Mel is our Partner Strategy & Delivery Manager and also a CIM Chartered Marketer, a testament to her commitment to excellence in the field. But Mel’s contributions don’t stop at the office door. Beyond her professional endeavours, she leads an active life as a qualified run leader and dedicated volunteer. Her experience in these roles has streamlined her leadership and teamwork skills, making her an invaluable asset when it comes to collaborating on projects and ensuring their success. Her sharp insights, strategic thinking, and knowledge have made her a backbone in our team’s ability to drive results for clients in this industry. Mel will make sure that we can approach marketing challenges from all angles and deliver outstanding results for our clients.
Posted on 11/09/2026 by Melanie Comerford
Latest Digital Marketing News: SEO, PPC, AI & Web Updates

Digital marketing changes quickly. Each week, Loop Digital rounds up the SEO, PPC, AI, content, and web developments that could affect how businesses market themselves online.
We focus on the developments worth knowing about, what they mean in practice, and where businesses may need to act.
Read time: 22 minutes
Updated 14th September 2026
AI search visibility is becoming easier to see, but understanding what that visibility actually delivers is still difficult. Google has acknowledged limitations in how Search Console reports AI Search, while new research using Wikipedia data adds to the debate around AI Overviews and referral traffic.
Elsewhere, a UK charity study suggests brand size alone does not guarantee AI visibility, content marketing performance has fallen to a 12-year low, and Amazon is testing another route for advertisers to access ChatGPT Ads.
Contents
- AI Search Visibility Is Easier to See, but Harder to Measure
- UK Charity Study Shows Brand Size Does Not Guarantee AI Visibility
- Content Marketing Success Falls to a 12-Year Low
- Amazon Opens Another Route Into ChatGPT Ads
- Key Takeaways
- Identify the Marketing Changes That Require Action
- What Should Your Business Act On Now?
- Stay Ahead of Digital Marketing Change
AI Search Visibility Is Easier to See, but Harder to Measure
Google’s generative AI performance reports in Search Console now show impressions from AI Overviews, AI Mode and generative AI features in Discover. But the data still has important limitations.
Google’s John Mueller has acknowledged that traditional Search Console metrics do not always translate neatly to AI Search. Links inside an AI Overview inherit the position of the overall feature rather than receiving an individual position, while standard impression rules do not necessarily show whether a user actually saw a particular link.
That matters because a University of Washington working paper estimates that the introduction of AI Overviews reduced monthly external-search referrals to English Wikipedia by around 5%. The research compared English Wikipedia with German and French editions during the initial US rollout.
However, the paper has not been peer reviewed and Google disputes whether the methodology can isolate the effect of AI Overviews. The 5% figure should therefore be treated as something worth monitoring, not as a benchmark for every website.
What should your business do?
- Record a baseline for AI Search impressions, pages and countries in Search Console.
- Compare AI visibility with organic clicks, landing-page performance and conversions rather than impressions alone.
- Avoid treating average position inside an AI Overview like a traditional organic ranking.
- Monitor important pages and queries over time rather than drawing conclusions from a single reporting period.
- Review unexplained changes in Direct and organic traffic alongside AI Search visibility.
Loop Digital’s Generative Engine Optimisation services focus on how businesses are represented across generative search while keeping visibility connected to wider search and commercial performance.
UK Charity Study Suggests Brand Size Does Not Guarantee AI Visibility
A study of 15 prominent UK charities found that strong brand recognition did not automatically translate into strong AI Search visibility. Cancer Research UK, British Heart Foundation, RSPCA and Mind scored highly, while some other well-known organisations performed more modestly.
The report suggests broader organisations can be associated with many subjects without establishing the same depth of authority around one particular area. More focused organisations may therefore have an advantage for specific questions.
Technical accessibility matters too. Useful research buried in PDFs or content heavily reliant on JavaScript can be harder for AI systems to access than clear, structured HTML content.
The same lesson applies to businesses. Brand recognition helps, but AI systems still need clear evidence of what you do, what you know and where you have genuine expertise.
What should your business do?
- Identify the products, services and subjects where you have genuine expertise.
- Answer specific customer questions rather than relying only on broad category content.
- Turn useful research, evidence and expertise into accessible website content.
