Here’s the question this article actually answers: how do you get eight pieces of content out of one research push without repeating yourself, diluting the message, or quietly writing the same article nine times with a different logo pasted over it? The short version — one thesis, forked by voice and by who’s reading, cross-linked into a map before anyone drafted a sentence. The long version is everything below, because the short version is easy to say and easy to get wrong in practice.

Claude and I didn’t set out to write a “content strategy” piece about ourselves. We set out to answer one real question — is WordPress worth staying on in 2026 — and along the way built something more useful than any single article: a repeatable pattern for turning one piece of real work into a cross-site authority web. This is that pattern, documented plainly, with links to every piece it produced, because a capstone that can’t point at real, live pages isn’t a capstone. It’s a promise, and promises aren’t the thing we’re in the business of selling.

Where it started — one thesis, two data points

The thesis fits in one line: WordPress isn’t dying, it’s retiring on a shelf-life clock — speed, security exposure, cost, and how readable a site is to an AI answer engine — and the real unanswered question underneath all of that is how a business owner keeps control of their own content without it. Everything downstream of this piece — eight articles, three sites, two audiences — is one argument, restated for whoever’s actually asking.

We let ourselves cite exactly two numbers, and they don’t carry equal weight, which is worth saying plainly instead of gliding past it:

  • WordPress’s market-share decline, tracked by W3Techs — third-party, methodologically solid, citeable as-is, no hedging required.
  • Our own site-staleness scan — how many small-business WordPress blogs had actually been touched in the last year.

That second number didn’t start as something we were allowed to publish as fact. It started as an informal read of a few hundred sites, useful enough to open a conversation, nowhere near rigorous enough to stand behind in NW eSource’s own voice. So we said so, out loud, in the early drafts — and held the line until the scan was actually scaled: stratified across ten-plus verticals, run to 1,800 U.S. small-business sites, documented with an actual methodology, and published as its own citeable asset — the NW eSource 2026 Small-Business Website Audit. Only after that shipped did NW eSource’s own content get to cite its own number as fact.

Of the WordPress business blogs in that 1,800-site audit, 48% hadn’t published a post in twelve months or longer, with a median gap of 11.5 months — a figure that held up when we re-ran it on a second sample to check it wasn’t a fluke.

That gating rule — don’t let your own brand cite your own number until it’s been scaled and documented — is the least glamorous discipline in this whole project, and it’s also the thing that makes the rest of it credible instead of self-congratulatory. Skipping it would have saved a few weeks. It also would have meant every article downstream was standing on a number we hadn’t actually earned yet.

The fork — one thesis, two voices, two funnels

Once the thesis and the data were settled, the next decision wasn’t “what do we write next.” It was “who is this for, and what do they already believe.” Two very different readers were asking the same underlying question, and writing one generic article aimed at both of them would have satisfied neither.

  • This site — CM — writes it peer-to-peer. First-person, “Claude and I,” aimed at developers and other agency operators who want to see the mechanism, not just the conclusion. It funnels to the Agency AI Operating System.
  • NW eSource writes it outcome-first. Buyer-intent, leadership-voiced, aimed at the business owner deciding whether their own site needs to change. It funnels to a consult, not a capability breakdown.

The clearest proof this fork actually works is that the cornerstone got written twice, on purpose, and neither version is a duplicate of the other. The CM variant — Better, Faster Websites & the WP Retirement Plan — opens with the four forces retiring WordPress and walks a builder through the five-model system for keeping an owner in control of their own content. The NW eSource variant — the 2026 audit — opens with the audit itself and answers a narrower, sharper question: is it time, for your business, specifically. Same research underneath both. Different entry point, different structure, different close, because a developer reading this blog and a business owner reading NW eSource’s are not, in fact, asking the same question in the same words — even though the honest answer underneath is identical.

The engine underneath — agency side and client side

Two articles came out of this same research push that aren’t rewrites of the cornerstone at all — they’re the machine that makes any of this repeatable past a single good idea:

  • How agency-directed content leads SEO/GEO covers the agency’s actual job: the category map, the answer-first structure, the FAQ-and-schema slot — built before a sentence of prose exists, the same discipline as designing a database schema before anyone enters a row.
  • Getting to know your client covers the half no agency or AI can manufacture on its own: the owner’s real frame of reference, the specific way this business already talks to its customers, pulled out through actual discovery rather than assumed from a template.
  • Where agency SEO/GEO meets client engagement is the keystone — the piece that turns the other two from a philosophy into something operational, with the actual brief template, the required-elements checklist, and the linked-card pattern any of this runs on.

That three-part shape — agency engine, client engine, a keystone where they meet — is itself the reusable part. Any agency running its own service, not just a web-design shop, can build the same triad around whatever it does: an architecture half nobody but the agency can supply, a voice half nobody but the client can supply, and a defined seam where the two meet instead of quietly overwriting each other.

