Key takeaways
- AI tools like ChatGPT will complement search, not replace it.
- People don’t just want answers; they want an experience.
- However accurate the answers sound, people want to know what led to them.
- Use AI as a research assistant, not a replacement for human creatives.
As a content marketing platform, we’ve always known content itself is only half the story. The other half is search: the ability to be found, getting your ideas in front of the people you want to reach. It’s why Search Engine Optimisation is such a major part of digital marketing – and why experts like the multitalented Neil Patel have made a fortune from it.
With AI lighting up the web in 2023, Neil recently penned an article: How ChatGPT, Bard, and AI Will Impact Search. Like a lot of his work, it’s a good one – stuffed with useful links and supported with hard numbers. And while we broadly agree with his conclusions, we think there’s more to be said. (Isn’t there always?) So here’s the Contentoo take on Neil’s piece, adding some perspective from our experience working with hundreds of content marketers and thousands of freelance writers and creatives.
TL;DR: We think ChatGPT, Bard, and similar are going to expand the universe of search – not replace it – and make it even more useful to marketers. Of course, feel free to disagree!
A wrong answer doesn’t mean zero value
Neil’s first point is about the content ChatGPT generates not being accurate – and why this isn’t a big problem. We agree. Yes, an AI search chatbot can generate wrong answers – but so do search results today. As search users, we’re aware that no search is 100% accurate, and we expect to explore a few results before finding what we want.
The point is that just as search got better with time (who visits the second page of results any more?) so will these AI tools. The “source data” is the same (web content) but the methods for making sense of it will improve, just as Google’s PageRank algorithm has evolved.
When it comes to content creation, we see these tools being used more like research assistants than writing partners, with savvy freelancers using them to build an understanding of what information their content needs to be relevant to an audience – just as they use traditional search to gather information today. And just as with human research assistants, content creators will accept that they can get things wrong … and compensate for it.
Why trust is key: human behaviour matters
Another reason Alexa and Siri didn’t replace the Google Home Page: a lot of the time, human searchers don’t want a singular answer. Not all questions are as simple as “Siri, what’s a good price for a hatchback?”
When searching, we want choice, to be aware of options available. It’s why few people ever use the “I’m feeling lucky” button at Google, which takes you straight to the top result.
After all, understanding a subject in depth means “reading around” the immediate question, seeing context and backstory and digging deeper into the topic. Often – this happens a lot to our freelance copywriters! – you’ll find the question you asked wasn’t actually the best phrasing, or even the right question at all.
So we’d add to Neil’s opinion here. Yes, people want quality of results, not quantity – but in most cases they do want more than one choice of answer. Meaning the AI of natural language answers and the AI of search engine results pages aren’t operating in the same space.
The web is about more than copy
Another point we’d expand on: the web is more than the text on its pages. Websites aren’t textbooks; they are Pillars of Content, destinations that people come to love and respect. Whether they’re selling products, services, or information, no two sites are successful in quite the same way – because no two audiences are alike.
An attractive web design may mean nothing to an AI, but a lot to a human. While the site’s writing style – thoughtful and poetic, laugh-out-loud humorous, shamelessly sales-focussed – can make all the difference to audience appeal and the decision to buy. The human has emotional triggers the AI lacks: sharing information in a chat window is not the same as sharing an experience on your favourite website.
And much of what people want from the web is an experience. We want to go shopping, not have someone shop for us. So while these new AIs are great at processing text, they’re less good at understanding what makes us human.
That’s why we expect AI-driven Large Language Models (like the OpenAI project ChatGPT is based on) to become a useful tool for goal-driven, single-answer searches – like “What’s the price of petrol at my local supermarket?” But for those exploring life’s depths and nuances, the traditional search engine with its pages of choices will prevail.
AI and search: the beginning of a beautiful partnership
So we’re in agreement with Neil so far: if you want an answer, ask an AI. If you want an experience, search the web. Is there any common ground where the two methods can mingle … leading to something new?
We believe there is.
Whether you’re querying a search engine, typing to a chatbot, or having a voice conversation with Alexa, you want the same thing: results that address your needs. And if those needs can be answered by a marketer’s product, service, or content, you’ll still need to take a second step: buy the product, book the service, or visit the marketer’s web pages.
An AI can advise you that you have a flat tyre. But to inflate it you’ll still need to buy a pump. And that’s why ChatGPT, Bard, and others ultimately add to the importance of having great content that plays well in search rankings, whatever the search method. (Voice commerce is expected to drive US$19.4bn in revenue this year.)
When Google launched its “knowledge graph” (the box at the top of the results page summarising information on what you searched for) some people thought it would kill content optimisation, because it answered the user’s query with no need to click through anywhere. It didn’t. While “zero click searches” do exist, they’re only a small percentage of the total. As Neil Patel confirms, most of the time people are clicking through to websites, whether they’re using a voice assistant, chatbot, or search engine. And that’s the marketing opportunity.
Use AI to find it – but not to write it
Let’s end with the biggest question of all. Since tools like ChatGPT can generate grammatically-accurate text in seconds, should you use it to write your content, not just research it?
As a platform for content creation, we’ve obviously got an avatar in this boss battle. So we’ll limit our answer to a single data point: no – because what builds interest in your content is your “voice”. Your personality, how you engage with your audience, the passion you show for your subject. All things the average AI, trained to generate statistical probabilities rather than refreshing originality, struggles with.
As Neil says, “No one wants to read regurgitated content that’s just a spin or existing content.” They’re looking for experience, expertise, authority, and trust – not the buzz of a thousand opinions, but a single confident voice that makes them feel they’ve made the right decision.
And when you use our human freelancers – the top 10% of the sector – that’s what you’ll get.
Discover what the Contentoo platform can do for you, talk to one of our Content Advisors today!
