Key Takeaways:
- With little warning, Google’s global rollout of “AI Overviews” in search, and the integration of ads into those generated results, accelerates deep questions about how brands will be built and how products will be promoted and discovered in a new era of the web.
- A parody ad for a fictitious fast fashion brand, aired on Saturday Night Live, prods at a raw nerve by reminding us that wilful consumer ignorance contributes as much to the negative impacts of fast fashion as the brands’ material choices, business models, and purchasing practices.
Google’s AI Overview rolls out, quickly gets ads, and challenges sellers to react to a new era of the internet.
Following hot on the heels of Google I/O (which we profiled last week) the annual Google Marketing Live conference took place this week, and the Google Ad team announced many new AI formats, new ad tools, and changes to the Google Ads platform that are coming soon.
Despite only recently rolling out the feature itself to a wider audience, Google has moved quickly to showcase ads directly in its AI Overviews, which are generative summaries that appear at the top of search results – and that can feature text, images, products, recipes, businesses, maps, and more, with links to original sources.
As The Interline has written about before, the impact of these changes to search are being downplayed and under-analysed when we consider just how significant of a transformation they represent to how people engage with the internet – and especially with eCommerce.
For more than two decades, finding content and products through the internet has meant being served a list of links and previews (even though the way those links are prioritised has been extremely mercurial) and then having to actively click or tap those links to visit websites and online retail frontends to find out more.
Now, very suddenly, that paradigm has flipped, and the default behaviour for a Google search will be to place an AI summary of unclear provenance and opaque weighting ahead of any links. This, readers of The Interline, is not a small change to a single Google product, but rather a massive overhaul to how the internet operates – and it is destined to have deep implications for companies that sell through the internet as even a secondary channel.
Consider this: do you know, right now, if your eCommerce frontend is optimised for AI scraping? Do you know how Google is determining what channels, what content, and what products merit inclusion in an AI overview? Does anyone outside Google have visibility into those criteria?
And this is just the first-stage consideration. Beyond the question of how, when, and why products earn an AI recommendation is the adjacent question of what mechanisms influence where advertisements are now shown next to AI overviews. And this, too, is a change that Google is seeking to downplay, with a short explainer video that offers very little clarity.
According to Google, advertisers that already run certain campaigns through Google will automatically become eligible to appear in AI Overviews, but the search giant will “continue to test and learn new formats, getting feedback from advertisers and the industry”.
Similarly unclear today is how Google’s own shopping platform will interact with AI Overviews. Like Google itself, which is a key referrer of traffic for many online brands, Google Shopping is an important source of product discovery – hence why it was designated a gatekeeper under the Digital Markets Act (DMA) by the European Commission in late 2023.
The DMA aims to “ensure fair and open digital markets” by applying a new set of pro-competition rules on how designated gatekeepers can operate their “core platform services”. Others on this list include Alphabet, Amazon, Apple, ByteDance, Meta, and Microsoft.
In all likelihood, this spells the end of the SEO era rather than a natural evolution of it. Brands and content creators are already working to optimise their content for what feels, organically, like the right structure (questions and answers) but ultimately this new approach represents a complete black box, and brands and advertisers are guessing blind as to what conditions need to be fulfilled for their content and products to earn a prominent position in the AI-first age of the internet.
Of course this will also entail opening up all their content for AI scraping; banning Google’s bots (or OpenAI’s, or others) from crawling content that isn’t already part of these models’ training sets will also prevent those pages from being surfaced as part of AI Overviews.
And this is all without addressing the fact that AI Overviews, at least in their launch state, are often deeply inaccurate – providing wrong answers, dangerous guidance, and a host of other undesirable outcomes. What this means for public liability is one thing, and what it means for brand protection is quite another.
What’s clear, though, is that we have quietly and over-confidently entered a new stage of the internet’s evolution – one that fashion brands cannot reasonably be expected to be ready for when the company insisting on taking us “forward” clearly isn’t ready either.
SNL commentary on fast fashion highlights consumer behaviour as a central concern
A new investigation by Swiss advocacy group Public Eye has followed up on its 2021 report, which found a number of staff across six sites at e-tailer Shein in Guangzhou were doing excessive overtime. That is, more than 75 hours a week with one or fewer days off. In a response to the report, Shein said long working hours in the sector were a “common challenge that brands, manufacturers, and other ecosystem players must work together to address.”
Coincidentally, also this week, Saturday Night Live (SNL) aired a skit that poked fun at fast fashion giants, but also raised critical questions about forced labour, worker pay and conditions, and product quality. The most significant moment comes when a customer character asks, “Is this shady?” and the narrator responds, “If it was, would you stop buying it?” All the customers reply: “No.”
And therein lies one of the biggest dead-ends when it comes to any proposal to rein in overproduction: overconsumption is the trigger, not just the symptom.
Of course consumer behaviour does not evolve in a vacuum, and current prevailing attitudes around “not worrying about it” are heavily shaped by the marketing practices of fast fashion brands. For the demographics that these brands primarily target, exposure to fast fashion marketing is an unavoidable element of social media usage, and is particularly good at convincing the purchaser that fast fashion is not only convenient but also harmless. The clothes are easy and quick to purchase, with short delivery times adding to the appeal. This hassle-free shopping experience creates the illusion that the clothes just about appear out of nowhere, allowing consumers to avoid thinking about who made them or what will happen to them once they are discarded.
This marketing also plays on human aspirations and insecurities, something that affects younger consumers in particular. With fashion lifecycles now shorter than ever and microtrends constantly popping up, many young consumers fear missing out and struggle to stay current with the latest trends. This urgency drives them to quickly purchase items from fast fashion stores, fearing that these items could disappear from shelves almost instantly.
And this is before we thread the two strands of this week’s news together and consider how this might evolve when an AI Overview begins surfacing new trends itself… and potentially repeating misinformation about how products are made and by whom, with the confident authority that’s become typical of generative responses.
Although it’s been said many times, consumers need to demand better quality – not just in the products being bought, but in the lives of the people making them. And this is for consumers’ own benefit too; many unknowingly purchase toxic products or products made under horrific labour conditions. Fast fashion items may seem exciting and the price tag too good to turn down, but consumers will ultimately end up spending more money on low-quality items that need to be replaced frequently than better-made, timeless products.
And if consumers start to switch to the mindset of treating themselves to the very best that they can, they will feel better about investing in and wearing high-quality clothing. Or if not clothing with a higher price tag, at least investing time in finding quality pieces from vintage or thrift stores. Everyone, including – especially, actually – those working in fast fashion’s supply chain, deserves better.
In short, transparency matters in the AI era more than ever. Not just in terms of visibility into the supply chain, but into how language models are creating their answers (and from what sources), lest we land ourselves in a thick soup where the truth is buried many geographical, contractual, and software layers deep.
