It has never been easier to ship a website. Describe it to an AI tool in the morning and by the afternoon you have a landing page, a pricing table, a signup form and a domain. It works. It even looks decent.
Then nothing happens. No visitors, no signups, no impressions in Search Console. A month later the site is still live and still invisible.
I see this every week, and it's rarely because the product is bad. The AI did its part. It built a website. Nobody asked it to make the website findable, trusted or worth sharing, and those parts were never in the prompt.
The good news is that every one of them is fixable, most of them in a day.
Why AI-built sites look the same and rank nowhere
Open ten sites generated this way and you'll recognize them in a second: a purple gradient, a centered headline that says "Build faster with AI", three feature cards with icons, a testimonial from someone who doesn't exist.
That sameness costs you twice. People skim past it, because it looks like the hundred other pages they saw this week. And search engines have nothing to tell it apart from those pages either.
Under the surface it's often worse. Many of these sites are single-page apps that send the browser an almost empty HTML file and build everything in JavaScript. Google can render that, slowly. Most AI crawlers read only the raw HTML, so to them your site is a title tag that says "Vite + React" and nothing else. No sitemap, no meta descriptions, one URL for the whole product.
None of this is the AI's fault. It's what you get when the only goal is "make it work".
Do SEO from day one, not after launch
Search is the one channel that keeps sending people while you sleep, and it rewards being early. Every week your pages aren't indexed is a week of history you don't get back.
The foundations are short:
- HTML that arrives rendered. Server-side rendering or static generation, so the text is in the page before any JavaScript runs.
- One page per question. Not one landing page for everything. A page for each use case, each audience and each comparison someone might search for.
- A real title and description on every page, written for a person deciding whether to click.
- A sitemap, submitted in Search Console on the day you launch.
- Structured data that says what the thing is: a product, a person, an article, a video.
When I launched this site, the domain was brand new. It reached the first page for my name in six weeks, and how that happened is almost entirely this list.
Build what the others don't have, or do badly
If your page says the same thing as the ten pages already ranking, there's no reason for Google to put you above them, or for a person to pick you.
Before you build, read what people complain about. One-star reviews of your competitors, Reddit threads, support forums, the questions under their YouTube videos. That's a list of features nobody has built properly yet, written in the exact words your future users type into a search bar.
When I built RevealTogether, the usual way to share a gender reveal with family abroad was a video call where everyone saw the moment at a slightly different time. The whole product is built around fixing that: one countdown, every screen, the same second. It's now used by more than a thousand families in over forty countries, and it ranks on the first page in English and Spanish, because it answers a question the existing pages didn't.
Tell the people you already know
This is the part people skip, and it's mostly psychological. You built something and showing it feels exposing. What if it's not good enough? What if someone you know thinks it's silly?
Share it anyway. Send it to friends and family. Post it on your own LinkedIn, Instagram or TikTok. Put it in the two or three communities where your users actually spend time, and say plainly that you built it and want feedback.
Your first twenty users almost never come from search. They come from people who already trust you. They'll also tell you, for free, where they got stuck, which is worth more than any ranking in the first month. And every share is a chance at a link or a mention, which is exactly what a new domain lacks.
Measure what people do, not just how many came
A visitor counter tells you almost nothing. You need to know where people came from, which page they landed on, what they clicked and where they gave up.
I recommend PostHog for this. It's product analytics, funnels and session replays in one place, with a generous free tier. Set up three things in your first hour:
- A funnel from landing page to the action that matters: signup, purchase, download.
- Session replays on that funnel, so you can watch where people hesitate.
- Events named after what the user did ("started checkout"), not what the code did ("button_click_3").
Watching five real sessions will change your homepage more than a week of guessing.
Read Search Console every week
Google Search Console is the most underused free tool on the internet. The Performance report shows the exact queries people typed before your page appeared, how often it appeared and how often they clicked.
Two things to look for:
- Queries you show up for but didn't write about. Those are people asking for something you could answer. Write the page they're looking for.
- Pages ranking between positions 8 and 20. They're close. A better title, a clearer answer near the top, or a few internal links from your other pages often pushes them onto page one.
Search Console also tells you what's broken: pages it couldn't index, structured data it couldn't read. Fix those first. Some alerts are harmless, and one real error can quietly cost you rich results. I wrote up one I hit on this site and the one-line fix.
Make it fast
Speed is part of ranking, and it's most of the first impression. Run your pages through Lighthouse or PageSpeed Insights and look at the three Core Web Vitals:
- Largest Contentful Paint under 2.5 seconds: the main content shows up quickly.
- Interaction to Next Paint under 200 milliseconds: the page reacts when tapped.
- Cumulative Layout Shift under 0.1: nothing jumps around while loading.
The usual culprits in AI-generated sites are oversized images, a font loaded from three places, and a JavaScript bundle that ships the whole app to render a landing page. Fixing those three often takes a page from red to green in an afternoon.
Ship the corrections
None of this is a one-off. The loop is simple: look at what Search Console and your analytics say, change one thing, ship it, and look again a week later. The sites that grow are rarely the ones that launched best. They're the ones that kept correcting course in public, every week, for months.
AI is good. It isn't the whole job.
I build with AI tools every day, and they've made me much faster. But speed at writing code was never what kept a website invisible. What's missing is the 360 view: search, content, design, analytics, speed and distribution, all treated as part of the product from the first day rather than as a cleanup after launch.
If your site is live and quiet and you'd like a second pair of eyes on it, write to me at consultingyassir@gmail.com. Tell me what it is, who it's for and what you've tried, and I'll tell you what I'd fix first.