The Crit Pick SEO case study: the first month of real data
Two weeks to get crawled, then 18 clicks and ~1,650 impressions. The first real Search Console data from Crit Pick, plus the dos and don'ts behind the build.
Two weeks after Crit Pick went live, Google finally crawled it. Here's the first stretch of real Search Console data, about 1,650 impressions and 18 clicks, and what those small numbers already say about the method. Crit Pick is an SEO experiment I run in the open, and this is the first checkpoint with data instead of just a plan.
Data through · Google Search Console
- The blog is the case study. Not a review site. Every page is chosen for keyword opportunity, not personal taste.
- We don't fight the head term. “rimworld” sits at difficulty 89. The cluster gets won on the long tail, difficulty 10 to 30, plus one high-volume utility page as the front door.
- Structure first, then pages. A hub routes down to deep dives; a mod only gets its own page when it has its own search demand.
- The first data is in, and it's small. Around 1,650 impressions, 18 clicks, average position 11. And it took two weeks before Google even crawled the site.
- Read it as a starting point, not a result. The numbers are tiny. The point is watching them grow, in public, with the method visible.
What Crit Pick is (and why it's games)
SeeIndie teaches indie devs how to get found. Telling people “SEO works” is cheap, so Crit Pick is the proof, built in public with the exact methods SeeIndie writes about. If it ranks, the method holds. If it doesn't, you'll read that here too.
Games, not another finance site, for a simple reason: the competitive shape matches indie marketing. Big audience, saturated content, a few dominant incumbents, and a long tail of low-competition searches nobody's answering well. The lessons carry over. The starting focus is Guild Wars 2 and RimWorld, with Guild Wars 3 covered as an upcoming release. Deep mechanics, strong long-tail demand, real player interest in guides and builds. Good ground to show keyword targeting at scale.
The strategy in one chart
Here's the whole idea in one picture. This is keyword difficulty across the RimWorld cluster, on a 0 to 100 scale where lower is easier to rank. The long bars are terms I don't chase head-on. The short gold bars are where a small site can actually win.
The method, start to finish
Nothing here is clever. It's the same five moves on every cluster, done in order and not skipped.
The dos and don'ts
Most of these I learned by nearly getting them wrong, and a couple I only understood once the first data came in. Here's the short table, then the two calls worth explaining.
| Don't | Do instead | Why |
|---|---|---|
| Chase the head term (difficulty 89) | Build the cluster on the difficulty 10 to 30 long tail, plus one utility anchor | The long tail is rankable in months. The head term takes years and links a new site hasn't earned. |
| Give every sub-mod its own page | A page only when the term has its own search demand, otherwise a section on the parent | Thin pages dilute the cluster and quietly compete with each other. |
| Reuse a slug you already used | One slug per page, checked before publish | Two pages on one URL collide. I caught a Mechanist and Scrapper build sharing a slug before either went live. |
| Split “armor” and “armour” into two pages | Real name in the H1, target the higher-volume spelling in title and meta, cover both in the body | It's one intent, not two. Splitting it hands Google a reason to rank neither. |
| Expect to get indexed fast | Submit a sitemap, then wait it out | It took about two weeks online before Google crawled the site at all. Nothing was broken, that's just the pace for a brand-new domain. |
| Ignore pages with impressions but no clicks | Rewrite the title and meta description, check the ranking position | The RimWorld lore page pulled 538 impressions and 2 clicks in the first weeks. That's a title problem, not a traffic problem, and it's fixable. |
| Let tag and author archives get indexed | noindex, follow on those routes | Keeps low-value archive pages out of the index without breaking the crawl path to real posts. |
Page or section? Let demand decide
This is the call I make most often, so it gets a rule. A single mod gets its own deep dive only when the data shows people searching for it by name. When they don't, it becomes a section inside the parent page. The RimWorld Vanilla Expanded series has dozens of modules; a handful clear the bar for their own page, the rest live as sections. That keeps the cluster dense instead of spreading it thin across pages that would fight each other for the same weak signal.
Impressions without clicks are a to-do list
The most useful thing in the first data wasn't the clicks, it was the mismatch. The RimWorld lore page showed up 538 times and got clicked twice. That's the highest impressions on the whole site paired with almost no clicks. It means the page ranks well enough to appear but the title isn't earning the click, or it's sitting at the bottom of page one. Either way it's a clear next job, not a disappointment. A page nobody sees is a content problem; a page people see and skip is a title problem, and that's a much faster fix.
The first data
Honest status: the site launched, sat uncrawled for about two weeks, then impressions started climbing from mid-June. Over the window that's roughly 1,650 impressions and 18 clicks, at an average position of 11 and a 1.1% click-through rate. Small numbers. But they're real, and they're up from a flat zero.
The top pages tell a clearer story than the totals. The one that pulled the most clicks was a timely one: a demo review for an upcoming release, which rode fresh search interest before the wider game did. Evergreen pages take longer, because they need the cluster to build authority first. Here's the early leaderboard.
Two movements worth watching. The demo review's impressions already dropped 61% from their peak, which is what happens to timely content once the moment passes. Meanwhile the RimWorld animal tier list jumped over 1,000% in impressions, a sign that page is starting to catch on as the cluster around it fills in. That's the pattern I'm betting on: news pages spike early and fade, evergreen pages start slow and compound.
What's next
Three jobs. Rewrite the RimWorld lore page's title and meta so those 538 impressions start converting. Keep filling the Vanilla Expanded long tail, since that's where the lowest-difficulty wins sit. And watch whether the animal tier list's momentum holds as more of the cluster goes live. Then, once these pages have had a few more months to settle, I'll publish the next set of numbers here: what climbed, what stalled, and what I'd change. The whole point is that you get to check my work.
Common questions
What is Crit Pick?
Why a games blog instead of more finance content?
Isn't this just an affiliate site in disguise?
How long until a new site starts ranking?
Can I use this method on my own game's site?
Follow the experiment
One practical discovery tip and whatever I'm currently testing on Crit Pick. No fluff, no pitch.
