For the past couple of years, the prevailing wisdom has been simple. Let the machines do the work. Let automated bidding decide where your budget goes, let AI search decide which brands to recommend, and let your analytics tell you what’s working.
This week showed why that only works if somebody checks.
Researchers invented a concept, posted about it on a handful of popular sites, and four of the biggest AI search tools were citing it within seven days. A sudden collapse in ChatGPT traffic sent marketers scrambling, until it turned out to be a tracking bug. And Google, after years of asking advertisers to trust automation, is giving back control over where budgets are spent.
The thesis running through all of it: the machines are making confident decisions on information nobody has checked, so the advantage now goes to marketers who question their data, use the controls they’re given and take charge of what the machines read about them.
Here is what mattered, why it mattered, and what to do with it on Monday.
Board Analysis, The Digital Maze’s weekly marketing briefing, goes deeper on all of this every Friday. Subscribe here.
The Big Story: Google Ads Gives You Back Control Of Where Your Budget Goes
Google Ads is introducing two new controls, Channel Prioritization and Budget Allocation, which “will roll out globally across Search, Shopping, Performance Max, and Demand Gen campaigns over the coming months”, according to Search Engine Journal. Advertisers will be able to mark channels as “High priority” or “Low priority”, and “Google will fund high-priority channels before moving any remaining budget to lower-priority inventory, while still considering Target CPA or Target ROAS goals.” Google also confirmed that 0% will be an option, which means you can stop spending on a particular channel or channel grouping, “including within Performance Max.”
For years, the message from Google has been to trust the automation. This is the first time in a long while it has handed a meaningful lever back. There are limits. In Performance Max and Demand Gen, Google-owned channels such as Search, YouTube, Discover, Gmail and Maps stay grouped together, and Google says it will keep to your caps on a “best-effort basis”. But for any marketing leader who has struggled to explain to a finance director where automated spend actually goes, this is a real step forward, provided someone on your team actually uses it.
Why this matters:
- Every channel defaults to “High priority”, so nothing changes until your team sets priorities and caps.
- The 0% option lets you switch off spend on a channel grouping, even inside Performance Max.
- From February 2027, the existing Display allocation in Search campaigns is retired. Google recommends setting Display to “Low priority” with a 50% allocation to keep your current setup.
Read more: Google Ads Introduces Channel Prioritization And Budget Allocation, Search Engine Journal, 14 October 2026
Search & Intelligence: AI Search Believes What It Reads. Make Sure It’s Reading You.
A paper from researchers at Fudan University found that “ChatGPT, Perplexity, Grok, and Google AI each cited posts about a made-up concept within seven days of the posts going up”, as reported by Search Engine Journal. The researchers posted about the fictional “Cactus Gravity Protocol” on sites the tools already cited often. Grok “linked to a post in 30 of 32 answers, Google AI in 15 of 33, Perplexity in 13 of 35, and ChatGPT in 7 of 20.” Meanwhile, “40 postings on a self-hosted WordPress site received zero references within seven days.”
This is the thesis in miniature. AI search tools aren’t weighing up truth; they’re trusting the places they already trust. In paired tests, “the post on the frequently referenced source produced the first reference/citation or planted detail” every time. The tests didn’t involve real brands, and the researchers ran them before Google clarified that its spam policies cover attempts to manipulate AI responses, so this isn’t a playbook for gaming the system. It is a clear signal that where your brand is talked about matters as much as what you publish on your own site.
Why this matters:
- Placement beat volume: posts on frequently cited sites were always picked up first.
- 40 posts on a self-hosted WordPress site got zero references in a week, so your own blog alone isn’t enough.
- Reddit was the most-cited site for Google AI and Grok, YouTube for Perplexity and Wikipedia for ChatGPT, so know which sources your audience’s AI tools favour.
Read more: AI Search Tools Cited A Made-Up Concept Within A Week, Paper Finds, Search Engine Journal, 19 October 2026
Also worth your time this week:
Paid Media & Performance: UK Budgets Are Growing. Spend The Extra Wisely.
UK businesses revised their marketing budgets upwards to the highest level in more than two years, according to the latest IPA Bellwether report, covered by The Media Leader. “A net balance of +9.8% of panellists made upward revisions to their main media budgets”, a nine-year high and up from +1.5% the previous quarter. Main media “was led by Other Online (+17.2%) and Video (+9.8%)”, and for the first time since tracking began in Q1 2020, all five main media subcategories registered upward revisions.
