July 2025 | ariel lim consulting

AI Mode Is Here: 3 Ways It Will Impact Traditional SEO

July 9, 2025 by Ariel Lim

This morning, I happened to use Safari on my iPhone to search for something. As with most users who have an iPhone, Google is my default search engine.

During my search, I saw an AI Overview. This has been showing up more and more for infornational searches recently.

Nothing new so far.

But as I expanded in the "show more" button, something caught my eye.

Google added a new button called "Dive Deeper in AI Mode."

Naturally this got me curious.

AI Mode Is Everywhere Now

AI Mode started out as a search labs feature where you needed to sign-up for access. This eventually changed to a no sign-up feature, then recently became publicly available.

What's new to me this time around is I was signed in to my Google Workspace account. I know this wasn't available before. It was only available on my personal Gmail account where I signed up fornit in search labs.

Next, I checked incognito, and it's still there. This tells me it has now completedly rolled out.

Perhaps the only place where I haven't seen it yet is on Google's homepage where other users reported as having a search in AI mode button there.

Nonetheless, Google's AI Mode is here.

And it's everywhere.

AI Mode = Panic Mode

Since I work in SEO, I am both anxious and excited at the same time.

See, when AI mode was introduced early this year, my first thoughts were my industry is going to change. A lot.

This hasn't happened yet but I am seeing shifts in behaviors as well as in the industry.

Perplexity has been my favorite search engine lately. That's why I havent used Google as much.

In a way, I think this will eventually be how AI Mode will become. But that's for another time.

But for the vast majority of people, Google remains the most used tool for search.

According to a recent data I read, Google handles 8-14 billion searches a day (depending on where you read) while the most popular AI tool in the market, chatGPT does ~1B searches.

What I don't know if the numbers in chatGPT are actual searches or just queries/messages.

But here's where my head is at—Google is the #1 search website. They are now introducing AI mode to the public which will significantly reduce the amount of traffic websites get.

Think about it.

If you can chat longer, which is what AI Mode offers, there would be no more need to visit the site, or any website for that matter. Whether they are linked or mentioned, it won't matter anymore.

And that is the first reason and impact on the SEO industry…

AI Mode Will Reduce Users Clicking Through a Website

Looking at my own behavior when I use Perplexity, I rarely go in and click on the sources anymore. In the beginning, yes, I was checking whether the sources were coming from credible ones. But after a while, I stopped.

All I did was refine my questions and/or followup with more prompts.

And now, Google released AI mode which means "regular people" will be doing the same thing. People who aren't in the SEO industry will be using it more and more.

Over time, this will significantly reduce the number of clicks to the websites even though they are cited.

For website owners and businesses, this represents a big shift.

Organic traffic is the single biggest source of "free" leads and customers.

While you can still get visibility, if people aren't reaching your site during that search, it's going to be difficult to convert them.

Traditional SEO metrics such as keyword rankings, click-through rates and organic traffic are all going to be negatively affected.

That means as a business, you need to find other ways to reach your audience.

And for many small businesses, venturing into other channels like paid or social media requires an entirely different approach.

Side note: this is already true to some extent. Just search for "zero click searches" and it is becoming more and more prominent.

This now leads us to the second impact on the SEO industry…

SEOs Will Need to Focus on SEO, Not SEO

I couldn't resist, I had to make a play on words.

SEOs (people in the industry) needs to start thinking more broadly. We need to start thinking about Search Everywhere Optimization instead of Search Engine Optimization.

This has been a shift that is becoming more clear over time—you need to be present everywhere in order to succeed as a business today.

Google has emphasized it's EEAT guidelines (formerly just EAT) over and over:

Experience: First-hand, real-world experience with the topic.
Expertise: Knowledge, skills, and credentials in the subject area.
Authoritativeness: Recognized authority or reputation in the field.
Trustworthiness: Reliability and integrity of the content and its source.

This makes it extremely difficult for agencies and third parties to do at scale. They'd need industry specializatiom in order to really showcase this.

Even with the use of AI, industry knowledge cannot be replicated. You can easily write long-technical content, optimize them for semantic search, and follow all the best practices of SEO—but industry experience and know-how is something only those who work inside the industry or take the time to study it can replicate.

And if you're on the client-side, you can't do this willy-nilly too.

The impact the rollout of AI Mode will have on SEOs means our scope of work needs to expand as well.

We cannot raise our hands in the air and just say, "nope, not my problem."

