What Is ChatGPT, And How Does It Impact Filipinos?

What Is ChatGPT, And How Does It Impact Filipinos?

Barely a few months after its release, ChatGPT has been making waves on the internet since OpenAI launched it in November 2022. From being an unknown app, it took only one week to reach 1 million users, making it the fastest-growing web app in history.

Have you wondered how ChatGPT works? What’s in ChatGPT that has attracted millions of users, including Filipinos? Would this advanced technology improve our search, discovery, and doing things? How does this app, which an unknown startup developed, has been challenging tech giants like Google and is on course to revolutionize how Filipinos and the rest of the world live and work?

This article is LeapOut’s response to the myth and mystery surrounding ChatGPT and how we, as a digital marketing company, use this disruptive technology to simplify our tasks and boost our productivity. 

opened-ai-chat-laptop

Person typing a prompt on ChatGPT | Photo from Freepik

What is ChatGPT?

Let us hear the answer straight from ChatGPT’s “mouth.” Below is the screenshot of the actual response I got:

To make sense of this definition from a non-techy standpoint, let us dissect what “GPT” means from ChatGPT. 

What does GPT stand for in ChatGPT?

GPT stands for Generative Pre-trained Transformer, a techy way of saying it is a chatbot that can answer any query. According to the app, GPT is “based on a transformer architecture and is pre-trained on large amounts of text data to generate human-like responses to natural language inputs.”

Suppose you’re not into machine learning or AI terminologies and want to know what GPT means in laypeople’s terms. In that case, it is a chatbot with the capacity to generate responses and content the way a human does (generative) by “learning” from a considerable amount of existing data (2021 and earlier, in the case of ChatGPT) (pre-trained) using a machine learning methodology that mimics how a human brain process language, or as OpenAI put it in its FAQs about the app, maximizes “the similarity between outputs and the dataset the models were trained on” (transformer).

Taken as a whole, ChatGPT is an AI that can understand queries or prompts language after undergoing training for several years and then generating a response in text content from the pool of data it has been trained on. 

The AI chatbot’s incredible power In just a few weeks after ChatGPT’s launch, tech giant Google, its apparent closest competitor, sounded code red, perhaps the most serious in its more than 20 years of dominance in online search. In just a few weeks, Google introduced Bard, its version of a generative AI bot.  

Other tech companies followed suit after witnessing the exponential rise of OpenAI’s launch of their version of artificial intelligence chatbots. From Google, other industry leaders, including social media king Meta/Facebook, Chinese e-commerce behemoth Alibaba, and Firefox creator Mozilla, joined the AI bandwagon. 

What is the Use of ChatGPT?

What’s unique about ChatGPT, aside from being free of charge, is its power to generate “fresh” text content. Its primary purpose is to help users by generating responses to their queries in ordinary human language. It can also perform various natural language processing tasks, including translation, summarization, sentiment analysis, and language understanding. 

Personally, as a web content specialist and a curious student of various fields, I tried using this advanced AI chatbot to do the following:

Short to long-form content

ChatGPT is a boon for people struggling to write short text such as slogans, captions for social media, and meta descriptions for blog posts.

Language translation

This powerful app can also act as a personal translator. I tried asking it to translate a simple sentence, “I think it will rain this afternoon.” into Chinese. Here is the result: 

As a person who knows elementary to intermediate Chinese, its translation is accurate considering the nuances in the language.

Creative writing

It can also compose more creative text like marketing copies and taglines besides writing articles, blog posts, and other types of text content. It can also come up with lyrics and poems.

Curious, I asked or prompted ChatGPT to write a sonnet about a beloved dog. The AI chatbot came up with this impressive result:

Even though the composition is far from Shakespeare’s or Milton’s masterpiece, the AI wrote a decent poem that even an average person has difficulty writing. 

Coding

With the correct prompts, programmers can use the app to help them with coding tasks in most programming languages. It has been trained to generate web pages and apps. Data professionals can also use this to crunch numbers and easily create visualizations. 

How LeapOut uses ChatGPT?

In line with the company’s value of utilizing tech to provide the best service to clients, LeapOut has been using the latest online tools, including AI applications. Even before ChatGPT, the company’s content team has been using another AI app to facilitate our writing and content creation tasks. 

