There is a whole constellation of AI tools promising to change the way you interact with the Internet. Perplexity. SearchGPT. Copilot. Google AI Overviews. All of these tools (and others) are trying to replace traditional search engines.
AI is here to stay. And we’re just starting to feel its impact on search.
How did we get to this point? What does it mean for your business?
It’s critical to understand how search has evolved … and what AI means for where it’s going.
A Brief History of Search
When the Internet became public, people realized quickly that they needed some way to find information. The amount of information at a user’s fingertips was powerful … but if they had no way to access it, it wouldn’t matter.
Enter search engines.
Early search engines were clunky. Some, like AskJeeves, leaned on human-created lists. Others, like Yahoo!, Altavista or WebCrawler, used early web crawlers to power their search — much like Google today, but with little algorithmic complexity. If you searched for “novelty lamp,” you might get a site that sells lamps, the National Lampoon, and a review of DuckTales the Movie: Treasure of the Lost Lamp.
Then things changed. I still remember when my grandmother switched from WebCrawler to Google for her searches. It stuck in my mind because I thought the logo looked funny … I would have been about 8 or 9. Little did I know that I was witnessing history. Google rapidly became the dominant force in search due to the strength of its search algorithm. It just seemed smarter than its competitors.
Smarter search mattered as the web began to explode. Google was founded in 1998, right before the Internet leaped into public consciousness and caused a huge growth in websites. Today, there are roughly 1.1 billion websites in the world (depending on how you count); in 2000, there were 17 million (and that was a 438% increase over 1999, when there were only three million). In other words, if you were searching the Internet in 2000, you were searching less than 2% of the amount of information that exists online today.
People had started to realize the Internet’s potential: with the right middleman, it became a question-answering machine. A good search engine (like Google) let you could ask everything from “Where can I find a Shadowrun Second Edition sourcebook?” to “How old is Brad Pitt?” and connected you to a page that answered your question.
Marketers realized the potential quickly. Questions meant a need for answers. Yahoo!, Altavista and Google didn’t create the content themselves … if there was no good answer out there, they couldn’t provide it. And if that answer involved selling a product or service, there was money to be made.
That birthed the SEO industry and the content revolution. If you were answering the questions your audience was asking, you could rely on middlemen like Google sending you a steady flow of interested traffic.
Fast forward to today.
There’s a broad consensus that middlemen (particularly Google) aren’t working the same way anymore. You’re no longer guaranteed a good answer; there’s so much content that even major players are having a hard time parsing which pages have good information, which pages are irrelevant to the query and what people actually want to see.
Enter AI search.
The Promise of AI Search
When search engines came on the scene, there was a simple process to be followed if you wanted information. You asked your question to the search engine (usually Google, but these days sites like YouTube, LinkedIn and TikTok also count). That search engine would try to figure out which pages answered your question best and display the links to you. Think of the search engine as your personal middleman — if you need something, he knows a guy, but he’s only arranging introductions, and he doesn’t ask any questions about why you need it.
The promise of AI search is to go one step further. AI search is more like a personal assistant than a middleman: it’s not just setting up a meeting between you and the person who can answer your question, it’s doing the legwork itself.
Why?
Because AI can parse both the question and the intent behind the question.
I’ll illustrate. I don’t have a personal assistant, but my late grandfather did. If he asked one of his other employees to get him a cup of tea, that employee would do one of two things: bring him all the tea in the office cabinet and ask which one he wanted, or make him whichever one they thought he would like best.
If he asked his personal assistant to make him tea, she knew that when he said “tea,” he meant Twinings Earl Grey. In a pot, with a teacup, not a mug, and a little milk on the side.
The current crop of search engines is like the employee. They know when you say “tea” that you want “tea;” they know how to find the tea and brew it. They may have some idea that you want a teacup if they’ve been watching closely (through cookies and personalized data). But they won’t bring you Twinings Earl Grey in a pot. They’ll bring you 20 different types of tea from the cabinet, a teapot, all the mugs and all the teacups in the cabinet, a carton of milk, a carton of half and half, a carton of cream, honey, sugar … you get the idea. Then you choose.
