Agentic Search: The Future of Search Isn’t Just Rankings
Search engines will do more than just recommend
Google are killing the golden goose which made them the titans of search.
Why? I suppose there are two reasons. Firstly, competition from the likes of ChatGPT has forced them to keep up. But secondly, they’re gambling on a new experience which they’re hoping will keep them in the lead.
It won’t be long before AI Overviews aren’t the only thing stealing clicks from your site. Only now, Google aren’t just coming for your informational content.
They’re coming for everything.
What if Google could help customers find your business and book your service — all without even visiting your website at all?
Here’s what we’re looking at this time:
Google’s latest play to retain its place as the “big dog”
How it’s going to change the search landscape
What businesses will need to do to keep up
Let’s check it out!
What’s happening to search?
For the last 20 years, Google’s model worked like this.
User types query into search engine: “historic sites in London”
A list of results related to the query comes up
At the top of this list are sponsored links from museums and tour companies
When/if user clicks on a sponsored link, they get taken to the business’ website. Google earns some money from the click.
Of course, not all links are sponsored.
Google still needs to understand what resources are relevant to the search. If you searched “historic sites in London” and didn’t see any reference to Westminster Abbey, you might think something was amiss.
You might even be tempted to run off and use Bing instead. *shudders*
This is the model that’s been powering Google since its start, and has led them to bring in over $200 billion in revenue in 2024. It’s their ability to pick the most relevant content which has propelled them to the top.
However, this model is changing.
It won’t be long before Google rolls out AI mode as the default, and it’s not just recommending the same content to you with a fresh lick of paint.
Google are going from a recommendation to a personal assistant, and they’re doing so with “agentic AI”.
AI agents
Imagine searching “cheap flights to Barcelona next weekend”.
In today’s experience, you’d get a list of links to airlines, aggregators and perhaps the odd travel blog. You’d click through to the site, try to complete the journey for the date you wanted. Maybe you got stuck or bored. You might go back to the results page, click through to another site and finish the journey there.
In tomorrow’s experience, you might just get this response instead:
“Here are three return flights for under £120. I’ve pre-filled your dates and times. Want me to book?”
All you’d need to say is “Yes, please!” and you’re off to the races.
This is what we call an agentic search; AI doesn’t just find results for you, but can act on your behalf, too. This could also mean things like making restaurant bookings, or even ordering things from eCommerce websites like Amazon.
Going forward, users will need fewer clicks to get more done before they’ve even seen your website — and that’s if they see it at all.
But Tommy! You said to keep doing traditional SEO!
That I did.
Here’s the kicker: having AI models scour the internet for the answers to every query will be prohibitively resource-intensive. So, much of the underlying tech that has powered Google’s recommendation engine all this time will still be in play.
So the good news about this is that a lot of the old SEO techniques will still get results. They just won’t be enough anymore.
That said, here are some things which will be fundamentally different.
Informational blogging will lose most of its value. We’ll touch on this again shortly.
Keyword-targeting will similarly lose most of its value. The goal will be to have clear information on your products which AI can use to decide if it’s a good fit for a buyer’s needs.
So, what does this mean for businesses?
Here’s how I think this will change the world of search.
Less traffic; more business will happen off-site
The goal of “informational” blog posts was once to just get people onto your site.
A company selling beauty products might have written a blog post on “What is retinol?” to get a handful of clicks.
Of course, the chances of someone going from wondering what retinol is straight to buying a £200 bottle were slim. But for the cost of writing one blog post, it was worth it.
This strategy has been dying a slow death since the dawn of Google’s AI Overviews; clicks are down 34.5% on informational searches in the year to March 2025. AI mode will be the coup de grâce; any top-level information on blogs won’t earn you any clicks or business anymore.
Even for content that can still earn business (middle-of-funnel blog information, which I spoke about in my last post), it’s going to be trickier to tell how it’s performing. Google might provide sites with analytics as to when pages were cited in AI answers; we just don’t know yet.
But in a lot of cases, customers will be interacting with your business off-site.
Over time, the raw number of site visitors will become less useful as a metric.
Programmatic access/structured data is survival
This already exists for a lot of businesses.
Think of how Skyscanner works with airlines, or how Expedia works with hotels. They use what’s called an integration.
This just means that engineers at Singapore Airlines might hand the engineers at Skyscanner a digital map, saying, “This is where to find what you need”. The map for British Airways might be identical, whereas the map for American Airlines might be completely different.