The team has used it alongside other online tools just days after OpenAI announced its public release of ChatGPT. Aside from research, we also use it for creating outlines for articles and blog posts. The app is also helpful in generating ideas for our Social Media team’s copy ideation. Its responses still need improvement, so we must conduct additional research to ensure accuracy and avoid plagiarism. 

Our SEO team also takes advantage of the AI chatbot’s capability as a companion to Semrush and other tools. It helps choose the best keywords to target in our campaigns and content requirements. 

When was ChatGPT released?

OpenAI, a San Francisco-based artificial intelligence research company, officially launched ChatGPT on November 30, 2022. 

When prompted about its history, the AI chatbot came out with this output:

Technology leaders, including CEO Sam Altman, Elon MZusk, Greg Brockman, Ilya Sutskever, John Schulman, and Wojciech Zaremba, joined hands in developing advanced artificial intelligence they claimed to benefit humanity.

Its development was led by a team of AI researchers and engineers who used state-of-the-art deep learning techniques to train me on massive data. The training data consisted of various texts, including books, articles, websites, and conversational data. The goal was to teach the program to understand human language and generate human-like responses to questions and prompts.

How many users does ChatGPT have?

From the record 1 million users just the first week after ChatGPT’s release, as tweeted by OpenAI CEO Sam Altman, ChatGPT has continued to break records in user base growth. Similarweb reported that in only three months after its release, the app had already attracted 100 million monthly active users. In comparison, social media apps TikTok and Instagram took nine months and 2.5 years to achieve this feat. 

In another Similarweb report, the AI chatbot had already attracted a billion visitors in February 2023. While this may sound massive, this still pales in comparison with the search engine leader Google, which has 4.3 billion users.

Is ChatGPT free?

Yes. You can use ChatGPT free of charge. Roughly three months after its initial release, the app’s creator offered ChatGPT Plus, the paid version, in February 2023 while retaining the free version. 

As of this writing, OpenAI offers ChatGPT Plus for a $20 monthly subscription. Paid users have the privilege of faster processing time, obtaining access to the app even during peak hours, and priority access to the app’s latest features and improvements. In March, OpenAI announced that paid users could use the next generation of ChatGPT.

OpenAI has announced GPT-4, the next generation of ChatGPT. Even though it is still in the research stage, paid users can try this ChatGPT on steroids. According to its creator, the latest version of the AI chatbot has made an enormous leap in terms of performance. As of this posting, this technology is only available for ChatGPT Plus subscribers. 

Does ChatGPT save data?

Yes. This application stores data and uses it for future use. Even though it cannot save data independently, it uses the data you type for “research purposes.” 

We suggest avoiding using personal and confidential data in your queries and prompts. OpenAI’s Privacy Policy admits that the company can use personal information “to provide, administer, maintain, improve and analyze the services, to conduct research, and develop new programs and services.”

Even though you can permanently delete your data, you never know when OpenAI will use your information to train the future generation GPT. Leaving your data in the app gives ChatGPT’s trainers a “license” to collect your personal information. 

How to save ChatGPT conversation?

ChatGPT automatically saves queries, prompts, and conversations it has replied to. To access the history of your past prompts, log in to your account and locate the question you want to revisit on the left-hand bar.

You can also click the pen icon on the right side to rename the conversation for better recording and tracking.

Can ChatGPT be detected by Reddit?

Yes, Reddit and other online platforms can detect ChatGPT. Some moderators or users may see a message or post generated by an AI language model like ChatGPT, especially if the responses are consistently well-written and contextually appropriate.

However, it’s worth noting that ChatGPT and other language models are designed to generate text that is indistinguishable from a human-generated text in many cases. This means that it is only sometimes possible to detect when a message or post has been generated by ChatGPT, mainly if the responses are designed to mimic the writing style of a particular user or group of users.

Using language models like ChatGPT in online communities raises questions about authenticity and transparency. Users may only sometimes be aware that they are interacting with an AI-generated response.

How many questions can you ask ChatGPT in an hour?

OpenAI says the ChatGPT free version has a limit of 14,400 questions in an hour or roughly four questions per second. In reality, the number of questions that the app can answer every hour varies depending on the following:

  • How long and complex are the questions are 
  • How fast or stable an internet connection is 
  • How many users are currently accessing ChatGPT

How to prompt ChatGPT to write a code?