AI search, on the other hand, will serve you a hand-tailored answer. You’ll never see the carton of milk or the box of tea; you’ll get a tray with the pot of Earl Grey, a teacup and the extras on the side.
The downside right now is that if you ask AI search to make your tea and there’s no Earl Grey, it’ll bring you whatever it wants (rooibos, chai, gunpowder green …). Or worse, it’ll just throw multiple teas into the same pot. The tools are not refined yet; I personally still primarily use traditional search engines for this reason. As the technology progresses, though, these sorts of hallucinations will decrease. By its very nature, this is the kind of task a large language model is made for.
What Does the Evolution of Search Mean for Your Business?
So … as a business owner, what does this mean for you?
First, it means that SEO as the only driver of business is dead. Search-based visibility is getting tougher anyway unless you’re paying for placement. Traffic is volatile; a lot of creators who’ve relied on search traffic are up in arms, as the recent Web Creator Summit showed. As content gets cheaper and easier to produce, it’s going to be harder and harder for traditional search engines to find you if you don’t stand out.
If search has been the primary marketing pillar of your business, it’s time to diversify. We’ve talked already about the importance of building a permission-based list so you can contact your audience directly; this is even more critical today than it’s been for the last 20 years, especially as the middlemen start hiking the rent up. The two subsets of creators hit hardest in the last year are those that rely on YouTube and Google for revenue. Brands you would think are so big and successful that they can live off traffic alone are now frantically building Patreon offerings, creating branded merchandise, building email lists and monetizing their audience. Visibility is down, costs are up, and the time to own your audience was yesterday.
You also need to find other areas to be visible to your audience. That means expanding your social media efforts, building partnerships with people who are relevant to your niche, ad campaigns, retargeting … a full-spectrum effort. The days of letting one channel carry your business are over — the smart money’s already moved.
Search is still relevant. It’s an ally that can help you organize your thinking, define your brand and business and establish your authority. But you can’t just publish and expect traffic anymore. You need to create content that’s useful and stands out from the pack and hyper-target your niche. Posting a blog every week won’t cut it.
AI Search Tools We’re Watching
Like the early days of traditional search engines, there are a lot of players vying for supremacy right now and no clear-cut best option. AI search is a tiny fragment of the overall search market, and the issues haven’t been solved yet. But you need to know where the market is going.
Here are a couple of AI search tools you should be aware of that have a mix of popularity, good backing and strong utility:
- Google AI Overviews: The 800-pound gorilla of search knows where the market is headed, and its AI Overview tool is already showing up for many searchers, condensing a page of results into a compact paragraph or two.
- Microsoft Copilot: Not to be outdone by its fellow tech giant, Microsoft has launched its own AI search engine. Copilot lets you ask natural language questions and get tailored answers from the entire Internet; it also offers voice search.
- SearchGPT: Built on the massively popular ChatGPT, this tool works similarly to Copilot.
- Perpexity.ai: Not as popular as the other options, but has a cult following among techies due to perceived accuracy.
Even if you don’t want to use these as a primary search option yet, take the time to test them out and see how they work. Knowledge is power.
The Time to Adapt is Now
As Google CEO Sundar Pichai notes, we’re only in the infancy of AI right now; the LLMs we’re currently using will continue to be refined as time goes on, and new technologies will make results faster and more reliable.
In a May 20, 2024 interview with The Verge, Pichai laid out what he thought the next five years would hold:
The way people use AI to actually solve new things, new use cases, etc. is yet to come. When that happens, I think the web will be much, much richer, too … there will be entrepreneurs who figure out an extraordinarily good way to [use AI], and out of it, there’ll be great new things to come.
AI search isn’t going to be a bed of roses for everyone; this is a time of upheaval in the way we interact with the Web. But that also means opportunity — so don’t wait. Get on board now and prepare for the next iteration of the Internet. And if you need a hand, we’re here to help. Just drop us a line, anytime.
Best Regards,
David Brandon
Copywriter
Rainmaker Digital Services