- Keep important information in clear, crawlable HTML wherever possible.
- Test how AI platforms represent your business for relevant category and comparison questions.
Loop Digital’s Answer Engine Optimisation services focus on making useful information clear, accessible and easier for answer-led systems to understand.
Content Marketing Success Falls to a 12-Year Low
Only 14% of marketers surveyed by Orbit Media said their blogs were delivering strong results, the lowest figure recorded across the survey’s 12-year history. AI adoption, meanwhile, has passed 92%.
That does not mean AI is responsible for weaker content performance. The survey found that several practices historically associated with better results have become less common, including original research, keyword research, expert collaboration, paid promotion, formal human editing and consistent use of analytics.
The distinction matters. AI can reduce the time required to research, structure or produce content, but publishing more quickly is not the same as publishing something useful, distinctive or commercially effective.
What should content teams do?
- Use AI to support the content process rather than replace research, expertise and judgement.
- Keep human editing and fact-checking in the publishing workflow.
- Add first-party experience, expert insight, original data or analysis where it genuinely improves the piece.
- Research what customers are actually searching for and asking before producing content.
- Measure qualified traffic, enquiries and conversions rather than publishing volume alone.
Loop Digital’s content marketing services focus on useful, user-centric content built around search demand, expertise and measurable business goals.
Amazon Opens Another Route Into ChatGPT Ads
Amazon is piloting access to ChatGPT Ads through Amazon DSP, allowing selected US advertisers to extend campaigns into ChatGPT through Amazon’s advertising platform.
Amazon is providing the buying and campaign-management route rather than taking control of where ads appear inside ChatGPT. The pilot is currently limited to selected advertisers in the US.
For UK advertisers, there is nothing to activate yet. The bigger point is that ChatGPT Ads is beginning to connect with established advertising platforms rather than developing as a completely separate channel.
What should advertisers do?
- Treat the Amazon DSP integration as a pilot rather than a mature advertising channel.
- Monitor whether access expands beyond selected US advertisers.
- Make sure conversion and attribution measurement is ready before testing emerging AI ad placements.
- Compare qualified leads, sales and acquisition cost rather than clicks alone.
- Consider how conversational advertising could fit alongside Search, paid social and other existing channels.
Loop Digital’s paid search and PPC services focus on lead quality, acquisition cost and measurable return across paid-media activity.
Key Takeaways
- Search Console can now show AI Search visibility, but its existing impression and position metrics still have important limitations.
- A working paper estimates AI Overviews reduced search referrals to English Wikipedia by around 5%, but the finding should not be treated as a general website benchmark.
- UK charity research suggests brand size alone does not guarantee AI visibility; focused expertise and accessible content still matter.
- Content marketing performance has reached a 12-year low despite widespread AI adoption.
- Amazon is piloting another route into ChatGPT Ads through Amazon DSP.
Identify the Marketing Changes That Require Action
This week’s updates show that visibility is becoming easier to track, but harder to interpret.
Businesses may now be able to see more of where they appear in AI Search, but that does not automatically explain how much traffic or commercial value that visibility creates. At the same time, the charity and content marketing research reinforces something more familiar: clear expertise, useful information and strong measurement still matter.
Loop Digital’s free digital marketing audit can help identify:
- Where AI and organic visibility are growing without a clear link to performance
- Content that lacks enough expertise, evidence or differentiation
- Gaps in how important information is structured and presented
- Reporting that focuses too heavily on visibility rather than outcomes
- Paid-media opportunities that need stronger measurement before testing
Claim Your Free Digital Marketing Audit
What Should Your Business Act On Now?
Review
- Which pages are gaining AI Search visibility and whether that is reflected in traffic, enquiries or conversions.
- How clearly your website demonstrates expertise around your most important products, services and topics.
- Whether useful information is easy to access in HTML rather than buried in PDFs or difficult-to-crawl formats.
- Whether faster content production is actually improving performance.
- Whether your current reporting can measure new paid channels properly if opportunities become available.
Prepare
- A baseline for AI Search visibility so future changes are easier to interpret.
- Stronger first-party content through case studies, expert input, original data and practical experience.
- More specific content around the questions customers actually ask.
- Better measurement around qualified leads, conversions and acquisition costs.
- A clear testing framework for emerging AI advertising channels when they become relevant.
