Search and SEO are evolving rapidly in the wake of new AI options. Many of our clients are concerned about continuing to receive a return on their SEO investment. They worry about putting effort into the right places. And they worry about how to prepare for a drastic shift in the landscape, should it come. 

The speed of evolution has made these questions difficult to answer with authority. But we conducted research, asked some experts, and have some theories that put these fears into context. Hopefully, they can help your organization navigate these uncharted waters.

Do AI Overviews reduce click-through rates?

In 2024, Google introduced AI-generated answers to queries in its search results. These “AI Overviews” are more likely to appear when a visitor phrases their search query like a question, using “what,” “how,” or “why” language. These overviews provide citations to their sources and a right sidebar (on laptops) with other references. Some are calling the traffic these overviews generate “zero-click” searches.

A screen capture of an example AI Overview as a result from a Google search

While the answer is yes, click-through rates have reduced by as much as 10%, others argue that most websites will be unaffected. For one, Google has scaled back their AI Overviews to only 1.28% of its billions of daily searches. This will likely increase now that AI has become less likely to provide incorrect answers, but the misconception that AI Overviews are everywhere is overblown. 

Further, the same article goes on to assert that 96.5% of all AI Overviews appear for informational keywords — meaning very few overviews are created for transactional, navigational, and local searches. Informational questions are much easier and safer for AI to answer and will likely remain the dominant use case.

Others argue that AI Overviews keep low-performing traffic away from your site. For many years, Google has already been answering queries with information cards. When you Google a business, you are likely to get a card with the business name, phone number, web address, and even a map with their location. Popular businesses might include reviews and specific details like daily open hours. These information cards have already been taking traffic away from your site. But was that the traffic that you wanted? 

These folks argue, if the searcher just wanted to know an answer to a question they had while having a conversation with a friend, they would have come to your site for that information and then left. Their visit would have counted as a bounce and negatively affected your monthly traffic data. Same with the ones that just needed a phone number or wanted to know what time you close. They would have come to your website for that one thing and then left.

Google’s own research says that when people use AI Overviews to start understanding a topic, they end up searching more frequently and express higher satisfaction with the results. Their position is that these overviews scratch the surface and help visitors ask more in-depth follow-up questions. Other recent studies have found that click-through increased for companies featured in AI Overviews, while those without an AI Overview lost traffic.

One thing is for sure: AI Overviews’ prominent position at the top of the results have pushed down organic results and made it harder for high-ranking organic websites to get noticed.

Takeaway:

Mixed. Yes, it is possible that AI Overviews are preventing click-through. It is also possible this traffic was not going to convert. And depending on your product and position in the market, AI Overviews might drive slightly more traffic than organic search. Either way, the result is an even more competitive search landscape than before.

Should I optimize my content for AI Overviews? 

The most obvious next question is “How can my brand rank for AI Overviews?” While this is an important question, remember that AI Overviews often include citations from multiple sources. So while your business may rank for an overview, it is likely not going to be alone. 

The answer to this question is more of the same things you should already know. In order to rank highly, you should: 

Lots of SEO companies want to help your business rank, and AI Overviews is the next frontier. But from all the articles we have reviewed (and there were many), the same best practices apply — there are no shortcuts to great content. 

Takeaway

Yes, optimize your content for AI Overviews, but this does not mean you need to do more than what you are already doing. To be a highly quoted source within your industry has benefits for brand recognition and trust, but just like long-tail keywords, these searches may have low volume. In the end, it is an investment vs. return question. There is a significant overlap between the sources cited in AI Overviews and the top organic search results, therefore, if your site already ranks well, you can’t do much more to get into an AI Overview.

Should I continue investing in SEO for Google?

Some clients worry that Google will be unseated as the dominant search engine now that tools like OpenAI’s ChatGPT have seen an explosion of millions of users. While these tools are indeed experiencing hockey-stick growth, Google completely dominates search volume.

SparkToro charted a 20% growth in search queries for Google in 2024, and crunched the numbers to conclude Google receives 373 times more searches than ChatGPT. 

To put that into context, Google handles 14 billion searches per day. The next closest competitor is Bing search with 613.5 million per day, followed by Yahoo, DuckDuckGo, and then Chat GPT. In other words, your investment would see a larger return if your team optimized content for Bing.com than for ChatGPT. 

These numbers are fresh from March 2025. Things can change, of course, but AI tools are not used only for search, have a relatively small market share, and do not get used daily. They suffer from not being the default tool at hand, which for most people, is a web browser. Google remains synonymous with search for a large percent of the population.