More money is good news. More money spent on autopilot is not. With Google handing back channel controls in the same week, UK marketers have both the budget and the tools to direct spend more deliberately. One detail stands out: “PR budgets have been revised upward in 13 consecutive quarters”, with the category likely benefiting from its role in supporting generative engine optimisation. In other words, businesses are already paying to get their brands into the sources AI tools trust, which is exactly what this week’s Search research suggests they should do.
Why this matters:
- The main media net balance of +9.8% is a nine-year high, so competition for attention is rising.
- Other Online led with +17.2%, so expect busier, pricier digital auctions into the golden quarter.
- PR budgets have grown for 13 consecutive quarters, likely helped by GEO, so earned coverage is now a search investment.
Read more: Main media budget increases rise to highest level in nine years, Bellwether finds, The Media Leader, 15 October 2026
Also worth your time this week:
Web Development & Design: Meta Changed Its Words. It Didn’t Change Its Crawler.
“Meta has removed training from the crawling clause in its new Commercial Terms, which take effect Oct. 30”, reports Search Engine Journal, but “the permission to crawl websites associated with a business’s products or services is still in the agreement.” Because the terms apply to any business use of Meta’s products, “they also cover companies that manage a Facebook Page without buying ads.” The UK’s News Media Association condemned the original wording, and as of 19 October the UK English preview of Meta’s Self-Serve Ad Terms still contained the earlier “may use any content” language.
This is the website side of the same story. Machines are reading your site, and the rules about what they can do with it are being rewritten in real time. Legal experts quoted in the piece think the revised wording is still broad, with one calling it “still broad enough to drive a truck through.” The practical point is that Meta’s crawler documentation lists separate crawlers for AI training and Meta AI search, so you can make a deliberate choice about each in robots.txt rather than leaving it to default.
Why this matters:
- The terms take effect on 30 October and cover Facebook Page owners, not just advertisers.
- The new language doesn’t mention robots.txt, so check the clause again before it takes effect.
- Meta runs separate training and search crawlers, so review how your robots.txt handles each one.
Read more: Meta Drops Training From Its Crawl Clause And Keeps The Crawl, Search Engine Journal, 19 October 2026
Also worth your time this week:
Brand & Connection: Fender And Marshall Are Writing Their Own Story For AI
“Brands can’t tell ChatGPT how to talk about their products”, writes Digiday. What they can do is give the models more to work with. Fender’s interim CMO Kelly Magrath told Digiday the company had overhauled its website, aiming to turn it into a “front door that we can use as a storytelling platform”, replacing a “transactional” shopfront with consistent brand messaging. Marshall is running a YouTube series, ‘Dead Venues’, which stages gigs by new bands in long-extinct venues, as part of its push into headphones. Marshall’s Christine Kennedy Grant described AI Overviews and LLM tools as “such an important place for us to show up”.
If AI tools trust what they read, the smartest brands are making sure there is plenty of distinctive, credible content to read. These aren’t SEO tricks. They’re brand-building projects that happen to feed AI search as well. As Digiday puts it, a brand’s own content “may be only one of several sources LLMs typically draw from when composing answers. But it’s one a marketer can control.”
Why this matters:
- Fender rebuilt its website as a storytelling platform, not just a shop, to shape how it’s described.
- Marshall’s ‘Dead Venues’ series shows brand content can support new categories and AI visibility at once.
- 75% of marketers plan to increase spending on visibility in AI assistants next year, per a Kantar survey, so the competition for AI attention is about to intensify.
Read more: Brands turn to content series and Jimi Hendrix to improve AI search performance, Digiday, 15 October 2026
Also worth your time this week:
AI & Automation: Your ChatGPT Traffic Didn’t Fall. Your Tracking Did.
“The number of ChatGPT-attributed sessions in Google Analytics 4 fell by 22% to 57%” for sites tracked by Graphite with meaningful AI Assistant traffic, starting 9 October, reports Search Engine Journal. The cause appears to be links in ChatGPT’s answers opening “without the utm_source=chatgpt.com tag or a referrer”, so the visits showed up as Direct. Adobe’s Cedric Huesler found ChatGPT visits to around 500 pages “fell by about 31,000 a day while Direct visits rose by about 26,000 a day.” Huesler says OpenAI confirmed to him on 14 October that it was a bug, and that a fix began rolling out early on 15 October.