Monitor
- How Google develops Search Console reporting for AI Search.
- Further evidence around AI Overviews and changes in referral traffic.
- Which brands and sources AI platforms continue to cite and recommend.
- Whether ChatGPT Ads expands beyond its current pilot activity.
- Whether content teams begin seeing stronger results by combining AI efficiency with research, expertise and human editing.
Not every development this week requires an immediate change. The priority is to improve how visibility is measured, strengthen the signals that demonstrate expertise and make sure new channels are judged by the results they produce.
Stay Ahead of Digital Marketing Change
AI is creating more places for businesses to appear, from AI Search results to conversational advertising. But more visibility also means more complexity when it comes to understanding what is actually working.
The businesses best placed to benefit will be the ones that can connect strong content, clear expertise and reliable measurement rather than chasing every new platform change in isolation.
Loop Digital helps businesses bring that together across SEO, AEO, GEO, content marketing, PPC and wider digital strategy, with a focus on visibility that leads to measurable growth.
Book Your Free 30-Minute Strategy Call
Updated: 7th September 2026
Two things stand out this week. What AI systems say about your business when the evidence is thin, and what happens to paid performance when AI Overviews sit between the searcher and the result.
WordPress has set out how it will handle a rising volume of security reports, Google is testing conventional Search campaigns inside AI Mode, and an analytics fault made 1st September look like a day with no traffic at all.
Contents
- WordPress Launches a Core Security Initiative
- When AI Has Nothing on Your Business, It Describes Someone Else
- ChatGPT Now Draws More US Traffic Than Bing or DuckDuckGo
- AI Overviews May Be Affecting Shopping Ad CTR and Impressions
- Google Tests Traditional Search Campaigns in AI Mode
- Google Analytics Showed Zero Traffic for 1st September
- What Anthropic’s Watermarking Means for Content Teams
- Preferred Sources and “Highly Cited” Labels in AI Search
- Key Takeaways
WordPress Launches a Core Security Initiative
WordPress has started a coordinated programme called the Core Security Initiative. It covers three things: a tighter release process, more volunteers on the report backlog, and AI-assisted scanning to catch vulnerabilities early. WordPress says report volumes have risen because AI has made it easier to find weaknesses in code, an advantage attackers share.
What should website owners do?
- Give WordPress and plugin updates a named owner and a schedule.
- Check a recent backup exists and has been tested.
- Note which plugins are no longer maintained.
Loop Digital’s hosting and maintenance services include WordPress updates, backups and security monitoring.
When AI Has Nothing on Your Business, It Describes Someone Else
Duane Forrester makes a sharp point that when a model lacks evidence about your company, it substitutes a competitor or a category average and states it with full confidence. The weakest test is asking a model to describe your business by name. Substitutions show up on category and comparison questions instead. Rich Sanger’s AI entity footprint audit offers a related framework, scoring identity, differentiation, evidence, consistency, relationships and specialisation.
What should your business do?
- Test category and comparison questions where your name is absent.
- Note any claim a model makes that no source of yours supports.
- Check your website, Google Business Profile and directories describe the same business.
- Build independent evidence: reviews, case studies, third-party coverage.
Loop Digital’s Generative Engine Optimisation services focus on how brands are understood across generative search.
ChatGPT Now Draws More US Traffic Than Bing or DuckDuckGo
Semrush data analysed by Greg Jarboe puts ChatGPT.com ninth among US websites, up 48.38% year on year. Bing fell 50.43%, DuckDuckGo 15.83%. One publisher’s Direct traffic had grown to 78% of sessions, a reminder that AI-generated links often arrive without referrer data.
What should your business do?
- Check whether your share of Direct traffic has been climbing.
- Compare GA4 against Search Console instead of trusting either alone.
- Ask new enquiries how they found you, and record it.
AI Overviews May Be Affecting Shopping Ad CTR and Impressions
Mike Ryan of Smarter Ecommerce has spotted Shopping CTR rising while impressions fall. Clicks stayed flat, so the CTR gain reflects fewer impressions, not better performance. His theory: Google may be serving AI Overviews more on queries less likely to earn an ad click. He is clear it is a hypothesis, not proof.
What should advertisers do?