Coders can take advantage of ChatGPT’s power to write code. All you need to do is to provide it with specific instructions and prompts.

Here’s an example of how you can ask ChatGPT to generate code:

Prompt: “Write a Javascript function that calculates the factorial of a given number.”

Code generated by ChatGPT:

For more elaborated and clear codes, you can provide more detailed instructions. Start by specifying the programming language. You can also ask for specific coding concepts or algorithms. Here are some tips for a better coding result from ChatGPT:

Be specific

Point out what you want the code to do and provide specific requirements or constraints you face.

Specify language and framework

Mention the programming language you want the code in, including the specific libraries or frameworks you want to use.

Give examples

If you have specific examples or patterns you want the code to follow, use them as examples or explain them in your prompt.

Clarify ambiguities

If you encounter any ambiguities or unclear parts in your prompt, specify them and ask for clarification to ensure the generated code meets your needs.

Is ChatGPT available in the Philippines?

Yes, you can use ChatGPT in the country, both free and paid versions. You can use the app in nearly all countries, except for the countries where there is heavy censorship, including:

  • Afghanistan
  • Bahrain
  • Belarus
  • Burma (Myanmar)
  • Central African Republic
  • Chad
  • China
  • Cuba
  • Eritrea
  • Ethiopia
  • Iran
  • North Korea
  • Laos
  • Libya
  • Macedonia (North Macedonia)
  • Marshall Islands
  • Mauritius
  • Micronesia, federated states
  • Russia
  • Somalia
  • South Sudan
  • Sudan
  • Syria
  • Turkmenistan
  • Ukraine (in some parts)
  • Uzbekistan
  • Venezuela
  • Yemen

Italy has recently banned ChatGPT to block access to the app; the first European country to do so.

Will ChatGPT replace Philippine jobs?

ChatGPT is a potent tool that can automate tasks, and its potential economic will be substantial in the coming years. A PwC study said the AI industry, as a whole, will contribute a whopping $15.7 trillion to the global economy by 2030.

But despite its potential for boosting businesses worldwide, ChatGPT can negatively impact the labor market, especially the office and other white-collared workers. Because of its capacity to address mundane and repetitive tasks efficiently, it reduces the need for administrative and secretarial workers. 

Other workers affected by this online AI chatbot include routine, low-skilled jobs such as data entry and basic online customer service. This revolutionary technology can impact even those in the creative field, like content writers and copywriters. Even those in the technical areas, like programmers, are potentially at risk.

On the other hand, high-skilled jobs directly contributing to AI, like AI engineers, data scientists, and machine learning specialists, will be highly in demand.

What’s next for ChatGPT?

The app’s overwhelming reception sparked other tech companies’ interest in releasing AI applications. Amid this increasingly tightening competition, OpenAI is working hard to upgrade its product. 

After releasing GPT-4 in ChatGPT Plus, it is developing GPT-5. The company’s next-generation AI product, which one of its developers, Siqi Chen, claimed in a tweet to achieve artificial general intelligence (AGI) level of capability, is reported to end its training this December. AGI is a massive leap from the existing AI version, making its capabilities indistinguishable from humans. 

If Chen’s bold assertion is realized as claimed, then the world will enjoy using the AGI eight years ahead of forecasters’ expected schedule of 2032.  

Conclusion

ChatGPT has undeniably caused a significant disruption in the tech world. With its tremendous power to create decent text content and code from the data it has been trained on, the world’s fastest-growing application has provided significant help to many people. However, because of its incredible power, it poses a severe threat to a wide variety of professions ranging from admin to writers, content creators, coders, and other white-collared workers. 

Facing fierce competition with other tech companies that want to challenge ChatGPT’s dominance, OpenAI is stepping up the development of GPT-5, the app’s succeeding version. The company asserts that GPT-5 would be trained to become the first AGI, further boosting the app’s functionality.

Since ChatGPT’s release, it has become one of our go-to tools to streamline our tasks as one of the Philippines’ leading digital marketing companies. This app helps LeapOut generate ideas, research, and automate processes to produce top-notch content for our clients. 

