Interviews, insight & analysis on digital media & marketing

The last Google search…

By Alex Hamilton, Copywriter at  isobel 

The internet is changing. 

For decades, Google Search has reigned as the go-to tool for accessing information on the internet. But with the recent explosion in AI Large Language Models, LLMs, is that all about to change?

News, people, elections, brands — all change. 

This is hugely significant. Google searches have always held immense control over the narrative. What links make it to the first page, and in what order is enough to sway opinions of newsworthiness, people, elections, and brands? Bigger still is thinking about the stuff that doesn’t make it to the first page or the second, or even the third. The filtering of information and resources at Google sort of makes Google a modern-day God.  

Well, this God’s days are numbered. There’s going to be a paradigm shift in how people interact with the internet. Instead of typing keywords into a search engine, people may find themselves just chatting to AI assistants like ChatGPT. 

Rather than getting the famous page of 10 links back, they’ll get personalised, contextualised answers. We might at last even be able to get the exact answer we were looking for. 

This poses a significant challenge for brands who are used to reaching their target audiences through sponsored ads on Google search. LLMs will allow people to bypass traditional ad placements, providing them with direct, unbiased information, potentially rendering traditional sponsored ads less effective. 

Adapting to the new world.

Brands will need to adapt their advertising strategies to integrate with LLMs or find alternative ways to engage with users. Of the current old guard, META is the biggest to gain from this, siphoning from Google’s market share by maintaining brand presence opportunities on social media. 

So, what does this mean for Google-reliant brands? As the landscape changes, the billboard that points to your business should probably move location. Finding more appropriate touch points like social media platforms is something brands can do today. 

In the future, they may want to keep an eye on the race to get the best ai assistant. It’s very likely there will be new ways to market within the chat-based environment. 

If anything, this is just a return to the old days, when you’d ask your mate for restaurant recommendations, and they’d list out a few of their favourites. 

But where’s this all going? 

So, will Google search disappear? I doubt it. People still use Yahoo after all. Heck, people still ride horses. Though it’s an interesting thing to think about. If Google search did cease to exist, what would be the final search? “How do I turn my AI assistant off?” or something inane like “ChatGPT”. 

In the end, if no one on the internet is searching anymore, there could be a decrease in forum activity, blog posting, and website-stored information in general. Oddly, when you break it down, a lot of the information available on the web turns out to be people looking for information. 

Take Mumsnet for example, the world’s most popular parenting website. New parents flock to it on mass every day, usually redirected by Google – searching up parenting issues and product advice from parents all over the world. But what if people start asking questions directly to their AI assistant? The assistant will give them a straight answer and presumably, they’ll forget all about Mumsnet. 

This is a particularly potent issue with forums like Stack Overflow, where developers can get help for the latest software-related problems. In a world where they just ask AI assistants for instant answers, why would they spend hours searching on forums? Or even searching on Google in the first place. And won’t the content of those forums diminish if people stop asking questions on them? 

And that brings us to the crux. If the training data for the LLMs is Google, then what happens if the data on Google goes stale? 

Can LLMs train other LLMS? Is there a conflict with information theory here? The old, “New information can’t come out of old information” rule. 

Whatever happens next, the internet is about to change in a big way. 

People and brands should be ready for it.