- Report impressions and CPA alongside CTR, not CTR alone.
- Check whether a rising CTR is matched by falling impressions.
- Re-baseline benchmarks set before AI Overviews became common.
Loop Digital’s paid search and PPC services focus on lead quality and measurable return.
Google Tests Traditional Search Campaigns in AI Mode
Google has begun testing exact and phrase match keywords serving text ads in AI Mode, limited to queries with explicit, direct intent. Advertisers may reach AI Mode without moving to AI Max or Performance Max. Google hasn’t said how widely the test runs.
What should advertisers do?
- Carry on with AI Max migration planning, as this changes no timelines.
- Watch for shifts in exact and phrase match performance.
- Resist loosening match types on the assumption broader targeting is now needed.
Google Analytics Showed Zero Traffic for 1st September
A widespread GA4 fault left many properties showing no data for 1st September. It was a processing issue at Google’s end, not a real traffic drop.
What should marketing teams do?
- Annotate 1st September in GA4 and any reports that use it.
- Check monthly and rolling averages before circulating them.
- Cross-check against Search Console and ad platforms, which were unaffected.
What Anthropic’s Watermarking Means for Content Teams
Steve Liu unpacks what Anthropic’s watermarking does and doesn’t prove. It’s statistical, applied worldwide, and responds to the EU AI Act. The key point: it shows text was processed by Claude, not authored by it. A light edit can carry the same signal as a full draft.
What should businesses do?
- Record where AI supports content, and at which stage.
- Keep named human review and fact-checking on anything published.
- Decide which content needs an extra approval step.
Preferred Sources and “Highly Cited” Labels in AI Search
Google’s Preferred Sources now appear inside AI Overviews and AI Mode, alongside a “Highly Cited” badge for widely referenced articles. The feature highlights original reporting that has been referenced by other publishers and can also show when an article links to a highly cited source.
The label is not a universal ranking badge. It is another reason for publishers to create distinctive, well-sourced work rather than repeat information already available elsewhere.
What should publishers do?
- Lead with original research, first-party data or useful expert analysis.
- Link clearly to primary sources.
- Keep authorship, publication dates and updates easy to verify.
- Measure links, mentions and qualified traffic rather than chasing the badge alone.
Key Takeaways
- WordPress has launched a Core Security Initiative covering releases, backlog and AI-assisted scanning.
- When AI lacks evidence about a business, it substitutes a competitor, and a content audit won’t reveal it.
- ChatGPT.com is now the ninth most visited US website, ahead of Bing and DuckDuckGo.
- Shopping CTR is rising while impressions fall, with AI Overviews a plausible but unproven cause.
- Google is testing exact and phrase match Search ads in AI Mode on high-intent queries.
- A GA4 fault showed zero traffic for 1st September and needs annotating.
- Anthropic’s watermark shows text was processed by Claude, not authored by it.
- Preferred Sources and “Highly Cited” labels now appear across Google’s AI search experiences.
Updated: 31st August 2026
Google completed two rollouts that change what website owners can see and control in AI search, while new research pointed to where AI citations are actually being earned.
Search Console’s AI reports went worldwide, ChatGPT became the first AI chatbot designated a search engine under EU law, and product pages look to be doing more work in AI answers than community platforms.
Search Console AI Performance Reports Roll Out Worldwide
Google confirmed its generative AI performance reports and opt-out control are now live for all websites. Reports show impressions from AI Overviews, AI Mode and Discover, by page, country and date. No click data yet. The UK’s CMA required this opt-out and expects click data by March 2027.
What should your business do?
- Open the AI performance report and record a baseline.
- Compare AI impressions against organic clicks for the same pages.
- Treat opting out as a decision with a cost; it removes visibility too.
Google Is Dynamically Expanding AI Overviews for Some Queries
Google confirmed AI Overviews may expand to their full state for some queries, removing the “Show more” button and pushing standard results further down the page. This gives the AI answer more space and can push traditional organic results further down the page.
What should your business do?
- Check which priority queries now show an expanded AI Overview.
- Compare impressions against clicks on those queries in Search Console.
- Make sure the answer a searcher needs is stated clearly in your content.
