Does On-Site Search Still Make Sense in an AI Search World?

Maybe you’ve had a recent positive experience searching for information (we’re not talking about products on an ecommerce site, but rather general details) on a company’s website. And maybe your customers love nothing more than spending hours searching random facts on your site. But ask yourself: was it worth it?

Of course you want visitors to find what they need in as simple and streamlined way as possible. For decades, that’s been the global search bar, the failsafe when your site architecture and navigation fails. But that seamless experience from search query to results listing takes a ton of work. And with answer engine optimization (AEO) or generative engine optimization (GEO) rising to prominence, it’s worth asking yourself this tough question.

Search engines like Google already are the starting point for 93% of all online experiences. And generative AIs like ChatGPT and Copilot are becoming more popular with the ability to send users to the direct page and information they're looking for. Meanwhile, 90% of website pages never get organic traffic.

It’s a tall ask for your website’s on-site search to compete with Google and ChatGPT. Think of it like a little league baseball team trying to play in the MLB. They’re following the same rules to achieve the same results, but one has vastly superior resources.  

Setting Up Website Search Is Painful

Setting up and configuring on-page search for your website is not an easy endeavor. On the outset, you need a separate tool to crawl, index, and serve results. That may be built into your CMS platform or require extra expense to operate.

Then comes the rules or prioritization of content – if a user types certain words, what should we show? What if they misspell words or use different regional words for a topic. Are there rules to rank certain types of content? Do we want to offer categorization and filtering, which then means we need to decide on a taxonomy? Pagination? Sorting? Layer in responsive mobile considerations to present filtering in a different UI pattern or lists of large results with lots of scrolling required? You get the picture.

How do you manage and control this easily? You don’t. Well, the longer answer is to do it well, it takes a significant amount of time and money to license a platform, implement a solution, and continue to maintain it.

And the statistics of on page search usage and it continually dropping don’t lie. Less than 30% of users say they use on site search, and that includes on ecommerce sites, so it’s safe to assume non-ecommerce search use is much lower. Many of our clients note how few users are actually engaging with search as confirmed by analytics.  

The Future of Website Search and Value

Practically every website is developed with a search bar in the global navigation. And the operations are the same as with the big search engines: enter a keyword and it will deliver links to the most relevant pages. But I just explained how that’s changing. Generative AIs aren’t just giving relevant links. They’re providing more robust answers.

This leads to three questions we have to ask ourselves:

1. Is the value of a website evolving?
2. Are we about to see a shift in the architecture of how sites are built and maintained?
3. How do you scale if it’s all going to change in three years? 

Let’s start at the top. I’m not ready to say that websites are less valuable now. But the metrics and methods we use to try and quantify their value are definitely changing. Metrics like number of visitors, time spent on site, and page views aren’t going to be as important as they once were. Soon, the fact that someone had to hit 4-5 different pages on your site may mean that your content and optimization strategy wasn’t focused enough to get them where they need to be to begin with.

And fewer people are likely to come in from traditional search channels. Last year, Gartner predicted that overall traffic driven by search engines will drop 25% by next year. What won’t change is that consumers will look to digital methods of getting the information they seek. It just might not be from a Google search bar as much.

What Else Could You Be Doing?

Strategy is all about prioritizing one area over another. Which means all that work and focus on an on-search experience is coming at the expense of something else your teams could be doing to improve your customers’ engagement with your company.

So if a small percentage of your visitors use on-site search and it’s only driving them to 10% of your site’s pages, don’t just ask if it was worth the effort, but also what other features you could be providing instead. Could those resources instead optimize your site’s architecture, content delivery, conversion, or AEO/GEO optimization.?

Again, there’s still value in on-site search for ecommerce products. But if visitors are hitting your site and not planning to enter their credit card information, there’s a better way to spend your time and money, such as prepping for the AEO/GEO era.