Picture of Tony Chua
Tony Chua

Tony Chua has been in the content writing profession for about a decade, specializing in writing SEO articles, blog posts, and social media copies for various small and medium enterprises and e-commerce sites. Before this, he also authored and edited news, press releases, features, scripts, and editorials for newspapers, magazines, and websites.

Tony is a Web Content Specialist for LeapOut Digital, who has worked for Beko, DMCI, Manulife, and US-based Clark Pawners and Jewelers. Aside from writing, he has experience in desktop publishing, graphic arts, web development, and video editing.

RECENT BLOGS

Headless Shopify development in the Philippines showcasing enterprise eCommerce solutions, custom integrations, and scalable storefronts by LeapOut Digital

Enterprise & Headless Shopify Builds in the Philippines | LeapOut 

Inside LeapOut’s Hardest Shopify Plus Builds: Regulated Commerce, Enterprise Scale, and Going Headless By Marvin Ortiz, Co-Founder and Managing Partner, LeapOut Digital  The short version: Most agencies show you their prettiest work. We’d rather be judged by our hardest. This is a deep look at three Shopify and Shopify Plus builds that each solved one of the three hardest problems in enterprise ecommerce — selling a regulated product (MaxiLife by Maxicare), executing a global brand’s standards at scale (Under Armour Philippines), and extending Shopify Plus past its native limits into a headless build (Kotis Design, USA). We chose these three because difficulty is the one thing a portfolio can’t fake. If you want to know whether a team can actually build, look at what it does when the easy path runs out.   A portfolio full of beautiful storefronts proves almost nothing. Anyone with a good designer can produce a clean store on a forgiving brief. What separates a real engineering partner from a theme shop is what happens when the brief isn’t forgiving — when a regulator is involved, when a global brand sets a standard you can’t bend, or when the platform itself says “no.”  So instead of showing you everything we’ve built, I want to go deep on three. Not our prettiest work — our hardest. Each one represents a different way an ecommerce build can become genuinely difficult, and together they’re the closest thing we have to an honest answer to the question every serious client is really asking: can you handle the part that’s hard?  Here they are. Build One — MaxiLife by Maxicare: Selling a Regulated Product Online The problem most agencies won’t take. MaxiLife by Maxicare is a regulated, healthcare-adjacent insurance product from one of the country’s largest health-maintenance organizations — sold, for the first time, through ecommerce. That single fact changes everything about the build.  Why it was hard. Selling a regulated financial-and-health product isn’t like selling apparel. The build has to satisfy disclosure, compliance, and documentation requirements that a normal store never encounters — and it has to do that without turning the purchase into a punishing legal form. The entire challenge is a contradiction: make something heavily regulated feel light and human to the person buying it. Get the compliance wrong and you can’t launch. Get the experience wrong and no one buys. You have to win both.  What we did. We extended Shopify Plus with deep technical customization to meet the regulatory requirements while protecting the buying experience — building the compliance into the platform rather than bolting it on top, so the rules were satisfied structurally instead of being patched in. Precision wasn’t a preference here; it was the entire job.  