads account), 24/7 security monitoring with personalised dashboards (flagged domains, dormant user review), and a revamped homepage with intelligent reporting and AI-powered data visualisation. The visualisation surface lets advertisers transform complex data into clear visualisations with a single prompt.

• Ask Advisor lands in the Merchant Centre as the merchandising expert assistant

  Ask Advisor in Merchant Centre launched in closed beta in the US in English: a Gemini-powered conversational assistant that diagnoses account and feed issues, surfaces performance and audience insights, and provides a roadmap to improve Store Quality Score. The advisor offers performance reporting and insights tailored to business goals, product listing and issue-resolution guidance for disapprovals, and personalised, actionable recommendations such as promotion opportunities or title optimisations.

• Ask Advisor reaches Google Merchant Centre

  The Ask Advisor chat, announced for Google Ads in May, is now appearing in Merchant Centre. Some accounts show an AI insights summary, a search box with suggested queries, and previous chats.

• Ask Advisor unifies the per-product advisors into a single agentic experience

  Ask Advisor is being unified across Google's marketing surfaces: Google Ads, Google Analytics, Merchant Centre, Display & Video 360, Search Ads 360, Campaign Manager 360, and soon more: into a single agentic experience that travels with the user across products. The unified advisor remembers business goals and context, surfaces the right product for each conversation, and proactively suggests next steps.

• Ask Maps brings conversational local search to Google Maps

  Ask Maps lets users query Google Maps conversationally, described by analysts as the strongest local UI change of the AI era and a product that keeps local queries inside Google's ecosystem. For you, this means local discovery shifts further into AI dialogue, so keep business profiles complete as answers draw on them.

• Asset Optimisation panel gets a refreshed layout

  Google Ads has updated the Asset Optimisation interface with a cleaner design and a new Manage opt-ins page. From this panel, advertisers can now control shorter videos (auto-generated from existing assets), resized videos (in different aspect ratios), and landing page previews (images pulled directly from your landing page).

• Asset Studio becomes the unified creative hub powered by Gemini, Veo, Nano Banana, and soon Gemini Omni

  Asset Studio is launching as the centralised creative hub inside Google Ads: bringing existing brand assets together, generating new creative with Google AI, and enabling one-click testing. The 2026 expansion includes four structural additions: Brand Guidelines (PDF Upload): upload free-form marketing briefs and PDFs to act as a permanent anchor for the brand's style, colour, and business info; Google Assets ingestion from Merchant Center, YouTube Studio, and Pomelli; Third-Party Assets integration with Adobe and Canva via the Google Ads API; and Asset Generation in the Google Ads API for programmatic integration into custom campaign platforms.

• Asset Studio can prefill Performance Max asset groups

  Google has documented a way to prefill Performance Max asset groups with material generated in Asset Studio, including headlines, descriptions, images and videos.

• Asset Studio turns websites and ad assets into video

  Google Ads is rolling out a beta workflow that converts assets from a website, ad, or asset group into short horizontal or vertical videos. Dario Zannoni's account test produced a 12-second, five-scene video with voiceover, while Google's August Demand Gen update describes wider multimodal video creation and editable storyboards.

• Attributed brand search conversions arrive as a Google Ads metric

  Google has added an attributed brand search conversions metric to reporting. It credits campaigns for the brand searches they generate, giving upper-funnel activity a measurable path to conversion.

• Attributed Branded Searches connects YouTube ad exposure to branded search lift

  Attributed Branded Searches (ABS) is now generally available globally as the "always-on" metric that connects YouTube ad exposure to subsequent branded searches. The metric counts users who saw a YouTube ad and subsequently performed a branded search, serving as a rapid leading indicator for campaigns aimed at driving brand interest and market share.

• Audience generation via Copilot prompts

  Microsoft previewed prompt-based Audience generation inside the campaign creation flow. Instead of manually picking attributes, advertisers describe their ideal customer in natural language ("Who's your target audience?