Takeaway

Yes, continue to invest in SEO for Google specifically. Google is still the biggest player in the search market, and their share is gaining, not decreasing (yet).

If we don’t implement structured data, are we losing out on AI crawler traffic?

Structured data is great for all SEO, so actually, you should implement structured data like Schema.org for across-the-board SEO value. 

For those of you using Google Tag Manager (GTM), you might know that you get some structured data for free. When a Googlebot crawls your site, it includes structured JSON data that it creates client-side, which means that Google gets the structured data but it is inaccessible to any other crawler. If the data existed server-side, other bots could access it. 

Most non-Google robot crawlers do not execute Javascript, therefore, they miss out on anything rendered in the browser. These crawlers include Bing, Yahoo, ChatGPT, Claude, and Perplexity. So again, server-side structured data would benefit all the search engine crawlers that are not Google.

But do LLMs really need structured data?

Large Language Models (LLMs) use statistical analysis to predict what word will follow the previous set of words. They do not understand language as much as they can mathematically reproduce its patterns. Therefore, they create structure from unstructured data all the time. 

But while LLMs process and understand unstructured text, providing structured data would significantly help interpret and categorize your content effectively and accurately.

Takeaway

The short answer is no, LLMs do not require structured data to create meaningful connections between content and search intent. But structured data would help them and any other search service to correctly label, tag, and organize your data. The longer answer is an investment in structured metadata would pay off in dividends for all search engines and crawlers.

How can we prepare for SEO’s evolving future?

In mid-2024, when Google first introduced AI Overviews, some in the SEO/SERP industry claimed sites could lose up to 25% of their traffic. That has not come to pass, with some sites reporting as high as 12% and others lows of 8.9% and 2.6% — not insignificant, but lower than expected. And the data is still coming in, with others reporting increases in traffic with specific kinds of intent.

While AI increasingly shapes search results, content strategy will need to shift for sites to remain visible and relevant. High-quality, authoritative, and authentic content that offers depth, accuracy, and unique insights is still valuable currency. AI algorithms are designed to identify and prioritize quality, trustworthy, and well-researched content for inclusion in their summaries. 

Sites should continue to target long-tail and question-based keywords to align content with visitor’s increase in natural language queries. This type of content is often more challenging for AI to fully synthesize and may still necessitate user click-through for a comprehensive understanding. Going deeper to investigate specific intents behind longer conversational queries could also be crucial for attracting relevant traffic. 

Finally, diversifying content formats by incorporating video, infographics, and interactive elements will continue to enhance engagement and provide unique value that text-based AI summaries don’t fully replicate. And optimizing content for featured snippets remains important, as appearing in these snippets increases the likelihood of a website’s content being cited within AI Overviews. 

Takeaway

The fundamentals of great content and best-practice SEO has not changed as dramatically as the tools that crawl your site and serve your content have.

Final Thoughts

Anything in the tech space evolves rapidly, and SEO is no exception. While the methods and the tools we leverage might change, the fundamentals remain strong. Keep doing what you have been doing, keep being curious, and keep asking these important questions of those in your circle whom you trust. We’re all figuring these things out in real time and can benefit from each other’s expertise.

If you have in-depth questions about SEO, content management, and the evolving AI-powered landscape, reach out to our team and we’ll always do our best to answer them thoughtfully and from multiple angles.

AI disclaimer: Google’s Deep Research was used for initial exploration and source gathering. All sources cited in this article were reviewed by the author. ChatGPT was used for follow up questions, as well as AI Overviews for examples of common questions. This article synthesizes these sources and was written by a human.

What’s it like to work at Oomph? It’s about more than just delivering world-class digital experiences. It’s about being part of a team that’s given the freedom to explore, experiment, and innovate. Enter: Innovation Labs — our twist on the classic Hack Day, designed for a remote-first workplace.

The Spirit of Innovation

At Oomph, we know great work starts with great opportunities. That’s why we give our team dedicated time to develop ideas, play with new tools, and push their skills in new directions. Our Innovation Labs are quarterly events where the entire company collaborates on solving real-world challenges — while having a lot of fun in the process.

For our first Innovation Lab this past fall, the mission was simple: use generative AI (GenAI) to save at least one hour of work per week. With AI reshaping the future of web development, this challenge gave everyone — from designers to project managers — the chance to learn about cutting-edge tech and apply it in meaningful ways.