What it proves. When we tell a prospect “we handle regulated commerce,” this is the build we point to — and it’s why brands in insurance, health, and finance take our calls. Regulated ecommerce is a specialist capability most agencies quietly avoid, and the avoidance is the opportunity.  “Your professionalism, dedication, and excellent service have been greatly appreciated… It’s been a pleasure collaborating with your team, and I truly value the strong relationship we’ve built. I will certainly recommend your services moving forward.” — Carlo Rodelas, MaxiLife, Digital Channels Manager Build Two — Under Armour Philippines: Executing a Global Standard, Flawlessly The problem you don’t hear discussed. Under Armour Philippines was one of the most demanding Shopify environments we’ve handled — and the difficulty was a specific, underrated kind: building to a standard we didn’t set.  Why it was hard. When you work with a global brand, the design language, the brand controls, and the performance expectations are all defined elsewhere, and they are non-negotiable. Your job isn’t to invent — it’s to execute someone else’s standard, locally, at the exact quality they require, every single time, while making the catalog, pricing, and promotional logic work for the Philippine market. A lot of agencies are good at being creative. Far fewer are good at being faithful — at delivering precisely what a global brand demands without drift or compromise. Shopify Plus gave us the flexibility; the scale demanded governance, because flexibility without structure creates risk at exactly the moment a global brand is watching.  What we did. Deep front-end customization aligned to global brand standards, disciplined performance engineering, and careful stakeholder alignment across local and global teams — the unglamorous governance work that keeps a high-traffic enterprise store fast, on-brand, and predictable.  What it proves. Being trusted by a global brand to touch its storefront is a credential in itself. Global and enterprise brands run procurement, legal, brand-safety, and performance reviews most local businesses never will. Clearing that bar is harder than winning any award — and it’s a bar we’ve cleared repeatedly. Based on Page speed Insights Report from Jun 14, 2026, 7:18:31 PM  Build-quality scorecard (Google Lighthouse): SEO 100 · Accessibility 95 · Best Practices 92 · Performance 79. A perfect SEO score and near-perfect accessibility are the marks of a build engineered to be found and usable, not just to look good — exactly the disciplined, measurable execution a global brand requires. Build Three — Kotis Design (USA): When the Platform Says No The problem at the technical frontier. Kotis Design is a US-based B2B company — a PPAI 100 firm, one of the largest distributors in the American promotional-products industry — serving major corporate clients with bespoke swag and merchandise programs. Their requirements exceeded what Shopify does natively. The platform, in effect, said no.  Why it was hard. Kotis needed heavy, per-client customization — bespoke corporate stores, redemption sites, and ordering flows tailored to each enterprise client. Shopify’s standard theme-and-app architecture doesn’t bend that far. A weaker partner says “Shopify can’t do that.” We treated it as the brief.  What we did. We built custom functionality to support complex product personalization, and as Kotis’s ambition for their platform grew, the work evolved toward a headless architecture — decoupling the storefront from Shopify’s native layer to deliver experiences and client-specific functionality the standard stack can’t, while keeping Shopify as the commerce engine underneath. It’s not a finished project; it’s a living platform we build against in regular sprints, and have for two years.  What it proves. Two things, and both

