I didn’t start because of a screenshot. I had twenty-five years of infrastructure experience and a straightforward question: does the modern approach to building niche content sites — static builds, Git-based workflows, edge CDN delivery — actually hold up when you treat it like a real project instead of a hobby?
So I ran the experiment. Over the past year, I built five properties from scratch: SafeHarborPrep, focused on preparedness and rural self-reliance; HomesAndGardenDecor, covering home systems and outdoor furniture; OpsForge Labs, for VPS and hosting decisions; MyCozyTrove, fishing and camping gear; and TheCoffeeCan Camp, a camping lifestyle blog. No course. No blueprint from a guy on a beach. I just started deploying and writing.
What follows is what actually happened — the order I did things, the technical debt I created by accident, and how long it took Google to acknowledge any of it existed.
The Order of Operations
The first decision wasn’t niche selection. It was architecture.
Three of the five sites — SafeHarborPrep, HomesAndGardenDecor, and MyCozyTrove — run on Eleventy with a Git-based workflow. Write in Markdown, push to GitHub, Cloudflare Pages builds and deploys automatically. No database, no plugin stack, just static HTML served from the edge. The fourth, OpsForge Labs, uses the same stack with Tailwind CSS layered on top. The fifth, TheCoffeeCan Camp, is still in early build.
That technical setup took about two weeks to get right across the first three sites. A WordPress install would have taken two hours. The tradeoff I made — and I’ll come back to whether it was worth it — was trading setup speed for long-term administrative friction.
Once the architecture was in place, the content approach was simple on paper: identify three to four seed categories per site, map a flat hierarchy so nothing was buried more than two clicks from the homepage, and commit to a foundation layer of roughly 30 articles per site before checking any ranking data. For SafeHarborPrep, the seed categories were power and energy, water and filtration, and preparedness gear. For HomesAndGardenDecor, water filtration, outdoor furniture, and kitchen infrastructure.
The part I underestimated was category infrastructure. You can’t just create a folder and start dropping articles in. Each category needs an index page, a nav entry, an Eleventy collection defined in the config, and at least one article linking to it from somewhere else on the site. I learned this the hard way on MyCozyTrove, where I had fishing articles sitting in a folder with no proper category scaffold around them. The articles were technically live. They just had nowhere to belong.
What Actually Took the Most Time
I thought the technical builds would be the bottleneck. They weren’t.
The real time sink was internal linking — specifically, doing it correctly instead of just doing it. On SafeHarborPrep, the preparedness and power categories needed to be genuinely connected, not just cross-referenced with throwaway anchor text. A solar generator article linking to an emergency water filtration guide isn’t decoration. It’s a signal to search engines that these topics belong to the same cluster of authority. Getting that right across 34 articles on one site, and 39 on another, required custom scripts to audit which articles were orphaned, which hub pages were missing spoke links, and which links were pointing in the wrong direction.
I spent close to forty hours on SafeHarborPrep alone just cleaning up the relationship between the power cluster and the preparedness cluster. That’s not a complaint — it’s the actual scope of the work. Anyone who tells you internal linking is a quick task is either managing five articles or not actually checking their link graph.
The other time sink was affiliate program approvals. I had content ready before the programs approved me, which meant publishing articles with placeholder CTAs and coming back later to wire in the real links. That process stretched some clusters by four to six weeks. I’ll cover which programs moved fastest and which ones stalled in The Gear and Outdoor Affiliate Programs That Actually Approved Me.
The Timeline of Silence
For the first four months, combined traffic across all five sites would not have filled a small coffee shop.
OpsForge Labs, which covers VPS and hosting decisions — a niche with real commercial intent but serious existing competition — was effectively invisible for about 150 days. HomesAndGardenDecor started showing up in search results around day 60, though “showing up” meant impressions, not clicks. MyCozyTrove lagged behind what I expected because I hadn’t accounted for the seasonal search patterns in fishing and camping gear. Queries spike hard in spring and early summer, then fall off. If your content isn’t indexed before the season opens, you miss it.
The rough timeline for a new niche site in the current environment:
Months 1–2: Technical setup, indexing lag, near-zero traffic. Google knows you exist. It isn’t sure yet whether you matter.
Months 3–4: You start ranking for long-tail queries — four- and five-word phrases that get maybe 50 searches per month. Traffic trickles. You check Search Console more than you should.
Month 6: A few articles break onto page one or two. Daily clicks move from single digits to double digits. The pattern becomes visible.
The fishing and camping content on MyCozyTrove is getting clicks now. It’s not yet converting at a meaningful rate, which is honest and probably more useful to know than a revenue number that would be too early to mean anything.
What I’d Do Differently
Three things, and I’d prioritize them in this order.
First, I would fix the category infrastructure before publishing the first article in any new category. The index page, the nav entry, the collection definition, and at least one inbound link — all of it needs to exist before the first article goes live. Publishing articles into an incomplete category scaffold and then cleaning it up later costs you the crawl window. On new sites, that window matters.
Second, I would stop chasing categories that are too broad. “Gear” is a good example from MyCozyTrove. I built several articles under a gear category that had no clear topical focus. The articles weren’t bad. They just didn’t reinforce each other, so none of them built much authority. I’ve since been migrating that content into specific categories where it actually belongs — backpacks, fishing, camping — and the cluster structure is visibly cleaner. There’s a longer version of this in How I Chose Topics and Categories.
Third, I would accept earlier that the content cluster architecture is the whole game. It’s not a nice-to-have on top of good articles. It’s what makes individual articles useful to search engines and AI systems trying to figure out if your site has real authority on a topic. I built this out properly on the backpacks cluster for MyCozyTrove — 11 articles, clean bidirectional linking, hub updated at the same time each spoke went live — and the difference in how that cluster performs versus earlier, loosely connected content is not subtle. The full model is in The Content Cluster Strategy That Gets AI Search to Notice You.
On the WordPress versus Eleventy question: I still run the five properties on Eleventy and Cloudflare Pages. For SideHustleBuilderPro, I went back to WordPress, because the tooling for bulk edits and content management at scale is more mature and I didn’t want to spend time on build pipelines when the point of this site is to document the process, not demonstrate it. That’s not a knock on static builds. It’s a reflection of what each tool is actually for.
The Bottom Line
The five sites are active, indexed, and getting traffic. None of them converted in the first six months in any meaningful way — that’s normal, and anyone who tells you otherwise is either in a rare niche or rounding up aggressively.
What I’d tell someone starting today: pick one site, build the category infrastructure before you need it, and connect your articles to each other in a way that a human can actually follow. The first six months are mostly waiting. Use that time to get the link graph right, because once Google decides what your site is about, changing that impression takes longer than building it correctly the first time.
The four articles below cover the specific mechanics — how I chose categories, how I structured the clusters, why MyCozyTrove took longer than expected, and which affiliate programs actually approved a new site: