The decisions behind running WordPress at agency scale.

Running one WordPress site well is a technical problem. Running thirty across a client estate is an operational one, and most of the published advice does not distinguish between the two.

This topic is about the second problem: which plugins earn a place in a standard build and which were cut after failing an audit, how updates get staged so a security patch does not take down a storefront, why plugin count is a useful proxy for maintenance surface, and what hosting decisions are worth arguing with a client about. We name the tools we actually deploy and the ones we removed, with the reasoning in both directions — including the cases where the popular choice lost to something less known. The bias throughout is toward decisions that stay correct eighteen months later, when the person maintaining the site is not the person who built it.