This is the clearest example yet of this week’s thesis. Plenty of teams would have looked at a near-halving of ChatGPT traffic and drawn the wrong conclusion, about AI search, about content, or about budgets. The data was confident. It was also wrong. Both Graphite’s Ethan Smith and Huesler suggest annotating the October dates in your reporting “prior to anyone interpreting it as a loss of traffic”, with Huesler now putting the window at 8 to 15 October.
Why this matters:
- ChatGPT-attributed GA4 sessions fell by 22% to 57% from 9 October, so your October AI search reports are likely understated.
- 95% of mobile ChatGPT visits to Graphite’s sites relied on the tag alone, so app traffic was hit hardest.
- Annotate 8 to 15 October now, and compare ChatGPT and Direct traffic side by side for the affected pages.
Read more: ChatGPT Traffic Drop Appears To Be A Tracking Problem, Search Engine Journal, 15 October 2026
Also worth your time this week:
Case Study of the Week: +403% Organic Revenue After Getting Back On Track
Boston Seeds is a leading grass and wildflower seed retailer that had worked with us before. When it paused its SEO investment, organic performance and revenue began to decline. In 2024, Boston Seeds came back on board, and our priority was clear: reverse the downward trend and build sustainable, long-term growth.
We combined SEO, including keyword research, on-page optimisation, valuable content and high-quality backlinks, with a content strategy across product pages, product guides and category areas. Continuous conversion rate optimisation made sure the extra traffic turned into sales.
- +403% increase in organic revenue (YoY)
- +350% increase in organic traffic (YoY)
It’s the same lesson as this week’s news. Visibility isn’t something you switch off and pick back up later. When Boston Seeds paused, performance slipped. When it invested consistently in content that search engines trust, revenue followed. As George Wallis of Boston Seeds puts it, “our rankings are the highest they’ve ever been, against the backdrop of ever-increasing competition.”
Read the full Boston Seeds case study.
Bringing It Together
Every story this week comes back to the same question: who is checking? AI search tools cited a concept that didn’t exist because it appeared on sites they already trusted. GA4 reported a ChatGPT traffic collapse that was really a missing tag. Meta rewrote its crawl terms but kept the crawl. And Google, after years of asking for blind trust in automation, handed back a set of budget controls that do nothing unless someone switches them on.
UK marketing budgets are rising at their strongest rate in more than two years, so the stakes are going up. The brands that win from here won’t be the ones with the most automation. They’ll be the ones that question their data, use every control on offer and make sure the content machines read about them is accurate, distinctive and in the right places.
Three Takeaways for the Week Ahead
- Check your measurement before you change your strategy. ChatGPT-attributed GA4 sessions fell by 22% to 57% from 9 October because of a tracking bug, not lost demand. Annotate 8 to 15 October in your reports and compare ChatGPT and Direct traffic for the pages ChatGPT sends visitors to before anyone touches an AI search budget.
- Take back the controls Google is handing you. Channel Prioritization and Budget Allocation let you cap, or set to 0%, spend on lower-priority channels, but every channel defaults to “High priority”. Plan your priorities now, and if you run Search with Display enabled, put the February 2027 change in the diary.
- Feed AI the truth, in the places it actually looks. AI tools picked up posts on frequently cited sites first, while 40 posts on a self-hosted WordPress site got nothing. Invest in distinctive owned content, earn coverage in the sources AI tools trust, keep your structured data in line with your product feed and review how your robots.txt handles each Meta crawler before 30 October.
Want help putting any of this into action?
Whether it’s fixing your AI search tracking, taking control of automated ad spend or making sure your brand shows up accurately wherever AI looks, our team can help you turn this week’s changes into a plan. We combine SEO, PPC and web development so your marketing works as one joined-up strategy, with every decision tied back to revenue.
Book a call with The Digital Maze.
Want this in your inbox every Friday? Board Analysis is our weekly briefing for UK marketing leaders: genuinely useful marketing updates, delivered weekly.
