AI is killing Google search: is your business next?
Google search is on life support, facing an AI-powered armageddon that’s forcing Australian businesses to completely rewrite their online growth strategies. But all is not lost.
Business
Don't miss out on the headlines from Business. Followed categories will be added to My News.
Google search is on life support and Australian companies are bracing for artificial intelligence armageddon.
Likening the collapse of online search to doomsday may seem extreme. But at the Winx Stand in Randwick this week, there was a funeral of sorts.
An image of the book: Inbound Marketing: Attract, Engage, and Delight Customers Online, was shown being ceremoniously burned on a barbecue as Google search traffic dives as much as 70 per cent thanks to AI.
HubSpot founders Brian Halligan and Dharmesh Shah wrote the book, considered the bible for companies seeking online growth, in 2014. The pair pioneered search engine optimisation – the art of drawing customer traffic from giants like Google – and built a marketing software empire now worth $US28.71bn.
But HubSpot chief executive Yamini Rangan tells The Australian that AI isn’t just changing the rules; it’s completely “rewriting the playbook”.
Chatbots are replacing Google searches. People can get information directly from AI tools like ChatGPT in a heartbeat and in “zero clicks”. As a result, traffic that marketers and companies have relied on for the past 20 years is plummeting.
Google, which controls about 94 per cent of Australia’s search market, has even disrupted itself. It launched AI Overviews late last year in an effort to compete directly with ChatGPT. These AI-generated summaries appear at the top of Google searches transforming the way people search and find information, negating the need to click on links to other websites.
How AI overviews kill search
“Almost everything in the way you go to market has changed with AI,” Ms Rangan said.
“Ten years ago, it was a lot about searching for information. Your customer would search and they would land in maybe the top 10 blue links within Google, and we did everything possible to show up. Then your customers would click on it and come to your website.
“That is getting completely disrupted. In fact, Google has announced their AI overviews have reduced search traffic by 20 to 70 per cent. That disruption is going to continue.”
Then there is the question of trust. Machines continue to struggle with what’s true or false. For example, Google’s AI summary said Christine Holgate was the chief executive of Linfox, when she in fact helms rival logistics company Team Global Express.
Other AI summaries can be more harmful, like suggesting users “eat at least one more small rock per day”. This was based on a satirical article but AI isn’t very good at understanding irony or hyperbole.
Website traffic in decline
Still, more people are relying on AI overviews in the search for information rather than clicking on links, sparking what has been branded an “AI armageddon” for news publishers. Traffic from HuffPost’s desktop and mobile sites halved in the past three years, according to Similarweb. The Washington Post, owned by Amazon founder Jeff Bezos, has experienced a similar decline.
Business Insider cut more than 20 per cent of its workforce last month, with chief executive Barbara Peng citing “extreme traffic drops outside our control”. That’s all happening in the US as tech companies launch new AI features there first, meaning Australia is about to face a similar onslaught.
But Ms Rangan says there is still time for businesses to adapt. She said Hubspot itself performed a complete U-turn in 2022 when OpenAI unleashed ChatGPT, hailed as the beginning of the generative AI era.
“We had an early decision to make in terms of our product. The decision was: ‘is this a separate AI feature that we will charge separately as a SKU (stock keeping unit) or product, or should we just build it into the platform?’” Ms Rangan says.
“And our decision was to build it into every hub and every platform, so there is no bolt-on, no additional kind of SKU, product. That’s because very early on we looked at it and said ‘this is a colour TV moment’. When colour TV came on board, nobody went back and said, ‘I want to buy a black and white TV’.”
How to get noticed by large language models
HubSpot has now built its business away from search engine optimisation (SEO) to capitalise on the AI shift, building a suite of tools to help other businesses generate various forms of marketing content and be discovered by large language models and AI-generated summaries.
Ms Rangan said showing up in AI overviews in Google is completely different to SEO, which relied on keywords and the reputation of a source to reach the top of Google’s blue link search results.
But with large language models, Ms Rangan said it was more about the “number of repetitions of your brand and your brand authority across a variety of sources”.
That is what helps an LLM find you. It’s the repetition of what you’re trying to do, so the playbook is quite different and that’s what’s exciting.”