Google Fixed a Citation Bug in Gemini 3.8 Flash
Google briefly experienced a citation problem after Gemini 3.8 Flash was introduced into AI Mode, with some answers showing few or no links. Google confirmed the behaviour was unintended and fixed the problem shortly afterwards.
What should your business do?
- Avoid drawing conclusions from a single week of AI citation monitoring.
- Record the platform or model version alongside your tracking.
- Separate platform faults from genuine visibility changes.
Product Pages Take 24% of AI Citations in One B2B Study
A Ten Speed study found product pages accounted for 24.1% of AI citations at the evaluation stage, against 4.2% for Reddit, YouTube and forums combined. Pages a business owns took 88.3% of citations overall. Comparison content punched above its weight. The caveats: descriptive rather than tested, and drawn from one agency’s B2B clients.
What should your business do?
- Write service and product pages so a stranger could explain what you do.
- Build comparison content deliberately, not just against one rival.
- Keep directory and marketplace profiles current.
Loop Digital’s Answer Engine Optimisation services focus on the clarity of information answer-led systems draw on.
ChatGPT Is Searching Individual Subreddits by Name
Suganthan Mohanadasan captured ChatGPT targeting one subreddit by name, giving it six of eight citations. Four days earlier, a different query pulled 84 Reddit threads and cited none. Reddit citation looks query-dependent rather than switched off, based on four queries over one week.
What should your business do?
- Decide where community presence genuinely helps, usually earlier in the journey.
- Invest in owned pages for questions asked when choosing a supplier.
- Track citation sources by query type, not as one overall figure.
ChatGPT Is Now a “Very Large Online Search Engine” in the EU
The European Commission designated ChatGPT a very large online search engine under the DSA, the first AI chatbot to receive it. OpenAI reported around 159.1 million average monthly EU recipients. There are four months to meet added obligations, including a public advertising repository.
What should your business do?
- Note the requirement if you advertise, or plan to, in EU markets through ChatGPT.
- Watch for the advertising repository, which will show competitor activity.
- Keep UK and EU compliance positions separate in planning.
AI Is Changing Website Security
Over 100 organisations, including OpenAI, Google and Cloudflare, warn that AI-enabled cyberattacks will become far more widespread. AI lets attackers move faster through weaknesses that already exist. A compromised site produces spam pages, redirects and crawling failures, all of which cost visibility.
What should your business do?
- Ask whoever manages your site to audit the codebase and dependencies.
- Keep packages, libraries, themes and plugins current.
- Set up monitoring and alerts so a compromise is spotted quickly.
Loop Digital’s web design services build and maintain sites that support search and paid performance.
AI Mode Shows Different Sellers and Prices to the Product Carousel
Productrise tracking found only 1.28% of carousel products also appeared in AI Mode for the same search. Where both appeared, the first-listed seller differed nearly half the time. Google says both surfaces use the same Shopping Graph but hasn’t verified the findings.
What should retailers do?
- Search your top products in both AI Mode and standard results, and note who appears first.
- Check your product feed is complete and accurate.
- Treat AI Mode as a separate surface in reporting.
Google Brings Agentic Video Understanding to Ask YouTube
A new video analysis system will power Ask YouTube on the watch page, letting Gemini inspect specific frames, audio and transcript sections rather than sampling a whole video. No confirmed date yet.
What should marketing teams do?
- Say important information out loud rather than leaving it in on-screen text.
- Keep accurate captions and transcripts.
- Add clear chapters to longer videos.
YouTube Adds Amazon Product Tagging for Affiliate Creators
YouTube has expanded its US shopping affiliate programme to let creators tag Amazon products in videos, Shorts and livestreams. YouTube reports 110% more product clicks with shopping tags versus description links. Note: this expansion is US-only; the affiliate programme itself opened to UK creators earlier in August.
What should brands do?
- Check whether your creators are already in the shopping affiliate programme.
- Clarify commercial terms, since a creator may now earn twice from one video.
- Keep disclosure requirements explicit in creator agreements.
Key Takeaways
- Search Console’s AI performance reports and opt-out control are live worldwide, with impressions but no click data.
- AI Overviews now expand fully on some queries, pushing standard results down the page.
- A short-lived Gemini 3.8 Flash bug removed citations from AI Mode before Google fixed it.
- One B2B study put product pages at 24.1% of AI citations against 4.2% for Reddit, YouTube and forums.