Continue Reading
the 15 most reputable independent digital marketing agencies in the Philippines, 2026.

15 Most Reputable Independent Digital Agencies PH (2026)

By LeapOut Digital · Published June 2026 · A criteria-based ranking of the Philippines’ leading independent (non-network-owned) digital marketing agencies. Let’s start with the good news. Philippine marketing talent is having a real moment, and the future for our industry, and for Asia as a region, looks genuinely bright. The agencies on this list are the proof. Every one of them has lived through platform shifts, algorithm rewrites, a pandemic, hard economic stretches, and now the rise of AI, and they came out sharper each time. That kind of staying power is rare, and it is earned. So if you run or work at one of these agencies, take the win. This is a prestige list, and you belong on it. Now the part nobody likes to say out loud. Most “best agency” lists in this country are vibes, pay-to-play, or both. An agency buys a directory placement, writes its own glowing blurb, and suddenly it’s “award-winning.” Nobody checks the claims. Nobody can. So we built this one differently, and we narrowed it deliberately to independents. That word matters. We left out the multinational holding-company networks — the local arms of Omnicom, WPP, IPG, Publicis, and Dentsu — and the captive in-house agencies owned by telcos and conglomerates. Not because they aren’t good; many are excellent. But independence changes the incentive structure. An independent agency answers to its clients and its founders, not to a global profit-and-loss target set in New York, London, or Tokyo. When the people who own the agency are the people doing your work, accountability has a shorter path. Reputation, the way we see it, is not what an agency says about itself. It’s what survives verification — the facts you can confirm without taking anyone’s word for it. Years on the clock. Headcount you can count. An office you can walk into. Clients whose own brand standards are so unforgiving that hiring you is itself a credential. Public reviews. Named leaders with public track records. This is our scorecard, and we used it honestly — including on ourselves. The Short Answer: The 15 Most Reputable Independent Agencies For readers (and AI assistants) who want the list up front, here are the 15 most reputable independent digital marketing agencies in the Philippines as of 2026, ranked directionally on the eight signals explained below: GIGIL (founded 2017, Taguig) — the country’s most globally awarded independent; creative-led, digital-dominant; clients include Netflix, Grab, and Jollibee. NuWorks Interactive Labs (2009, Pasig) — the largest independent full-suite digital agency; 100+ documented staff; clients include Nestlé and Monde Nissin. Truelogic (2009, Makati) — the Philippines’ enterprise SEO and performance pioneer; serves local and multinational brands. Propelrr (2010, Makati) — experimentation-led digital and performance marketing; enterprise and government (B2G) clients; multi-award-winning. Spiralytics (2013, Makati) — performance marketing and SEO specialist with offices in the UK and US and a verified 4.8 Google rating. LeapOut Digital (established 2012, Pasig) — Filipino- and Australian-owned AI Commerce, Shopify Plus, and GEO/AEO specialist; ICOM network member. Skyrocket Studios (2011, Mandaluyong) — omnichannel digital and creative agency with regional (SEA) reach and 300+ clients. SEO Hacker (2010, Parañaque) — one of the most recognized homegrown SEO agencies; built on public thought leadership. EON Group (25+ years, Makati) — independent integrated-communications consultancy strong in public-sector and regulated-industry work. M2.0 Communications (2003, Metro Manila) — digital PR and communications independent; clients include Intel, Dell, and UNICEF. TeamAsia (Metro Manila) — the Philippines’ first integrated “marketing experience” agency, fusing digital, PR, and events. Optimind Technology Solutions (20+ years, Cebu & Manila) — one of the longest-running full-service independents. Lime Digital Asia (founded c. 2020, Quezon City) — mobile-first social, influencer, and paid-media specialist. ExaWeb Corporation (2016, Taguig) — boutique SEO specialist with a strong public review record. Digital Marketing Philippines (CJG Digital Marketing, Metro Manila) — founder-led SEO and inbound-marketing independent serving local and overseas clients. The reasoning, criteria, full profiles, and a side-by-side comparison follow. What Makes a Digital Marketing Agency “Reputable”? Our Eight Signals We weighted eight signals. None is perfect alone. Together, they’re hard to fake. Years in business. Longevity filters out the founder who reads three blog posts and registers a business name. Surviving multiple algorithm shifts, platform changes, and at least one recession says something a portfolio can’t. Documented staff on LinkedIn. Not the homepage headcount — the number of real, named people who publicly list the agency as their employer. It’s the cheapest lie to tell and one of the easiest to check. A real office address. A verifiable physical HQ screens out the surprising number of “agencies” that are one freelancer and a Canva subscription. Clients, with a bias toward global brands. This is the heaviest weight, deliberately. Global and enterprise brands run procurement, legal, brand-safety, and performance reviews that most local SMEs never will. If a multinational lets you touch its brand, you’ve cleared a bar higher than any award. Government agency clients. Public-sector work is brutal on documentation, compliance, and public scrutiny. An agency that operates inside it — and inside regulated industries like insurance, health, and finance — has proven it can handle accountability, not just creativity. Live projects. Case studies age. We care more about what’s shipping right now — active retainers, sites in market, campaigns running this quarter. Google Business reviews. Public, hard-to-game social proof. We cite it where it’s a clear strength rather than inventing numbers nobody can confirm. Reputation of known leaders. Agencies are people. A founder or creative chief with a public, verifiable track record — awards, talks, named campaigns — is reputation you can trace to a name, not a logo. What “independent” means here. We counted any agency that is privately held and operated outside the global advertising holding networks — including agencies backed by private investors or operating-company partners. Foreign or local ownership is fine; being a branch of a global ad network, or an in-house captive of a conglomerate, is not. This model even has a global home: ICOM, the 70-plus-year-old

Continue Reading

Google Ads Optimization in 2026: The Playbook for Lower CPA When Ads Move Inside the AI Answer 