How We Innovate

Here’s how we structured the Innovation Lab:

What We Built

The projects were as diverse as the AI models we worked with, showcasing the creativity and problem-solving power of the Oomph team:

  1. Team OpenAI: Project Estimating

The team developed EstimAItor, a bot designed to make project estimating more efficient by analyzing historical task data from Jira. They exported a year’s worth of task titles, descriptions, and time log data, integrated it with OpenAI’s API, and built a Python-based app. The bot could estimate task durations, provide confidence levels, and identify related issues. The result? A tool that eliminates guesswork and streamlines the estimation process.

  1. Team Anthropic Claude: Project Estimating

Similar to Team OpenAI, this group leveraged Anthropic’s Claude model to connect with Jira and estimate task durations. Claude was tasked with analyzing past data to validate and refine project estimates. While the tool wasn’t fully functional by the end of the day, the team identified several exciting use cases, such as using Claude to double-check estimates against past projects, combining data fields (like titles, descriptions, and time logs), and producing insightful summaries.

  1. Team Hugging Face: Meal Prep

This team aimed to tackle a time-consuming part of everyday life: meal prep. They used the Hugging Face Inference API along with meta-llama and Llama-3.2-1B-Instruct models to scrape data from recipe websites like AllRecipes. Their tool could simplify recipes, generate concise shopping lists, and help users reduce time spent on meal planning. While the AI successfully summarized recipes and provided shopping lists, the team noted some uncertainty about whether the data extraction was based on the recipe URL or similar recipes in its training set.

  1. Team Mistral: Project Estimating

Using Mistral, Jira, Postman, and GitHub, this team created a tool to make task estimation more precise and context-specific. Their solution used historical data to suggest time estimates, which users could either accept or refine further. The team also explored how the tool could help assign the right team members to specific tasks based on past work. By eliminating manual guesswork, this tool could save significant time and effort for project managers.

  1. Team Gemini: Meal Prep

Gemini’s project added a dose of personality to meal planning. Using a photo of the user’s fridge contents, their tool suggested recipes — complete with a cheeky Gordon Ramsay-inspired twist. For example, if the fridge contained eggs, milk, and vegetables, Gemini might suggest an omelet while delivering a signature Ramsay quip. The tool assumed basic pantry staples (like flour and rice) and allowed users to customize meal types and dietary preferences. When one of the team members took a photo of their fridge, it even recognized cat food in the photo. Thankfully, Gordon didn’t suggest a cat food souffle. 

  1. Team Llama AI: Time Tracking

The Llama Log tool automated time tracking by connecting GitHub commits with Jira Tempo timesheets. Using GitHub Webhooks for notifications and the Jira Tempo API for updates, the team streamlined the process of logging hours for developers. Despite running into storage limitations with Ollama and compatibility challenges with Meta models, the team successfully demonstrated how AI could automate workflows and even integrated branding and content generation into their project presentation.

The Winner: Team Gemini

At the end of the day, the entire Oomph team came together to present their projects and vote for the winner. No judges, no top-down decisions — just a democratic process where every voice counted. The winning project? Team Gemini, whose creative and functional recipe tool stood out for its innovation and personality.

Ready to Innovate?

If this sounds like the kind of place where you’d thrive, we’d love to meet you. Explore opportunities at Oomph and bring your ideas to life with us.

JOIN OUR TEAM

Was this blog written by ChatGPT? How would you really know? And what impact would it have on Oomph’s site if it were?

Yes, we know there are some great AI-detecting tools out there. But for the typical reader, picking an AI article out of a crowd can be challenging. And with AI tools like ChatGPT delivering better-quality results than ever, many companies are struggling to decide whether to hand their content and SEO reins over to the machines.

While AI can add value to your content, companies should proceed with caution to avoid some potentially big pitfalls. Here’s why.

Quality Content Is Critical to SEO

All the way back in 1996, Bill Gates said “Content is King.” This phrase became ubiquitous in the early years of SEO. At that time, you could rank well simply by writing about a search topic, then optimizing your writing with the right keywords.

Since then, search algorithms have evolved, and the Google search engine results page (SERP) is more crowded than ever (not to mention the new continuous scroll). While ranking isn’t as easy as it used to be, content — whether it’s a video, an image, a product, a blog, or a news story — still matters. When content ranks well, it’s an ad-spend-free magnet for readers that eventually become customers and subscribers. What else on your website can do that?

That makes your content special. It also puts a premium on producing a high volume of relevant content quickly. For years, brands have done this the old-fashioned way: with copywriters and designers researching, writing, revising, creating images, and publishing ad infinitum.

Until AI.

AI-Powered Content Generation Changes How We Make Content

There’s no point in denying it: AI will impact SEO. But it’s still up for debate just how deep that impact will be.