• Audience signals can be assigned to individual creatives

  Bram Van der Hallen spotted an option to choose age and gender signals for each creative within a single ad set. Meta can still deliver beyond the selected audience when it predicts a better response, so the setting is a signal rather than a strict targeting boundary.

• Austin Lau (Anthropic) builds a Google Ads MCP management system using Claude Cowork

  Austin Lau, a performance marketing PM at Anthropic, has published a detailed account of building a custom Claude Cowork plugin to manage Google Ads via local MCP: encoding his personal search-term review methodology as a reusable skill that connects to the Google Ads API. The plugin outputs a bulk-upload-ready CSV for Google Ads Editor with a reasoning column for each negative keyword recommendation.

• Australia opens consultation on its next privacy reform bill

  Australia released an exposure draft of the Privacy Amendment (Personal Data Protection) Bill 2026 and an accompanying consultation paper on 31 August, with submissions due by 18 September.

• Austria's Supreme Court orders Meta to release user data after 11-year legal battle

  Austria's Supreme Court has ruled that Meta must provide privacy campaigner Max Schrems with access to all of his personal data within two weeks. The decision ends an 11-year legal case and includes €500 in damages.

• Auto end screens rolling out for video ads

  Google Ads has introduced auto-end screens that display product or service details after video ads finish. The feature pulls information from your campaign, including app name, icon, price, and install link.

• AWS and OpenAI expand partnership with Bedrock integration

  AWS and OpenAI have announced a major expansion of their partnership, bringing OpenAI's latest models into Amazon Bedrock, launching Codex on Bedrock for AI-powered software development at scale, and introducing Amazon Bedrock Managed Agents powered by OpenAI for production-ready agents in the cloud. All three components are in limited preview.

• AWS UAE data centre hit amid Iran strikes, mec1-az2 offline

  Amazon Web Services reported that one of its UAE Availability Zones (mec1-az2) in the ME-CENTRAL-1 region went offline after "objects struck the data centre, creating sparks and fire." The fire department shut off power while extinguishing the fire, with AWS stating restoration will take several hours.

• BatchJobService request size increased fourfold

  Google has raised the request size cap for BatchJobService.AddBatchJobOperations from approximately 10.5 MB to approximately 41.9 MB. The change reduces the number of separate requests required when batching multiple large assets: particularly images, which can each reach up to 5.24 MB.

• Best Buy checkout appears inside ChatGPT

  Juozas Kaziukėnas spotted a Best Buy-controlled checkout flow that lets shoppers complete an order without leaving ChatGPT, using an MCP-powered experience rather than a generic external checkout. The observed flow indicated that some membership benefits and reward points were unavailable.

• Bid Strategy Report adds new performance metrics

  The Bid Strategy Report gains new metrics: Avg. Target ROAS, Avg.

• BigQuery adds a persistent bytes processed limit

  BigQuery Studio now allows users to set a persistent "maximum bytes processed" limit that follows their user profile across projects. The setting acts as a personal safety net, preventing massive data scans triggered by typos or exploratory queries from generating unexpected costs.

• BigQuery adds multi-level aggregation

  Google has added multi-level aggregation to BigQuery SQL, letting you nest one aggregate function inside another in a single query, as flagged by Luka Cempre on the GA4 export. It removes the subquery workarounds analysts used for aggregate-of-aggregate calculations.

• BigQuery adds time-series functions to the query editor

  BigQuery's 20 August release notes added preview support for table-valued functions, including ML.TREND, ML.SEASONALITY, and ML.DETECT_CHANGE_POINTS in the query editor and conversational analytics. The functions are designed to expose patterns and anomalies in time-series data through reusable queries.

• BigQuery adds update-based triggers for scheduled queries

  Google has launched update-based triggers in BigQuery, flagged by Johan Strand and Luka Cempre. Scheduled queries can now run when a table changes, rather than only on a schedule, so aggregated tables update as soon as new data lands in the warehouse, not waiting for the next scheduled run.