- ChatGPT is targeting individual subreddits by name; Reddit citation looks query-dependent.
- ChatGPT is now designated a very large online search engine under the EU’s Digital Services Act.
- Over 100 organisations have warned that AI-enabled attacks will accelerate.
- AI Mode rarely shows the same products as the carousel, and often a different seller.
- Ask YouTube is getting a new video analysis system; YouTube has added Amazon tagging for US creators.
Updated: 24th August 2026
Website security and paid-media deadlines require immediate attention this week, while changes across Search, AI products and analytics show how quickly digital discovery continues to evolve.
Elementor Pro users have a critical patch to apply, Google has completed its August spam update, and advertisers have several AI Max and Performance Max changes to review. Google and OpenAI have also introduced developments spanning personalised search, social and video reporting, teen safety, advertising and AI-generated results.
Critical Elementor Pro Vulnerability Requires an Immediate Update
The flaw, CVE-2026-32475, carries a CVSS score of 9.8 and affects Elementor Pro 4.2.1 and earlier. Exploitation depends on a published Elementor Pro Form widget containing at least one File Upload field that is not marked as required. Elementor released version 4.2.2 on 19th August to address it.
The researcher who reported the issue received a $15,600 bounty. The more important point for website owners is that the attack does not require an account when the vulnerable form configuration is present.
What should website owners do?
- Check the Elementor Pro version and update to 4.2.2 or later immediately.
- Identify published forms that contain a non-required File Upload field.
- Verify recent backups and confirm that a restore has been tested.
- Review upload folders, administrator accounts and server logs for unexpected activity.
- If compromise is suspected, isolate the website and involve a qualified security specialist.
Loop Digital’s hosting and maintenance services include WordPress updates, backups and security monitoring.
Google Completes Its August 2026 Spam Update
Google’s August 2026 spam update ran globally, across all languages, from 18th to 21st August.
Google did not announce a new spam policy with the rollout. A ranking change during this window is therefore a reason to investigate, not proof that a website was penalised.
Short-term movement can settle after an update finishes.
What should your business do?
- Annotate the 18th to 21st August rollout window in GA4 and Google Search Console reporting.
- Compare query and landing-page changes with an earlier, comparable period.
- Separate ranking movement from tracking faults, website releases and demand changes.
- Review Google’s spam policies and content quality if losses persist after the rollout.
- Avoid large-scale content changes in response to one or two volatile days.
Loop Digital’s SEO services connect ranking analysis with technical, content and commercial priorities.
Google Expands Personalisation Across Search, Discover and News
Google has introduced new personalisation features across Search, Discover and Google News.
Publishers can now embed a Preferred Sources button that lets readers select their site and return to the page they were viewing. Preferred publications can appear more prominently for those readers in Top Stories and may be highlighted in AI Overviews and AI Mode.
What should publishers do?
- Consider adding the Preferred Sources button if the website publishes useful content consistently.
- Place it where engaged readers are likely to see it, such as near the end of an article.
- Continue building direct audience relationships rather than relying on personalisation alone.
Search Console Platform Properties Roll Out Globally
Google made Search Console platform properties globally available on 29th July. The feature shows how content posted on Instagram, TikTok, X and YouTube performs across Google Search, Discover and Google News.
Google also published a social and video performance guide covering trending queries, the 24-hour performance view, cross-platform comparison and opportunities to update or extend successful content.
What should marketing teams do?
- Add eligible social and video platform properties in Search Console.
- Compare the queries and formats generating visibility across each platform.
- Use the 24-hour view to identify emerging posts that may warrant further promotion.
- Record content changes with Search Console annotations where appropriate.
Google Expands Automatic Video Resizing in Performance Max
Google’s asset-optimisation guidance says landscape videos may be reformatted into square or vertical versions to reach more inventory.
Advertisers that do not want the feature applied to existing campaigns can use Google’s opt-out process by 4th September and can change the setting later.
What should advertisers do?
- Audit Performance Max campaigns with only landscape video assets.
- Upload approved square and vertical versions where brand control matters.
- Review automatically reformatted assets for cropping, claims, captions and product accuracy.
- Decide whether to use the opt-out before 4th September.
- Record a performance baseline before newly formatted assets begin serving.
