The real story behind client builds — bugs included, not edited out.
Written by Gardy·Building in public
An audit of FlowHaxa showed a four-stage framework with almost all its weight in one stage. Here's how I rebalanced the ends, fixed a newsletter that led nowhere, and set the site up to measure what earns, without claiming results I don't have yet.
FlowHaxa looked finished. Underneath, two copies of the site competed for the same rankings, six images loaded from someone else's servers, and mobile PageSpeed sat at 65. Here's what one day of foundation work changed, including one bug that turned out not to exist.
DryOutHub had 48 real pages of water-damage and mold content. Google still flagged it as low value. The content wasn't the problem — and finding out what actually was taught me more than writing the content ever did.
Small-business owners don't lack cash-flow tips. They lack an order to apply them in. FlowHaxa started as four verbs, Control, Optimize, Automate, Build, and every hub on the site is one of them. Here's how the framework became the structure, and where it didn't.
There's a specific moment in building BizCredX where I realized I was answering the wrong question. Someone doesn't want to know what business credit is. They want to know what to do next — and those turned out to be very different sites.
A burst pipe at 11pm isn't an information problem. It's a five-minutes-ago problem. Building FixRN meant admitting that a good article was never going to be enough on its own.
How building an HSA eligibility website taught me about search behavior, information architecture, connected digital experiences, and learning by building.
Nobody wants to read fifteen articles to understand their own retirement. Building SavedNest meant admitting that publishing more information was never actually the hard part.
Every tech-career site will tell you bootcamps exist. Almost none of them will tell you which one fits your actual budget, timeline, and starting point. Building TechPathU meant sitting in that gap on purpose.
After 8,739 impressions and 13 clicks exposed HSAMaxGuide's first problems, I stopped adding pages and rebuilt the site's information network around seven topic hubs. Here's what changed, and what I'm measuring next.
Trial, error, a wrong phone number on a flyer, a site that wouldn't deploy, and how a 31-page beauty brand website actually got built.
Three sessions of HSAMaxGuide fixes, each one driven by real Search Console data: page-one rankings nobody clicked, a grammar bug on 17 pages, a www duplicate hiding 12 pages from the index, and the honest math on what the traffic was actually worth.