Google’s AI agents might use a similar map, but it’ll take a different form. Similar to how sitemaps work, founders may be asked to provide a “service map” to teach Google how to make bookings and purchases (note — this is an informed guess, Google hasn’t published anything about the existence of a “service map” to date).
Whilst Skyscanner wouldn’t work without the map from the airlines, Google would similarly find life much harder without the service map from your business.
The form it’s most likely to take is structured data/schema.
The SEO community seems torn on whether LLMs like ChatGPT and Perplexity are utilising structured data.
But once Google’s AI agents go live, it’s not even going to be optional anymore.
Trust is the tiebreaker
User searches for “unusual gifts for Father’s Day”.
Google checks its indexed data and finds 2 sites with decent collections (it will be more in reality, but we’ll say this for the sake of argument). However…
All product descriptions are decent
All prices are similar
They both offer free postage and packaging
I guess what I’m saying is that, when it comes down to it, not a huge amount separates some eCommerce stores.
Which does it recommend to the user?
Google will scour the internet for any signals of:
Trustworthiness; reviews on sites like Trustpilot or recommendations on Reddit
Authority; brand mentions in reputable publications, and consistent branding across the internet.
Clarity; clear returns policies, accurate stock levels, and shipping that matches what’s promised.
This is to say, the “winner” in these situations won’t just be the brand with the best SEO. It will be the brand with the most trustworthy digital footprint.
Remember that users trust AI recommendations.
If AI trusts your brand, you’re set to pull ahead in this race.
To conclude — adapt or be left behind
This shift won’t happen overnight.
But going forward, your site’s job won’t just be to “attract clicks” anymore. It will be to prove to machines that your business is worth trading with.
The site will need to be:
Easy to understand (structured data)
Easy to trust (reviews and authority signals)
Easy to transact with (clear offers and policies)
For my part, I think it’s a huge shame that websites, from being an attractive digital storefront for customers, might just become glorified directory listings with “stats” that AI learns to listen to.
But the direction of travel is obvious. The brands that thrive in this new age will be the ones which leave behind the notion of optimising for rankings.
The successful brands will be the ones who adapt and start optimising for eligibility in an agentic world.
Remember to subscribe to Seen by Search to be kept up with how to stay ahead of the curve as more changes roll out with this ever-changing landscape.
Written by Tommy McDevitt, a Senior Engineer with 10+ years experience. Now advising eCommerce businesses about SEO and Digital Marketing. Want to work with me on your online visibility? Let’s see what we can achieve together!
Further resources…
What does “Agentic AI” mean?
Agentic AI refers to AI systems that can take actions on behalf of a user. This is in contrast to generative AI, which generates outputs like text, images and code. Agentic AI is designed to act like a digital assistant or concierge, with the ability to plan tasks, make decisions (within set parameters) and execute real-world actions, like:
Booking flights and hotels or making restaurant reservations
Filling out forms or completing online checkouts
Comparing products and recommending the best option
Automating multi-step workflows across different apps or services
Agentic AI goes beyond just “answering questions”. It can interpret intent, decide on next steps and implement them, often with little to no extra input from the user.
Are AI agents the future?
Yes, AI agents are widely seen as a major part of the future of computing and search.
Unlike traditional AI tools which generate text or images (like ChatGPT or Midjourney), agents can plan tasks, make decisions and act on behalf of users without much input from them.
Tech leaders like Google, Amazon and OpenAI are investing heavily in agentic AI because it shifts search and digital services from just finding information to getting tasks done. Analysts expect these systems to become more common in everyday workflows, especially in areas like eCommerce, customer service and productivity software.
While agents won’t fully replace existing apps or search engines, they’re likely to become the default interface for many tasks. AI agents represent a key direction in the future of how people interact with the web.
Will Google’s AI mode replace SEO?
No, it won’t. But it will change how SEO works.
AI mode shifts the search experience from showing “10 blue links” (the SERP) to providing direct answers, and actions, through agentic AI. This means fewer clicks will come from informational blog posts or keyword-targeted pages.
However, SEO fundamentals like crawlability, site speed, structured data and clear product information will still be essential as AI systems rely on those signals to understand and recommend websites.
What’s changing is the goal: instead of only ranking for traffic, businesses will need to optimise for eligibility; this means making their content, products and brand trustworthy and machine-readable so AI agents choose them when completing a potential customer’s task.