Google Tests Channel Prioritisation Controls for Performance Max
The controls use positive or negative channel adjustments. A positive adjustment relaxes the acceptable CPA for that channel, while a negative adjustment tightens it. They do not assign a fixed percentage of campaign budget to each channel.
What should advertisers do?
- Do not plan account changes on the assumption that the controls are available.
- Review Performance Max channel reporting and how each channel supports the wider journey.
- Avoid reducing upper-funnel activity based on last-click results alone.
- Define the evidence required before applying a positive or negative adjustment.
- Monitor account eligibility and any broader release announcement from Google.
Google Adds New AI Max Testing and Planning Tools
Advertisers will be able to test budgets and return targets across multiple Search campaigns in one experiment, while retaining brand and location controls in AI Max tests.
Performance Planner will also model the effect of bidding and budget changes and allow suggested changes to be applied.
| Date | What changes |
| 3rd August 2026 | Creation of new campaign-level Broad Match and legacy Automatically Created Assets configurations stopped. |
| 1st – 30th September 2026 | Eligible Broad Match and ACA campaigns migrate automatically to AI Max. |
| September 2026 | In-account notices encouraging voluntary Dynamic Search Ads upgrades begin. |
| 15th January 2027 | Google sends reminders to advertisers still using legacy Dynamic Search Ads. |
| 1st – 28th February 2027 | Dynamic Search Ads migrate automatically and new DSA ad groups can no longer be created. |
| Around September 2027 | Older Google Ads API versions supporting the legacy entities are expected to reach their normal sunset. |
The campaign migration begins in September 2026. The approximate September 2027 date applies only to the expected retirement of older API versions. The new testing tools give advertisers a short opportunity to establish a reliable baseline before the first automatic changes land.
What should advertisers do?
- Identify campaigns that are suitable for a controlled AI Max experiment.
- Audit campaigns, scripts and API integrations that still depend on legacy configurations.
- Benchmark CPA, ROAS, lead quality and conversion volume before testing.
- Keep brand and location controls aligned with commercial requirements.
- Define success criteria and a minimum test period before launch.
- Treat planner recommendations as scenarios to review, not instructions to apply automatically.
Need a clearer view of where automation is helping or limiting your account? Loop Digital’s paid search and PPC services focus on lead quality, acquisition cost and measurable return.
ChatGPT Ads Expand Across 31 European Markets
The UK is already listed as available and is not one of the newly added markets. At the time of writing, OpenAI’s availability page still labels the 31 markets as ‘Coming Soon’, so advertisers should check live eligibility rather than assume access.
Initial access is being offered through selected agency, technology and Ads Solutions partners, with wider self-service access expected later in the summer. Ads continue to appear for Free and Go users, while Plus, Pro and Enterprise remain ad-free.
What should advertisers do?
- Check whether target markets and the relevant buying route are live before planning spend.
- Confirm that ChatGPT’s Free and Go audience fits the campaign objective.
- Set up conversion measurement before launching a test.
- Use a controlled budget and compare qualified leads or sales, not clicks alone.
- For cross-border campaigns, review language, landing pages, consent and market-specific requirements.
OpenAI Introduces ChatGPT for Teens
OpenAI introduced ChatGPT for Teens on 18th August, with stronger safety protections, learning features and optional parental controls for users aged 13 to 17.
People who state that they are within this age group, or whom OpenAI’s systems estimate are under 18, are placed into the experience automatically. It includes Study Mode, responsible homework reminders, quizzes, learning visualisations and configurable Study Hours.
What should relevant organisations consider?
- Review how AI tools are discussed in education, safeguarding and acceptable-use policies.
- Avoid assuming that every ChatGPT user receives the same product experience.
- Keep human oversight in place where AI is used to support young people.
ChatGPT Is the Sixth-Most-Clicked Destination From Google Search
An iPullRank study placed ChatGPT as the sixth-most-clicked destination from Google Search. It analysed 13.1 billion search events from 9.1 million opted-in users between October 2024 and December 2025.
ChatGPT also received a higher share of paid clicks than the other leading destinations examined. However, much of this activity is likely to include navigational searches from people using Google to reach ChatGPT, and the historical dataset should not be treated as a current traffic benchmark.
