A close friend said it to me years ago, almost in passing, as if it weren't a big deal. «Opportunities don't disappear... Someone else takes them.» And ever since, I haven't been able to shake it. Because I've seen it play out more times than I'd like to admit. In projects I didn't start on time. In decisions I postponed because they weren't urgent.
The phrase sounds simple. Almost obvious. But it has an edge that stings when you live it firsthand.
The problem with «later»
We live in permanent urgent mode. The inbox, the 10 AM meeting, the month-end close, the client who needs something yesterday. All of that is real, and I'm not going to minimize it. But there's a side effect we rarely name: when everything is urgent, nothing important that doesn't hurt today ever gets done.
And problems that don't hurt today have a very annoying habit: they show up anyway. In a year, in two, in ten. They show up.
The difference between the person who acted and the one who waited isn't that one had better information. They almost always had the same. The difference is that one decided the future problem was real enough to address today.
The other decided their future self would handle it.
And it works. Until it doesn't.
What's coming with AI isn't a prediction — it's a straight line
I'd bet that within two years, a huge number of companies are going to be asking how to show up better in AI. Not as a technical curiosity. As a CEO demand in the monthly board meeting. As a client question asking why the competition is being recommended in ChatGPT and they're not.
And when that moment arrives, agencies will promise to deliver it even if they don't know how. Because that's how markets work when demand explodes before serious supply exists.
It will be hard. There will be a lot of competition. Prices will go up. Wait times too. And companies that start building AI visibility at that moment will be playing with a two-year disadvantage against those who already have that ground covered.
Why two years? Because AI visibility, just like traditional SEO, isn't something you buy in a month. It's built. The authority that language models like ChatGPT, Perplexity, or Google AI Overviews recognize comes from accumulated signals: structured content, consistent data across multiple sources, mentions on reference sites, well-applied semantic schema. None of that appears overnight.
And none of it can be compressed with budget urgency once the CEO is already demanding it.
The advantage you can't buy later
The authority that language models assign to a brand is cumulative. There's no shortcut. Whoever starts earlier arrives earlier, and whoever arrives earlier defines the space. By the time the others show up, they'll have to fight for territory someone else already claimed.
What this looks like in practice
Let me be specific, because this isn't abstract. Think about any B2B product or service category: tax consulting, last-mile logistics, payroll software, corporate clinics. Today, if you ask ChatGPT or Perplexity who's reliable in that category, you'll probably get three or four names. Some will be the obvious market leaders. But several will be companies that simply did their digital homework before everyone else.
Not necessarily the biggest. Not necessarily the oldest. The ones with structured content, consistent sources, documented presence across multiple corners of the web, and a brand narrative that language models can read, process, and cite.
That's what we call AI visibility — the umbrella that covers GEO (Generative Engine Optimization), AEO (Answer Engine Optimization), and the technical SEO layer that makes both possible. And the difference between being there and not being there isn't about budget right now. It's about when you decided to start.
There are companies with perfectly Google-optimized websites that still don't show up when someone asks an AI assistant about their category. Not because they're bad companies. But because the work of becoming visible to language models is different from the work of becoming visible to traditional search algorithms — and it has to be done deliberately.
The opinion nobody asked for:
The underlying problem isn't technical. It's a mindset about time.
Most companies make marketing and positioning decisions when the pain has already arrived. When traffic dropped. When the competitor is already ranking above them. When the director is already asking why they don't appear. And at that point they start with the urgency of someone putting out a fire, not the calm of someone building something.
SEO took years for companies to take seriously. There was a moment — maybe between 2010 and 2015 — where the companies that bet on organic positioning while others said «what for, nobody searches on Google» built advantages that are still hard to replicate today.
We're at that same moment, but with AI. And it doesn't last forever.
When the market mass-produces demand for «I want to show up in ChatGPT,» the cost of doing it will rise, timelines will stretch, and competition will make the work harder. Not impossible. Just more expensive and slower.
The window exists now. Not because it's a trend. But because market demand hasn't yet caught up with the supply of companies actually doing it well.
Three things that determine whether AI recommends you or not
Without running a full course on GEO and AEO, there are three levers that carry the most weight in whether language models include a company in their responses:
Entity Authority
Language models build a «picture» of your company from what they find on the web: your site, mentions in media, profiles in directories, reviews, structured data. If that picture is consistent, coherent, and documented across multiple trustworthy sources, the models know you and cite you. If your digital presence is fragmented or sparse, you simply don't exist to them — regardless of how many years you've been in the market.