Remember, as consumer search behavior changes, i.e. no longer clicking through as mentioned above, organic traffic will tank.

Over time, this channel won't be the same way it was.

When there's mass adoption for AI, which is certainly possible because of Google's AI Mode, what SEOs use to work and report on will all change.

As for how long or when that will happen, I don't know.

What I do know is I'm not waiting to be displaced nor hope I won't be affected.

I urge you to do the same.

Content Creation Must Match Content Consumption

A niche industry in SEO is growing—it's called Generative Engine Optimization (GEO). Others call is LLMO (large language models optimization). But whatever you want to use, this new industry operates on the fact that AIs or LLMs use more conversational copy as opposed to keyword-rich searches in traditional SEO.

This is actually one of the most actionable tasks you can do today for your website and content.

However, beyond that, the way you optimize and write content needs to match the new behavior that AI Mode (and other AIs) is shaping.

For instance, the Q&A format I just mentioned above is one.

Focusing on real questions people ask, instead of keyword-specific, would work best. This is one of the behaviors that are slowly shifting. AI searches are more complex, often reaching 15 words or more whereas majority of traditional searches are only 3-4 words long.

Formatting is also key. Today, large language models access content in smaller chunks. Tokens, at the moment, remain the most prominent limiting factor. That's why organizing content that can be easily read and understood by AI is also important.

Perhaps the biggest factor I'm foreseeing happening is the emphasis on industry-insider knowledge. This is going back to EEAT guidelines again. The more "brand authority" you can build, the better. With generative AI, it's easy to create content. But only "trustworthy" brands are most often cited by AI.

I believe that building a trustworthy brand is going to be the most crucial aspect for search in the AI era.

As I reflect more on how I use AI, I constantly refine my prompts to find "hidden gems" or nuggets of information I can't find easily.

Because AI Mode (and other AI tools) allow for more follow-ups without losing context or history, the entire search evolves. Think of it as having a dialogue with any expert in any field.

A good analogy I can think of right now is a visit to your doctor. She is explaining your most recent blood results. You ask what this means, then you follow up with more questions based on your medical history and other information such as current symptoms.

This entire "search" is personal and highly relevant only to you.

That's the promise of AI search.

And no, I'm not advocating you should be using AI to figure out your own symptoms and health problems (or legal issues or other topics requiring professional help). What I'm saying is this conversational nature changes the way we interact with content, which in turn, should also change how we write or create them as marketers.

The conversational nature AI searches encourages deeper exploration – users don’t reformulate the searches. They build on previous answers to find edge cases and specific scenarios.

These “hidden gems” represent huge SEO opportunities because:
1. Zero competition: Most content doesn’t address these nuanced scenarios
2. High conversion potential: People asking specific follow-ups are closer to taking action
3. Authority building: Answering complex edge cases establishes expertise

And the only way to execute this is by taking the time to really think through the questions and providing insider knowledge to stand.

AI Mode and the Future of SEO

The public release of AI Mode by Google will affect traditional search engine optimization as we know it. Metrics such as traffic, keyword rankings , and click-through rates will start to become obsolete. This means we need to change the way we approach SEO. We must move beyond a single channel approach and focus more on providing better user experience through Search Everywhere Optimization instead.

The rising popularity of AI have recently led to the growth of Generative Engine Optimization. This represents the start of an evolution as consumer's search behavior continues to change.

As with most things, SEOs need to adapt or risk becoming obsolete.

What do you think? I'm curious what your thoughts are on this topic. Let me know in the comments below.

3+ Years of Pure Digital Notetaking

July 8, 2025July 7, 2025 by Ariel Lim

Taking down notes—whether it's for your own projects or simply trying to remember something for referenve later—is a behavior that won't go away anytime soon.

At work, this can be as simple as writing down a task list, key decisions made in a meeting, or simply brainstorming ideas for a new marketing campaign.

At home, this can be the instructions by your veterinarian on how to care for your pet, a grocery list, or projects you want to get done in the house.

If Your Notes Aren't Personal, You're Doing It Wrong

Notes are written records of information. Whether this lives in a Post-it, or in your iPhone's Notes app, its purpose definition and its purpose doesn't change.

The purpose of any note is as a reference for the future.

Think about it.

You write something down so you don't forget or easily remember in the future.

your passwords (that you probably re-use over and over)
journal entries
next vacation itinerary

Notes are inherently personal.