What this means is businesses now need to find ways to produce content, whether written articles, videos or podcasts on forums where their customers are likely to be. Content is still king, and it needs to be unique to stand out.
“The first part, in terms of diversification, is to start with a couple of channels. If you would rely primarily on Google, a good place to start is YouTube. Another good place to start is Instagram, Tiktok, depending on where your customer personas are,” Ms Rangan says.
The HubSpot software allows people to spread their content across different channels. For example it can take a blog post and turn it into a TikTok video or a LinkedIn post.
Getting closer to customers
This is crucial because in the large language model era, driving audience and customer growth isn’t about just attracting people to your website.
“You need to be where your customers are, so your entire content marketing strategy needs to change,” Ms Rangan says.
“And this is also where AI really helps, because when you understand the intent of your customers, your buyers, and you can map your information to where they exist in YouTube or Insta or Reddit then you can drive conversion.”
So how does AI help understand a customer’s intent? The solution is looking backwards to move forwards. Confused?
AI is essentially about drawing insights from copious amounts of data. For a company it can tap into queries from customers to support staff, online forums and other channels to better understand and create content that eases pain points. Then once a company gets that right it can then start to anticipate customer needs.
National Australia Bank jumped on this early, creating what it calls a “customer brain”. It is powered by 200 machine-learning models and has so far taken in 36bn data points, covering everything from transactions to the way customers use the NAB banking app.
The “brain” is designed to deliver the precise products to the right customers while sorting out day-to-day banking problems before they arise.
Construction software company Canibuild has adopted a similar strategy. Before AI, its customer resolution rates were 20-30 per cent with 24-hour response times. Now it has a 70 per cent resolution rate, freeing up sales staff to prioritise speaking to qualified buyers.
AirTasker founder and chief executive Tim Fung told this masthead earlier that he has spent $51m on media partnerships – including a Formula One sponsorship in the past year to get closer to customers rather than rely on Google.
“As AI has come to the fore, there’s so much pressure on a lot of these digital performance marketing channels like Google. Google used to be the go-to place to advertise but now that game has completely changed,” Mr Fung says.
“Brands are looking more and more towards being able to tell their own story, rather than relying on Google.”
Is AI getting more stuff wrong?
But when it comes to AI, Elly Strang, global content marketing lead at brand tracking business Tracksuit, says the technology isn’t designed to replace human faces or input.
“Say we’ve done a webinar and we want to redistribute that into blog posts or social posts or whatever – it (AI) can take care of that and we can focus on the parts that matter,” Ms Strang says.
“In the age of AI, brand is the most powerful differentiator and that’s everything from like your tone of voice, to great creative, the curation of the right resources for your audience. So I would rather my team focus their efforts on that and what sets us apart than some of the stuff that’s more automatable.”
Another reason why humans are important is OpenAI’s own research suggests that rate of hallucinations – or AI making up answers – is getting worse despite the technology becoming more sophisticated.
Ms Rangan agrees and warned that delegating tasks completely to the technology would create generic slop.
“A lot of people start with AI and they say ‘write me an article’. You can’t do that yet because it will give you a super generic answer that will not gain trust with your audience.
“The first thing to do is to start with a human voice, human strategy, human intuition and human point of view. When you start there, and then you use AI to augment, you know, your thought process and your narrative, then it’s super valuable. A lot of people miss that step.”
Ms Rangan says this is the way she prepares for company earnings calls and speeches.
She will feed an AI tool like ChatGPT or Claude past call transcripts or podcasts she has recently appeared on and LinkedIn posts so it can better understand her tone of voice.
“It helps refine my narrative and thinking. I obviously don’t share public information, but I’m using all of that information to start building my thought process. And I’ve done that for earnings scripts. I’ve done that for internal memos, as I’m working with, you know, our teams on what our strategy needs to be. So it’s now a necessary part of how I get work done.”
More Coverage
Originally published as AI is killing Google search: is your business next?
Mr Potato’s tax debts revealed
Court documents have lifted the lid on how much tax is owed by a company behind a spud-themed fast-food franchise.
Sonos Ace: why new software beats a new model
A year after its app disaster, Sonos has released a major software update that it says transforms its Ace headphones as it aims to win back trust. Is it worth the hype?