I registered yassirhalaoui.com on 7 September 2026. On 21 October, the searches I ran for my name showed it on Google's first page, fifth, behind my LinkedIn profile and a few people who happen to share part of my name. That's not the top spot yet. For a domain with no history and almost no links, six weeks to page one is about as fast as this goes, and most of what made it work isn't specific to me.
Where it started
Before the site existed, searching my name returned two results that were actually me, LinkedIn and TikTok, followed by namesakes: a researcher with a similar surname, and encyclopedia pages for people called Yasser or Yassin.
That's a weak result page, and a weak result page is good news for a new domain. There's nothing strong to displace. The real risk is ambiguity: Google has to be confident that this site, that LinkedIn profile and those other pages are the same person, and that the namesakes aren't.
Make the person unambiguous
Most of the work was making one entity impossible to misread.
An exact-match domain. A domain that is literally the name does a lot of the work before a single page exists.
One Person, described once. Every page's JSON-LD points at the same
Person node: name, job title, city, employer, education, credentials, and the
profiles that belong to him. Pages reference it by a stable @id rather than
describing a slightly different person each time.
Profiles that point back. sameAs lists the profiles that are really mine.
That only fully works once each profile links back to the site in its website
field, and getting every one of them there is still in progress. A one-way
claim is just an assertion.
The same facts everywhere. Title, city, years of experience, spelled the same on the site, the CV and the profiles. Small contradictions, a different city on one profile or a year missing on another, are exactly what keeps an entity from settling. This is still the part I'm finishing.
Give Google pages that are about the person
A one-page portfolio is thin. Over the first weeks the site grew pages that each answer a real question about the same person: an about page, a CV in HTML rather than a PDF, case studies for each product, and field notes on problems from production work. Later came English and French versions of the main pages, with hreflang pointing both ways.
None of those pages target the name directly. They make the site the most complete source about the name, which is what a person-name search is looking for.
Fix what Search Console reports, and ignore what it doesn't mean
Two alerts arrived in the first six weeks, and only one needed fixing.
The first, on 16 September, said some pages weren't indexed because of "page
with redirect". On a new site that's usually Google doing its job: it found the
www and http versions and trailing-slash variants, followed their redirects,
and indexed the destination. I checked that every URL in the sitemap returned
200, that every canonical pointed at a URL returning 200, and that no internal
link used a redirecting form. All three were true, so the alert needed no fix.
Removing the redirects to make it go away would have created duplicates.
The second, on 20 October, was a real structured-data error on the video's
creator field, and it took a one-line fix in a shared helper. The details are
in invalid object type for field
creator.
That same week of checking turned up a bug no alert had mentioned: /fr/
redirected to /fr, which returned a 404. A French homepage now lives there.
Reading the site the way a crawler does found it. Waiting for a report wouldn't
have.
What didn't matter much
Some things I expected to help did little on this timescale.
Volume. Twenty queued articles don't outrank three good ones on a new domain. The notes publish weekly from a queue rather than all at once, which keeps the dates honest and avoids a pile of pages arriving with no history.
Clever metadata. The title that works for a name search is the plainest one: the name, the role, the city.
Tools that promise instant indexing. IndexNow gets pages into Bing quickly, which matters because Bing feeds ChatGPT's search. Google ignores it. For Google, the sitemap and Search Console are what count.
What's left
Fifth isn't first. The remaining work is unglamorous: making the profiles agree with each other and with the site in every detail, and earning a few links from pages that already rank for technical topics. LinkedIn will probably stay in the top three for good, and that's fine. The goal is for the site to be the first result, not the only one.
If you're starting from a new domain, the order I'd follow is the one above: make the entity unambiguous, give it pages worth reading, fix what's actually broken, and stop there for a few weeks. Most of the waiting is just waiting.