You write them in a way that makes sense to you (and hopefully to your future self).

The way how and the reason why you choose to write something is a personal decision that can only be described as personal.

A term that closely describes this is resonance.

You heard or saw something that resonated with you, and that is why you wrote it the way you did.

If you are taking minutes or transcripts, which is word for word, you are doing it wrong.

I'd bet that you are probably comlaining that you're not getting any value from it.

And that's because it's not personal.

What you wrote didnmt resonate with you.

So, here's my advice: stop.

Think about what you value. Think about what is important. Ask yourself why multiple times. This self-reflection is key to taking effective notes.

Why I Switched from BuJo to Obsidian

A few years ago, I wrote about my 7 years of Bullet Journalling. Check it out. You'll see my stacknof Moleskine notebooks throughout the years.

But after the pandemic, I completely switched to digital notes. I first transitioned to Agenda, then now to Obsidian.

There are two main reasons:

Remote work
PARA system

Remote Work Permanence

I realized I won't ever go back to an in-office role. Since before COVID-19, I had been working remotely for my freelance business. When the pandemic hit, remote work became the norm for companies around the world.

This shift changed the way organizations operate entirely. While remote work (globalization of talent) was already a thing before, this was mainly as a supplement to shortages.

But the pandemic accelerated this shift and has now become a strategy to attract and retain talent around the world.

That shift, and my decision to never go back to an office, is the turning point.

I then started figuring out how I can improve my notetaking.

This led me to the PARA method.

The PARA Method of Notetaking

The PARA Method, developed by Tiago Forte (not affiliated), is a simple way of thinking about and organizing your notes.

PARA is an acronym that stands for projects, areas, resources, and archives.

At a high level, here's what they mean:

Projects: stuff you're actively working on. have end dates.
Areas: roles and responsibilities you're managing over time. usually don't have an end date.
Resources: topics you're interested in that may be useful in the future.
Archives: completed or inactive items from the other three categories.

Pretty simple, right?

Combined with Getting Things Done (GTD) concepts and task management, you reach an entirely new level of sophistication amd simplicity.

I'll write more about this some other time. Here are some posts I wrote about GTD before:

For now, I invite you to read more about these if you are unfamiliar.

Digital Notes Solves the Discovery Problem of Analog Notes

Have you ever said to yourself, "I knew I wrote this down somewhere."

The biggest problem of analog notes is discovery—finding information you wrote.

Digital notes solved this.

Sure, there are solutions that actually try to bridge this gap.

For instance, Livescribe is one I particularly used before.

You simply have the pen record the audio (e.g. a meeting or a lecture). Then, you take your usual
notes on a special paper/notebook.

Later on, you can sync the pen to the phone/computer and your notes will show up. It will
create an audio and actual notes. Then later on if you want to review something? you tap a phrase or term, then the audio will
hop to when you wrote that word/phrase.

Anyway, that was before. I haven't kept track of the company but it was pretty cool.

But the key point is searching for your notes is the most difficult and persistent problem in analog methods.

Obsidian Solves This Issue and More

I found out about Obsidian during my research and fell in love with it. Sure, it took some tweaking and time getting used to, but it's really worth it.

I'm using it mostly for it's basic functionality.

If you decide to go down this rabbit hole, there's an entire community around personal knowledge sysyems (PKM) and a lot of people really customize their Obsidian.

As for me, I remain practical.

I use it and apply the PARA method.

I have basic templates for meetings and project notes. I add details like agenda, attendees, action items, etc.

It's a combination of administrative details and what resonates with me.

Then I use the basic features of Obsidian such as internal links and search as needed. I won't dive into that but it's a neat feature to help find relationships among your notes.

Private by Default

Perhaps the most crucial benefit that I didn't anticipate is that my notes here are actually mine. They are private by default.

This means that all the notes are stored locally. The company doesn't even know what I put in the app.

Contrast this with other popular notetaking apps where you won't be able to use it without any Internet. This means that your data is actually with them. Of course these companies would say they won't check or snoop or have access into your data and other notes, but I wouldn't count on it.

Notetaking in the AI Era

Perhaps another benefit that I got when I transitioned to Obsidian is that the way you write notes is using markdown.

Markdown is a simple way of styling notes (versus HTML or other markup language).

Why was markdown a benefit?

Because it allows me to write better prompts in AI.

I can also read and understand the output better because that is how it is displayed across the different LLMs.

