I was wrong about SEO. Here is the data.

I was wrong about SEO. Not a small wrong, a "no one searches this category" wrong. I said it out loud to my team, working off instinct, and I said it with confidence. Google Trends proved me wrong. That should have taught me something about the difference between what I believe and what I have checked. It did not, quite, because a few weeks later I made a version of the same mistake again, this time about my own site.

The instinct I trusted, and the check I finally ran

The instinct this time was smaller: that FirstCustomer would simply show up when someone searched its own name. A brand query. The easiest kind of ranking there is, or so I assumed. I had shipped meta tags, an og:image, a title tag with the product name in it. Surely that was the work.

It was not the work. I ran the actual searches on 2026-09-06, five queries, through Google directly, and read what came back instead of what I expected to come back.

firstcustomer.net, the full domain, ranks first. That one I got right. But that is the easy query. You are asking Google to find the URL you already told it about, and Google is happy to oblige.

firstcustomer, the actual brand name as a single word without the domain suffix, does not appear in the top 20 results at all. The page that owns that search is firstcustomerpodcast.com. A podcast, not us, sitting on the name I want people to associate with my product.

FirstCustomer app does not go how I expected either. Google auto-corrects it to "First Customer app," splitting my one word into two. It does not yet recognize the brand name as a single entity. A competing product, firstcustomer.ai, shows up in that result. My own .net does not show up at all.

"first customer" in quotes, the generic English phrase, returns the podcast, Reddit threads, a Harvard reference, and a company called EarlyCustomers. That query was never winnable and I am dropping it from the keyword plan entirely.

Why meta tags could not have fixed this

Here is the mechanism, because the mechanism is the part I skipped past the first time. Ranking is not a property of a page. It is a property of a page relative to what the rest of the internet says about it. Google's own site: search shows exactly two indexed URLs for the domain: the homepage and the app. subdomain. Not the privacy page, not the terms page. Two. There is no sitemap, so there was nothing for Google to discover beyond what it happened to crawl on its own. Two URLs is not a site to Google. It is a stub.

Then I went into Search Console and inspected the homepage directly, field by field. The field that mattered most was labeled "Referring page." The value recorded there was "None detected." Zero. No page anywhere on the internet links to firstcustomer.net. Not one. Google is not hiding this from me or making me guess at it. It states the fact directly, and that fact is the actual reason the brand query loses to the podcast. A podcast that has presumably been linked to, mentioned, and shared for years has exactly what my domain that is not even two months old does not have: other pages vouching for it.

The same inspection showed no user-declared canonical tag and no structured data of any kind. The last crawl on record was from six days before I checked, which by itself is unremarkable. What it confirms, alongside the two indexed URLs, is that this is not a broken site. It is a site Google looked at once, found almost nothing worth returning to, and moved past.

None of this is fixed by a meta tag. A title tag tells Google what a page claims to be about. It says nothing about whether the rest of the internet agrees the page is worth showing to someone who typed a name into a search box. I had the causality backwards. I treated indexing and ranking as one problem, solved by the same fix. They are not the same problem. Getting indexed is a technical checklist, the kind of thing you can, in fact, finish in an afternoon. Getting ranked for your own name against an established competitor is an entirely different kind of work, and it runs on time and on other people's pages linking to yours, not on markup.

The part I would rather not publish

There is an uncomfortable finding underneath the technical one, and burying it would defeat the point of writing this at all. firstcustomer.ai already exists, positioned around getting a founder their first customer, and it outranks my .net on more than one of these five queries. firstcustomerpodcast.com has apparently owned the bare word "firstcustomer" long enough that a brand-new domain has no realistic path to displacing it quickly. I did not choose an empty name. I chose a name with two established occupants already sitting on the search results I want.

I do not have a clean answer for that yet. The honest one is: I do not know if a domain that is not even two months old with zero backlinks can out-rank a podcast for a one-word brand query, no matter how much content I publish under it. It might. Content and links accumulate, and podcasts do not automatically win forever just because they were first. Or it might not, and in three months I will look at this same set of five queries again and find the same order, in which case the honest conclusion is that this is simply the cost of the name I picked, not a problem SEO work can solve. I would rather say that plainly now than dress it up as a temporary setback I am confident about beating.

What actually changes because of this

The fix, correctly scoped, has two parts and only one of them is quick. Getting indexed properly, meaning a real sitemap, canonical tags, and an actual Organization entry telling Google that "FirstCustomer" is a proper noun and not a typo of two words, is a checklist, and most of it is already shipped. That part was never the hard part. It only felt like the hard part because it was the part I could finish in a single sitting and call done.

The part that is actually hard is the one with no shortcut: getting other pages, real ones, written by people who are not me, to link to firstcustomer.net. That is what "Referring page: none detected" is telling me to go build, and it is not something a checklist produces. It comes from writing things worth linking to and waiting for people to decide, on their own, that they are worth sharing. This post is one attempt at that. It will not be the last, and I am not going to pretend one post changes a zero-backlink domain into a well-linked one.

I would rather have found this out from Google Trends the first time than needed a second lesson from Search Console. The pattern is the same one from before, and I keep falling into it at the exact point where checking would have been cheap and skipping it felt harmless. Assume less. Look at the actual numbers before deciding what is true. I am writing this down mostly so that the next time I am tempted to skip the check, I have something on record to remind me what it cost the last two times.

What I am telling the team

Stop writing anything that assumes ranking follows from shipping the technical checklist. The checklist matters, it is necessary, and it is nearly done. But the number that decides whether "firstcustomer" ever beats a podcast is "Referring page: None detected," and the only way that number moves is other people deciding, on their own, that something we published was worth linking to. That is a slower, harder thing to build than a sitemap. It is also the only thing that actually works.