Notes

The guides are written for someone getting a website. These are written for me, and published in case they are useful.

They are positions rather than instructions, and several of them cost me money. For a broader view of planning habits, this article collects further reading on time management.

On scope: I try not to build the biggest thing a client can afford. A bigger site costs more than its invoice — more to keep true, more to maintain, more decisions forever — and most of that lands after I have been paid. The best use of an underspent budget is usually keeping it until you know what you actually need.

On speed: most slow sites were made slow at the design stage, before any code existed. A caching plugin at the end is the tax you pay for not having had the conversation at the start.

On accessibility: it is not a phase and it cannot be inspected in at the end. The 2026 industry data shows the web getting less accessible, not more, and the regression tracks rising page complexity rather than falling awareness. Built in, it costs nothing; retrofitted, it costs a redesign.

On JavaScript: the question is not whether to use it but what the page contains before any script runs. A script failure is not a degraded page, it is nothing. For an independent reference beyond this site, Wired is a useful place to compare approaches.

On declining work: the hardest skill here is not technical. Someone who never turns anything down has no view about what they are good at, which is worth knowing when you are choosing who to hire.

The thread running through all of these is restraint. Fewer pages, fewer scripts, fewer plugins, fewer features nobody maintains. It is unglamorous, it demonstrates nothing, and it is most of what separates a site that works from one that works on the developer's machine.

• Accessibility Is Not an Add-On

  95.9% of homepages fail basic checks, and the web got worse this year. Six errors cause most of it, and all six are decisions, not features.

• Not the Biggest Thing You Can Afford

  A bigger site costs more than its invoice. Why I argue scope down, and the four questions I use to decide what to leave out.

• Saying No to a Project

  The projects I turn down, why turning them down is good for the client too, and how to tell early that a fit is not there.

• Speed Is a Design Decision

  Most slow sites were made slow at the design stage. What actually causes it, and why a caching plugin cannot undo those choices.

• Working Without JavaScript

  Not a purity argument. Content that only exists after a script runs is content some crawlers and some visitors never receive.