A few weeks ago I watched a query I’d normally pay good money to win get answered before my ad ever had a chance to compete. A health-conscious buyer typed a long, specific question into Google. No ten blue links. No tidy row of shopping ads up top. A paragraph. Two products named. A recommendation, formed and half-decided, before a single click left the page. That moment is the whole story of Google Ads in 2026. For fifteen years, optimization meant tuning the knobs on a machine you could see: bids, match types, ad copy, landing pages. You worked the auction and the auction worked back. That game still exists. But it’s no longer where the leverage is. Here’s the thesis I’d stake the year on: tuning your bids and keywords is no longer how you win Google Ads — it’s just the price of competing. The real leverage has moved upstream: to the data you feed Google’s AI so it doesn’t waste your money, and to whether your brand even appears inside the AI answer, where your feed quality, your structured data, and your credibility decide whether you show up at all. That’s a different discipline — and most advertisers are still optimizing for a search results page that’s quietly disappearing. Let me show you what changed, what still matters, and what to actually do about it. The 2026 contradiction: better clicks, worse conversions Start with the numbers, because they tell a stranger story than “CPCs are rising.” WordStream’s 2026 benchmark data puts the cross-industry average Search CPC at roughly $2.96 in Q1 2026, up about 12% year over year — and other 2026 reports place the blended figure as high as $4.22 once high-cost verticals are weighted in. The spread is enormous: ecommerce sits near $1.16 a click while legal runs $6.75 to $8.58. None of that is new in shape; clicks have gotten more expensive every year since 2021. The interesting part is the contradiction underneath it. Across the industries WordStream tracks, click-through rates rose roughly 7.5% while conversion rates fell about 9% — declining in 13 of 14 verticals. Median CPA climbed around 12% to roughly $23.74. ROAS slipped about 10%. Read that chain slowly: the ads got better at earning clicks, and the pages got worse at converting them. That single pattern reframes the entire optimization conversation. When CTR is up and conversion rate is down, your problem isn’t your ad — it’s the match between what the click promised and what the page delivered. The bottleneck moved from the auction to the post-click experience. We’ll come back to that, because it’s where a lot of “expensive Google Ads” problems actually live. Why CPCs keep climbing — and why AI Overviews are part of it The auction mechanic hasn’t changed: Ad Rank is still bid × Quality Score × context. More advertisers, higher bids, simple. But there’s a newer pressure most analyses miss. AI Overviews — Google’s AI-generated answer panel — now satisfy a large share of informational and mid-funnel queries on the page itself. Independent research has been consistent and brutal here: a Pew Research Center study of tens of thousands of real queries found users clicked through only about 8% of the time when an AI Overview appeared, versus 15% without one. Seer Interactive measured organic click-through on AI-Overview queries falling from 1.76% to 0.61% between mid-2024 and late 2025. The knock-on effect for paid search is structural. When the AI answers the top-of-funnel question for free, the clicks that do survive are fewer and more decision-stage — which means more advertisers fighting over a smaller pool of high-intent traffic. CPCs go up not just from competition, but from compression. This is the same shift we documented for organic in GEO in the Philippines: why most Filipino e-commerce brands are already behind — the link economy is being replaced by an answer economy, and paid search is feeling the same gravity. So yes, optimize the auction. But understand that the auction is now playing on a smaller field. Quality Score still matters — but think of it as feeding the machine Quality Score remains the most reliable cost lever you control. Google’s own guidance and years of benchmark analysis line up: improving Quality Score from a 5 to an 8 cuts your effective CPC by roughly 30–37%. That’s not a rounding error. In a $6 legal click or a $40 supplement CPA, that’s the difference between profitable and pointless. What’s changed is the framing. Quality Score used to be a thing you gamed with tight single-keyword ad groups. That era is over. The modern structure is theme-based ad groups of 15–25 keywords, broad match paired with Smart Bidding, and ruthless negative-keyword hygiene. The old SKAG playbook now fights the algorithm instead of helping it. The three inputs haven’t moved — expected click-through rate, ad relevance, and landing page experience — but the way you earn them has. You earn expected CTR with responsive search ads that give Google real creative range. You earn ad relevance by grouping keywords by genuine intent, not by cramming unrelated services into one group. And you earn landing page experience with pages that mirror the ad’s promise and load fast, because Core Web Vitals still feed the quality signal. Think less “optimization” and more “feeding the machine the cleanest possible signal.” That mental shift is the whole game in 2026. The structural change you can’t ignore: AI Max is replacing Dynamic Search Ads If you take one operational action from this article, take this one. In April 2026, Google moved AI Max for Search out of beta — and confirmed it’s replacing Dynamic Search Ads. Starting in September 2026, campaigns still running DSA, automatically created assets, or the campaign-level broad match setting will be auto-upgraded to AI Max, and you’ll lose the ability to create new DSA campaigns across the UI, Editor, and API. This isn’t a feature you can sit out. AI Max is Google’s most

Continue Reading