Content that answers specific questions
AI assistants are built to answer questions. If your content isn't written to answer the questions your potential clients actually ask, the models have nothing to cite from you. AEO (Answer Engine Optimization) is the practice of structuring content exactly for that: direct answers, natural language, no filler.
Technical signals that AI crawlers can read
Most AI assistants crawl the web with their own bots. If your site blocks those crawlers, misdirects them, or has a technical structure that confuses them, you become invisible — even if you have the best content in the market. Schema markup, structured data, a well-configured llms.txt: the boring details that actually make the difference.
Our SEO, GEO, and AEO service works across exactly these three layers — not as separate services but as an integrated system. Because all three are needed for it to work.
Before investing in a service, verify where you stand today
There's something you can do right now, free and with no setup required: the AI visibility test at aeon42.com/test-ia. You leave your company name, industry, and main product. The system asks an AI assistant the question a real client would ask when looking for what you offer — without mentioning your name. You get the result on screen and by email: a score, what the AI knows about you, who it recommends instead of you, and three concrete actions.
It's a snapshot of one moment, not continuous monitoring. But it's enough to know whether you have a problem or not. And if you do, to understand how deep it goes.
If after that you want ongoing tracking, historical metrics, and actionable recommendations week over week, AEON42 is the tool we built for that: it connects your Google Search Console, analyzes 16 months of real data, and measures your AI visibility every week: Google AI Overviews and AI Mode from USD $49 per month; ChatGPT, Gemini, Perplexity and Claude on the Studio and Agency plans. With a 14-day trial.
And if you'd rather have someone do it for you
There are two ways to approach this. The first is the tool: you use it, you decide what to implement, you set the pace. The second is the team: we run the diagnostic, design the strategy, implement the changes, and measure results.
Play.Interactive has over 26 years of digital project experience for clients in Mexico, Latin America, the United States, and Asia. The AI visibility work we do today isn't fundamentally different from the SEO work we've done since Google existed: build authority, create content that answers real questions, make sure the technical infrastructure works. What changed is who we need to be visible to.
Before, it was for Google's algorithm. Today it's also for language models.
And tomorrow there will be more.
The most expensive decision is usually the one you postpone
Back to my friend's line. Opportunities don't disappear. Someone else takes them.
In two years, when a marketing director at some company in your sector demands their agency put them in ChatGPT, they're going to discover that three or four competitors have been building that work for a while. That they already have Entity Authority established. That they're already the ones AI cites when someone asks about what they do.
Those companies won't necessarily be the biggest or the most well-known. They'll be the ones who decided the future problem was real enough not to leave it for their future selves.
If that's you, the first step is knowing where you stand. Subscribe to the newsletter at /en/subscribe and next week you'll get access to the SEO/GEO/AEO diagnostic: 12 questions, three blocks, five minutes.
Or start today with the free test at aeon42.com/test-ia. No setup, no access to grant, result in one minute.
Frequently asked questions
AI visibility is the set of practices that determine whether your company gets recommended by assistants like ChatGPT, Perplexity, or Google AI Overviews. Traditional SEO optimizes for you to appear in search results when someone clicks. AI visibility aims to have language models recognize you as a trustworthy source and cite you in their responses — without the user needing to click anything. They're complementary, not mutually exclusive, but they require different work.
The timeline is similar to competitive SEO: technical improvements can show up in 4 to 8 weeks, but building the authority that language models recognize takes 3 to 6 months of consistent work, depending on your existing domain authority. There's no shortcut. That's precisely why it makes sense to start before the pressure arrives.
They're not sequential steps, but they are interdependent. A technically solid site with structured content and domain authority accelerates AI visibility work. If your SEO is neglected, addressing that while simultaneously working on GEO and AEO is the most efficient path. Going layer by layer when you could work all three together just slows down the result.
Yes. The free test at aeon42.com/test-ia does exactly that in one minute: you leave your company name, industry, and main product, and the system queries an AI assistant the way a real client would — without mentioning your name. You get the result on screen and by email: a score, what the AI knows about you, who it recommends instead, and three concrete actions. It's a snapshot of a single moment, not continuous monitoring.
AEON42 works well for those who have the internal capacity to read the data and execute the changes: it connects your Google Search Console, analyzes 16 months of data, and delivers actionable recommendations week over week, from USD $49 per month. Play.Interactive's managed service makes sense when you want someone to design the strategy, implement the technical and content changes, and deliver results without you having to coordinate anything. One doesn't replace the other — they're two different speeds for two different realities.