The wider finding is that traditional search and AI platforms are part of the same discovery journey. Users may move between them rather than replacing one with the other.
What should marketers do?
- Analyse journeys involving both traditional search and AI platforms.
- Separate navigational brand demand from broader discovery activity.
- Treat the study as market context rather than a benchmark for current performance.
Business Websites Lead Gemini Local Citations as Sources Shift
In one US local-search study, business websites accounted for 59.9% of citations in Gemini. The dataset covered 14,472 citations from 1,487 queries across 50 metropolitan areas and 10 local-service categories.
Separately, Promptwatch measured an 86.4% fall in Reddit’s share of ChatGPT citations over four days. The monitoring identified the change in timing, but it did not establish why it happened.
These are useful snapshots, not durable ranking rules. The local study covers US queries, results differ between AI platforms, and citation share does not automatically translate into traffic, enquiries or sales. The stronger lesson is to invest in reliable owned content while measuring visibility repeatedly across more than one system.
What should your business do?
- Make service, product and location pages clear enough to answer common selection questions.
- Keep business details, expertise and evidence consistent across owned and trusted third-party sources.
- Monitor a repeatable set of prompts across Gemini, ChatGPT and other relevant platforms.
- Record citation sources and changes over time instead of relying on one-off searches.
- Connect AI visibility with referral traffic, enquiries and revenue where measurement allows.
Loop Digital’s Answer Engine Optimisation services focus on improving the clarity, authority and usefulness of information that answer-led systems may draw on, without promising citations or rankings.
AI Overview Traffic Can Be Misclassified in Analytics
A case study of one transport brand found that some visits identified as AI Overview traffic appeared as Direct in GA4. The analysis covered 51,200 tracked events and 1,661 AI Overview snippets between September 2025 and June 2026.
The study estimated that 22.4% of traffic was misattributed to Direct. However, the finding should be treated cautiously: its tracking method used URL fragments, which may also follow clicks from Featured Snippets and People Also Ask, while the analysis mixed event-level and session-level data.
What should your business do?
- Review unexplained changes in Direct traffic alongside organic-search trends.
- Explore a custom GA4 dimension for text fragments only if the team can document its limitations.
- Compare GA4 findings with Search Console and tracked AI Overview appearances.
- Treat attribution estimates as directional rather than exact.
- Do not use the study’s percentages as a performance benchmark for your website.
Google Pauses AI-Generated Images Within AI Overviews
Publishers raised concerns about accuracy, attribution and generated visuals appearing alongside citations to original recipe content. Google confirmed that the experiment was no longer running and distinguished it from the separate feature that creates images when a user explicitly requests one.
What should publishers do?
- Keep dated records of how important image-led searches appear.
- Use original, attributable images with accurate surrounding information.
- Monitor future experiments before making large changes to visual-content strategy.
Key Takeaways
- Elementor Pro 4.2.1 and earlier has a critical file-upload vulnerability; version 4.2.2 contains the patch.
- Google’s August spam update finished on 21st August after a global, all-language rollout.
- Google released new personalisation controls across Search, Discover and News, including an embeddable Preferred Sources button.
- Search Console platform properties now report how social and video content appears across Google surfaces.
- Performance Max can generate missing video formats from existing assets, subject to advertiser settings.
- Google is testing, but has not broadly released, Performance Max channel prioritisation controls.
- New AI Max tools will support multi-campaign testing and planning as legacy Search campaign migrations begin in September.
- ChatGPT Ads is moving into 31 additional European markets, with access still being phased in.
- ChatGPT for Teens introduces a protected experience for users aged 13 to 17.
- One large historical dataset placed ChatGPT as the sixth-most-clicked destination from Google Search.
- One US study found business websites led Gemini local citations, while separate monitoring showed Reddit citations falling sharply in ChatGPT.
- A single-brand case study found that some AI Overview traffic may be misclassified in GA4.
- Google ended a small experiment that inserted unrequested AI-generated images into recipe AI Overviews.
Looking for your next opportunity?
Digital marketing careers
We’re always on the lookout for talented individuals to join our ever growing team. If you think you’d be a great match for Loop Digital, we’d love to hear from you.
Join 300+ business owners getting weekly growth strategies - subscribe now.
"*" indicates required fields