I’ve been watching the AI landscape closely this week, and there’s been a flurry of activity that’s worth paying attention to. Google’s launched Gemini 3, Microsoft’s rebuilding Windows around AI agents, and there’s some interesting insights from brands about how they’re navigating AI-powered search. For those of us running e-commerce businesses and paid campaigns, this isn’t just tech industry noise – these developments are going to shape how your ads perform and how customers find your products.
Let me walk you through what’s actually happening and what it means for your business going forward.
Google Launches Gemini 3: The AI Arms Race Heats Up
Google unveiled Gemini 3 this week, claiming it’s now leading the pack in mathematics, science, and what they’re calling “agentic AI” benchmarks. Now, you might be wondering why you should care about Google’s latest AI model when you’re just trying to sell products profitably online.
Here’s the thing: Gemini isn’t just powering Google’s search engine – it’s increasingly the brain behind your Google Ads campaigns. Remember when Performance Max campaigns first rolled out and felt a bit like handing over the keys to a teenager? That was early-stage Gemini at work. Each iteration of these AI models directly affects how your campaigns optimise, how your product feeds get interpreted, and how Google matches your ads to search intent.
What does this actually mean for you? In my experience with accounts, when Google upgrades their underlying AI, we typically see a settling-in period where campaigns need to relearn patterns. If you’re running Performance Max or Smart Shopping campaigns, don’t be surprised if you see some fluctuation in the coming weeks as this new model rolls out into the advertising platform. It’s not your campaigns breaking – it’s Google’s AI getting smarter, albeit with some growing pains.
The multimodal capabilities Google’s bragging about – that’s their AI’s ability to understand images, text, and video together – matters particularly for product advertising. Google’s getting better at understanding what your products actually are, which should mean better matching to relevant searches. Should being the operative word.
Microsoft’s Building an AI Agent Army (And Why That Affects Your Advertising)
Microsoft’s been busy too. They’ve announced Agent 365 and they’re literally remaking Windows to accommodate autonomous AI agents. This might seem completely disconnected from your Google Ads account, but bear with me.
What we’re seeing here is Microsoft’s vision for how businesses will operate going forward: AI agents handling routine tasks, making decisions, and taking actions without constant human oversight. They’ve also forged an alliance with NVIDIA and Anthropic to build out the infrastructure for this.
Why does this matter to you? Because Microsoft also owns Bing, and Bing Ads (sorry, “Microsoft Advertising”) is increasingly worth paying attention to. With lower competition than Google and Microsoft pushing hard into AI-powered search, the platform’s evolving rapidly. I’m interested to see how Microsoft’s AI capabilities translate into their advertising platform over the next six months.
More broadly, this AI agent trend is something I’m watching closely because it changes how people search and shop. If AI agents start handling purchasing decisions or product research on behalf of users, the way your ads need to communicate will shift. We’re not there yet, but the groundwork’s being laid.
Real-World Insight: How Scotts Miracle-Gro Is Navigating AI Search
This week brought some refreshingly practical insights from Matt Taylor at Scotts Miracle-Gro about how they’re actually dealing with AI search and retail media in the real world. This is the stuff that matters – not what the platforms claim will happen, but what brands are experiencing on the ground.
What caught my attention was their focus on how AI search is changing the customer journey. People aren’t just typing “lawn fertiliser” anymore – they’re asking questions like “how do I fix brown patches in my lawn?” and expecting complete answers. AI-powered search results are increasingly giving those answers directly, which means your product needs to be part of that answer, not just waiting at the end of a traditional search results page.
For e-commerce businesses, this is crucial. Your product descriptions, your ad copy, and your landing pages need to answer questions, not just list features. The old approach of cramming keywords into product titles? That’s becoming less effective as AI gets better at understanding context and intent.
The retail media piece is important too. If you’re selling through Amazon, Tesco, or other retail platforms, their AI-powered advertising systems are evolving rapidly. What worked last quarter might not work this quarter because the underlying technology keeps shifting.
Grok 4.1 Launches: Lower Hallucination Rates Matter More Than You Think
Elon Musk’s xAI launched Grok 4.1 this week, claiming lower hallucination rates. “Hallucination” is tech-speak for when AI makes stuff up – confidently stating things that aren’t true.
You might think this doesn’t affect your advertising, but it does. As AI becomes more integrated into how people search and shop, the accuracy of these models matters enormously. If someone’s using an AI assistant to research products and that assistant hallucinates features your product doesn’t have (or claims your competitor offers something they don’t), that affects the customer journey before they ever see your ad.
Plus, if you’re using any AI tools for writing ad copy or product descriptions – and many businesses are – the reliability of these models directly impacts your advertising content. Lower hallucination rates mean you can trust the output more, but you still need to check everything. I’ve seen too many product descriptions that clearly came from AI and clearly weren’t reviewed by a human.
What This All Means Going Forward
Here’s what I’m taking away from this week’s developments and what I’m advising businesses to focus on:
Expect more volatility in AI-powered campaigns. As Google and Microsoft upgrade their AI models, your Performance Max, Smart Shopping, and automated campaigns will go through adjustment periods. This is normal. Don’t panic and make dramatic changes during these periods – give the algorithms time to restabilise.
Question-based content is becoming crucial. Your product pages, landing pages, and ad copy need to answer questions, not just describe products. Think about what customers actually ask, not just what keywords they type.
Multi-platform strategy matters more. With Microsoft pushing hard into AI and building serious infrastructure, keeping some budget on Microsoft Advertising alongside Google is worth considering. Competition is lower, and their AI capabilities are developing fast.
Don’t trust AI blindly. Whether it’s Google’s campaign automation or AI writing tools you’re using for content, review everything. These systems are getting better, but they still make mistakes – and those mistakes can cost you real money.
The AI race is accelerating, and the advertising platforms we all rely on are right in the middle of it. Things are going to keep changing, probably at an increasing pace. That’s both a challenge and an opportunity – businesses that stay informed and adapt quickly will have an advantage over those who stick their head in the sand and hope their campaigns keep working the same way they did last year.
I’ll be keeping tabs on how these AI developments translate into actual advertising platform changes. Because that’s where it matters – not in the benchmark tests and press releases, but in whether your campaigns drive profitable sales for your business.