In addition, it reads far easier when using it as an input (knowledge base, attachment, etc.).

Other formats, for example, docx or gdoc have a ton of formatting on them that may not be seen visually but is present in the code.

Which leads me to another benefit in this AI era—smaller file sizes.

Smaller file sizes mean less tokens used.

Less tokens used mean more context can be added, longer conversations, and overall
improved quality.

That's my current setup right now. I think this will last for a couple more years but we shall see. With the pace at which things are developing,it's really important to start flexible.

How about you? What's your current setup like today? Are you thinking of changing things up soon?

Using AI Beyond Marginal Gains

July 3, 2025 by Ariel Lim

I've been using AI for quite a while now. It has definitely helped a lot in driving efficiencies particularly in the creative aspect of my work. But I recently started asking myself, "is there something more?"

As with many of you have noticed, generative AI tools like ChatGPT, Claude, and Gemini really shines in creative writing:

paraphrasing words or sentences
writing documents, reports, social media posts
creating plans and spreadsheets

If you go a step further than that, you can use these tools for further research and analysis. Personally, I prefer Perplexity for research purposes as the results are more accurate and recent. While the popular tools all now have web seach functionality, the default settings for the average user yields better results in Perplexity.

When it comes to the analysis part, I'd say they are relatively good depending on the model and prompts you use. In my case, I've used them for the following:

analyzing CSV files for keyword research
analyzing CSV and PDFs for reporting purposes (think Google Analytics and Google Search Console data exports)
analyze code and provide fixes /recommendations (think HTML, CSS, python, etc.)

Then there are your advanced users. These are people who developed complex prompts and delved (yes I used this hated word) in developing advanced system prompts or custom GPTs or AI agents. With this advanced use, it's easy to see 10-20% improvements in work/task completions.

I consider myself to be in this group.

But I've been wondering– is there more?

50+% Efficiency Gains

When I started researching about this, I had a belief of gaining over 90+% gains. But based on what I've been reading, that is mostly refuted.

What remains though are the unicorns that truly transform their work and business, generating over 50+% improvements across the organization.

I think that is where we are still heading. That for me is the allure of agentic AI.

But as that relies complicated setups and not readily accessible to the regular user, the question remains–how can we use AI today to transform (my work/my business/my organization)?

According to McKinsey, the promise of 50+% efficiency gains lies in a strategic approach, rather than a functional one.

Organizations that fundamentally rethink business processes rather than adding AI layers to existing workflows achieve dramatically superior outcomes.

And that's where business leaders across all levels need to be aligned.

That's where managers need to learn how to manage and how to implement change initiatives.

The real culture needs to be addressed before this transformation can happen.

Sidenote: culture is often portrayed superficially as perks or disguised under the words or adjectives the C-Suite often says publicly. But real, deep culture goes beyond that. It's the underlying system of shared values, beliefs, attitudes, behaviors, and norms that shape how work is done and how people interact within an organization.

As you can see, no matter how much the CEO proclaims the culture they have or their AI initiatives, if this breaks down somewhere in the org chart, this transformation can at worst, easily become out of reach, or at best, delayed.

More to Explore

These are just some thoughts running through my head right now. Obviously there are a LOT of things I don't know or even grasp about genAI. And with the speed of development, it's hard to keep up.

One of the things I'm experimenting on is agentic AI, MCP servers and all those autonomous use cases for AI.

I'm not even diving deep into special AI use cases like image creation (dall-e, midjourney), videos (veo, invideo), voice / audio (descript, elevenlabs), and many more.

There's simply no time to learn and test all these tools.

Gen AI Is Creating a New Problem

Underneath all this generative AI hype, a new problem is emerging.

This problem is the oversupply of content and limited availability to filter and consume them.

Think about it. With the ease of creation, one can create thousands of pages in just a few hours.

Couple this with vibe coding, you can create multiple websites or social media posts and videos easily.

And I'm mot even considering the cybersecurity problem all these present.

This is what Google is actually trying to prevent–focusing on helpful content and their EEAT guidelines to filter out generic and regurgitated content.

But even then, that can be easily bypassed.

So what do you do now as the "end-consumer" of these content?

How do you filter out the noise?
How do you find real credible guides or brands to follow?
Who do you trust?

I don't have the answers.

All I know is that the way I consume content and the way I do my work have significantly change.

And I'm sure this is happening across the entire world.

Which means businesses need to adapt amd change the way they do things as well.