• BigQuery previews analysis of changes after a campaign launch

  BigQuery introduced AI.CAUSAL_EFFECT in preview on 10 September. The function compares results after an intervention with a forecast based on earlier observations, letting analysts examine a…

• BigQuery conversational analytics reaches general availability

  Google has made BigQuery conversational analytics generally available, flagged by Luka Cempre. The capability lets teams build data agents on top of BigQuery data: tables, views, and other sources: and query them in natural language, lowering the barrier to self-serve analysis.

• BigQuery adds market basket questions to conversational analytics

  Google has made market basket analysis questions generally available in BigQuery conversational analytics, allowing users to explore which products are commonly purchased together through a conversational interface.

• BigQuery Global Queries enable cross-region joins

  BigQuery has launched Global Queries in preview, allowing analysts to join datasets across different regions in a single SQL statement without manually moving data. Alicia Williams tested the feature by joining bike-sharing data from Austin (stored in the US region) and London (stored in the EU region) to compare rental patterns by hour.

• BigQuery Graph capabilities launch in Preview, enabling GQL queries without ETL

  Axel Thevenot has flagged that Google Cloud has launched BigQuery Graph capabilities in Preview, enabling Graph Query Language (GQL) to be used directly within BigQuery on existing tables with no ETL required. SQL and GQL can be mixed in the same workflow, built-in graph visualisation is included, and the feature is optimised for large-scale analytical queries.

• BigQuery launches "chat with your data" at the dataset level

  Balazs Vajna has flagged that BigQuery is rolling out "chat with your data" at the dataset level: a shortcut to Conversational Analytics (the Agents feature) that's spreading across the BigQuery interface. The capability lets analysts launch a chat-based exploration on entire datasets rather than only query results.

• BigQuery Studio adds Git repository integration

  Luka Cempre has flagged a new BigQuery Studio capability: Git repositories integration is now available in preview, providing direct integration to remote Git repositories. The setup requires enabling APIs, naming the repository, and authenticating with GitHub: within a few minutes, a GitHub repo is synced with BigQuery Studio.

• BigQuery Studio notebook gallery launches general availability

  Alicia Williams has announced that the BigQuery Studio notebook gallery is generally available, providing pre-built templates that remove the boilerplate setup required to start a new analysis. Templates cover SQL and no-code visualisation cells, Python and BigQuery DataFrames workflows for cleaning and machine learning, Spark SQL guides for serverless analysis, and multimodal generative AI templates using Gemini.

• BigQuery surfaces recent queries per tab.

  Luka Cempre spotted a BigQuery UI update that makes recently run queries easily accessible alongside their results, apparently independent for each tab. A small quality-of-life change that saves jumping into query history to manage jobs across many open tabs.

• BigQuery previews TabFM predictions without custom model training

  BigQuery now offers a preview of TabFM, Google’s pretrained model for structured tables, supporting regression and classification without training a separate model for each task.

• Bing ads pop out of the results as you scroll

  Bing is showing ads that visually pop out during scrolling. The animated treatment appeared for multiple observers this week.

• Bing is testing Asian owned tag for Shopping ads

  Microsoft Bing has been spotted testing an "Asian owned" business tag on Shopping ads, similar to other business attribute labels. For you, this means watching for potential business identity tags that could differentiate your listings.

• Bing labels a good time to buy using price history

  Khushal Bherwani has flagged Bing analysing price history inside the product detail window and labelling some items as a good time to buy. The cue sits with the product before the shopper decides.

• Bing ships an AI opt-out toggle for search results

  Microsoft has shipped a preview extension that lets Bing users turn Copilot-style AI answers on or off with a single click, announced by Microsoft President and Head of Search Jordi Ribas. Users can also append "-ai" to any Bing query to suppress AI-generated responses for that search.

• Bing tests a new images section in the results.

  Bing is testing an enhanced images section within its search results, as spotted by Sachin Patel. The test adds visual content to result pages where it previously showed standard listings.