AuthorTom Gandhi
LinkedIn
Our Head of SEO, Tom Gandhi, is collating his monthly reading each and every month around SEO, and eCommerce industry updates and changes. Keep up to date with the latest news, and hear Tom’s take on how it might impact your brand in the future.
New LLM’s are struggling to differentiate from competitors due to limited access to unique data.
Writer’s Take:
There are limited avenues for LLM’s to scrape data, especially unique data, which means that other than interface, speed and slight improvements in relevance of response, most LLM’s do the same as one another, no matter who built them (GPT, Anthropic, Perplexity etc).
What is new and unique is Perplexity's collaboration with Shopify. As the first LLM to product-tise the LLM, this is huge. Not only does Perplexity allow the use of pretty much all major LLMs (Anthropic, GPT, Gemini), it also connects directly with signed up Shopify stores which means, UNIQUE DATA NO ONE ELSE (other than Google) has. This means its ahead of the curve of other LLM’s, and ahead of AI Overviews too - a lot of work still to be done to improve relevance and accuracy of products displayed, but a true front-runner in the commercial LLM landscape.
Useful related reading;
Perplexity announcement source
Perplexity’s AI shopping agent
OpenAI’s early mover advantage is diminishing, and quickly.
"Among closed-source models, OpenAI’s early mover advantage has eroded somewhat, with enterprise market share dropping from 50% to 34%. The primary beneficiary has been Anthropic,* which doubled its enterprise presence from 12% to 24% as some enterprises switched from GPT-4 to Claude 3.5 Sonnet - (those that I have spoken with will also know I have pretty much completely moved over to Claude now too!) - when the new model became state-of-the-art. When moving to a new LLM, organizations most commonly cite security and safety considerations (46%), price (44%), performance (42%), and expanded capabilities (41%) as motivations."
Writers Take:
Improvements in competitor LLM’s means GPT has eroded over time and become not only the only player in space, but unable to deliver the same quality of relevance / output as others. An example of this comes in the source references, it took some time to catch up with Claude (by Anthropic) who leveraged source references from the get go, just 1 example of GPT trailing other services.
Brand awareness converts to higher conversion (2.6x times higher than low brand awareness brands) - emphasis on LOYALTY!
Writers Take:
Loyalty has always been a theory, but not something traditionally executed in the D2C environment, well not until 2023/24 that is. Loyalty has since boomed, and become a significant player in the way brands can establish themselves, and in their pursuit of growth. Take Represent for example, a streetwear brand that started on hype (still does to some degree), but as it’s become more saturated e.g. more people can get access to the product devaluing its once prominent scarcity, the founder established himself as THE BRAND, George Heaton is Represent, his brother? He’s a co-owner but you don’t see nearly as much from him. Why? George is the brand, he drives the brand, he powers the brand through his commitment to community and to their engaged audience. Run 247 is why the brand continues to be so successful, as well as their core product offering, run clubs have exploded in 2024, they’ve tapped into that, and that has catapulted them further into growth.
Therefore, a brand with limited brand awareness e.g. loyalty / customer buy-in clearly will impact their ability to drive performance in 2024 and onwards, you need to focus on community, loyalty and building a truly valuable product (unless you’re a SHEIN or TEMU where price just can’t be challenged and loyalty means nothing), you won’t be able to convert nearly as well as your established counterparts.
Will AI eat the traditional browser?
“He wants to transform the browser from a mere viewer to an operating system-like entity that maintains personal preferences and behaviors at the system level, allowing us to use “AI” across devices without replicating our choices at the app level. His new browser-based OS will understand user context and preferences at a fundamental level, making it easier to create personalized experiences. Rather than having applications dictate how we interact with information, our usage patterns and preferences will shape how information and services are presented to us.”
Writers Take:
I don’t see how AI can completely take over the human experience when browsing, delivering results so accurately that they are able to essentially make decisions for us based on previous buying habits, preferences and choices. Can AI be better at decision-making than a human?
Humans are innately lazy, so if this can be perfected, I could see it being a success, however it breaks the entirety of the buying journey where choice & selection are paramount to humans abilities to decide on a product, not to mention actually trying clothes on before buying them - though with the explosion of D2C eComm stores without brick and mortar, clearly this isn’t such an issue.
Google has released a quick, more accurate version of their Gemini AI which powers AI Overviews, rumoured to answer questions much more accurately than before.
Want to understand how these changes are likely to impact your business? Contact Tom Gandhi today at tom.gandhi@thisisnovos.com to consult on impact, and how to address.