The rise of AI-powered language processing tools like ChatGPT and Meta’s Llama makes quick content generation a reality. They can easily produce high-quality content that will likely only get better with time. ChatGPT can produce an article in minutes, not hours, and even suggest keywords for you.

For all those reasons, marketers have embraced these new tools — ChatGPT shattered records when it reached 100 million daily active users in a mere two months. As the saying goes, though, just because we can, doesn’t mean we should — especially if using it means compromising on quality or losing what makes us, well, us.

After all, AI is a tool, not a human replacement. If you’re considering integrating AI into your martech stack, here’s what to keep in mind.

4 Things to Know Before Using AI for Content

1. Google may ding you.

In August 2022, Google released an algorithm update called the “helpful content update.” You can read more on Google’s blog, but what’s important is this: Google prioritizes content by people, for people.

In other words, using AI solely to churn out articles that feed the search engine beast won’t work. You should continue to create content that puts your audience first, is genuinely helpful, and is 100% unique to your brand.

But Google’s stance does beg the question: Will they really know if you used AI to generate your content? Well, they do have a very special set of skills…

Animated gif of Liam Neesen mouthing the words “I will find you” from the movie Taken (2008)

While AI content can be good quality and easy to read, tools like Crossplag’s AI Content Detector and OpenAI’s GPT-2 Output Detector can detect whether something is written by a human or by an AI bot. That comes down, in part, to the fact that AI lacks EEAT – experience, expertise, authority, and trust – which is not just a major factor in SEO, but a major reason why people consume content in the first place.

2. AI isn’t always accurate.

AI culls information from across the web, and it’s remarkably good at it. But aggregating massive amounts of information isn’t the same as fact-checking. AI tools often get it wrong – even with easily verifiable, fact-based questions.

Take the question: What is the largest country in Central America that isn’t Mexico?

AI’s answer:

The largest country in Central America that isn’t Mexico is Guatemala. Guatemala has an area of about 108,890 square kilometers, making it the largest country in Central America by land area. It is bordered by Mexico to the north, Belize to the northeast, Honduras to the east, El Salvador to the southeast, and the Pacific Ocean to the south.

Source for AI’s answer

Google’s answer:

Nicaragua is the largest country in the subregion, with a total area of over 130 thousand square kilometers, followed by Honduras, with more than 112 thousand square kilometers.

(Hint: Google is right)

This is a problem for your business and for your SEO. Accuracy is an important part of EEAT. It’s especially critical for “your money or your life” (YMYL) content, like financial or medical advice. In these cases, the content you publish can and does impact real people’s lives and livelihoods.

Spotty accuracy has even prompted some sites, like StackOverflow, to ban AI-generated content.

3. You don’t have the rights to your AI-generated content.

AI-generated content isn’t actually copyrightable. Yes, you read that right.

As it stands, the courts have interpreted the Copyright Act to mean that only human-authored works can be copyrighted. Something is only legally defensible when it involves at least a minimal amount of creativity.

We’re all familiar with this concept when it comes to books, TV shows, movies, and paintings, but it matters for your website, too. You want your content and your ideas to be yours. If you use AI-generated content, be aware that it isn’t subject to standard intellectual property rules and may not be protected.

4. AI-generated content can’t capture your voice.

Even if you fly under Google’s radar with your AI content, it still won’t really feel like you. You are the only you. We know that sounds like it belongs on an inspirational poster, but it’s true. Your voice is what readers will connect with, believe in, and ultimately trust.

Sure, AI may succeed at stringing together facts and keywords to create content that ranks. And that content may even drive people to your site. But it lacks the emotional intelligence to infuse your content with real-life examples and anecdotes that make readers more likely to read, share, and engage with your content and your brand.

Your voice is also what sets you apart from other brands in your industry. Without that, why would a customer choose you?

AI and SEO Is a Journey, Not a Destination

AI is not the end of human-driven SEO. In reality, AI has only just arrived. But the real opportunity lies in finding out how AI can enhance, not replace, our work to create winning SEO content.

Think about content translation. Hand translation is the most premium translation option out there. It’s also costly. While machine translation on its own can be a bit of a mess, many translation companies actually start with an automated solution, then bring in the humans to polish that first translation into a final product. If you ask us, AI and SEO will work in much the same way.

Even in a post-AI world, SEO all comes down to this guidance from Google:

“If it is useful, helpful, original, and satisfies aspects of E-E-A-T, it might do well in Search. If it doesn’t, it might not.”

If and when you do decide to leverage AI, keep these tips in mind:

At Oomph, we believe quality branded content is just one component of a digital experience that engages and inspires your audience.

Need help integrating SEO content into your company’s website? Let’s talk.