Proof it isn’t theoretical

None of the above matters if it stays a framework nobody’s actually tested against a real decision. The use-case survey is where the thesis gets checked against evidence that isn’t ours: real, named companies’ documented results across headless WordPress+WooCommerce, Cloudflare Pages, and modern static-JS stacks — including the honest fifth answer, that plenty of businesses are still right to stay on WordPress. That last point matters more than it looks: a framework that can’t say “not you, actually” isn’t a framework, it’s a sales script with extra steps.

Reaching outward — the guest post and its follow-on

At some point a content play stops being internal and starts compounding through someone else’s audience, and that’s what the last two pieces in this suite do:

  • Retire WordPress the Right Way carries the same core argument to NW Business Links’ B2B-networking readership, in Charles’s own author voice, with a bio that cross-links his NWBL business profile, this site, and NW eSource — three destinations from one guest post. It was also built as a reusable guest-post template, so any other NWBL member business can run the same pattern for their own content instead of starting from a blank page.
  • How Portland’s small businesses stack up online is the follow-on that guest post earned — a locally scoped data piece, scanning Portland-area small-business sites specifically, for the same B2B audience that read the how-to first. One good guest post didn’t just place an article. It opened a channel for a second one.

The map, laid flat

Skimmed on its own, the shape of this looks like a diagram. Written out plainly, it’s just a sequence of decisions, each one made before the article it produced:

  • The WP retirement plan sits at the center — cornerstone, written twice, once per brand voice — and links down to the agency/client/keystone triad and across to the use-case survey.
  • Agency-directed content and getting to know your client both feed up into the keystone, which hosts the actual reusable templates.
  • The use-case survey cites real external deployments as evidence, sitting alongside the cornerstone rather than beneath it.
  • The NWBL guest post is the outward on-ramp — one post, three linked destinations — and it opened the door to the Portland audit that followed it.
  • This article sits on top of all of it, publishing last, on purpose — every link above resolves to something that was actually live by the time you’re reading this, not a placeholder waiting to catch up.

That publish-order discipline is worth naming directly, because it’s easy to skip under deadline pressure: a capstone that links to draft pages isn’t a capstone, it’s a promise. Same rule as the stat-gating story earlier — the boring, unglamorous choice (wait until the thing you’re pointing at actually exists) is the one that makes the rest of the piece trustworthy instead of decorative.

What any agency can copy

None of this required a bigger team, a bigger budget, or a cleverer thesis than “WordPress is retiring, not dying.” It required three ordinary disciplines, run in order, every time:

  1. Data worth citing — and the restraint to label the informal version as informal until it’s actually been scaled and documented.
  2. A clear voice-fork by audience — one argument, deliberately rewritten for who’s actually reading, never one generic post trying to serve two readers at once.
  3. A link map built before the first draft, not stitched on after the fact once eight articles already existed and someone remembered they should probably connect to each other.

Any agency operator, or any solo developer-consultant running their own content, can run the same three steps around their own real work — a migration they did, a system they built, a client problem they actually solved. The scale doesn’t need to match ours. The order does.

This entire suite — eight articles, three sites, one research effort — is what the Copywriting & Content capability inside the Agency AI Operating System actually produces when it’s run as a system instead of a one-off blog post: the category map, the required-elements checklist, the cross-link discipline, and the voice-fork, all applied to real research instead of generic AI filler. See it at the Copywriting & Content card.

FAQ

What is a multi-site content strategy, and why would one agency run more than one blog? It’s deliberately forking one piece of real research or work into separate articles for separate audiences and sites, instead of writing one generic post and hoping it lands with everyone. Claude and I run three properties across this suite — this blog, NW eSource’s, and a B2B guest outlet — because a developer, a business owner, and a networking audience are asking the same underlying question in different language.

Doesn’t publishing the same idea multiple times hurt SEO through duplicate content? Not when each version is genuinely rewritten for a different voice, a different reader intent, and a different funnel destination — that’s reframing, not duplication. The risk shows up when a team copy-pastes and swaps a logo; the fix is treating each site’s version as its own draft from the same source material, which is why the two A1 cornerstones read nothing alike despite sharing one thesis.

How do you decide what to cite as fact versus what stays labeled as informal? Rigor tier decides it, not convenience. A third-party number like W3Techs’ WordPress market-share data gets cited outright. Our own site-staleness read started as an informal scan of a few hundred sites, labeled as exactly that, and only became something NW eSource could cite as fact once it was scaled to 1,800 sites, stratified across verticals, and published with a documented methodology as its own audit.

What actually holds a cross-linked content architecture like this together? The link map gets built before the first article is drafted, not stitched on after. Every piece has a defined home, a defined set of things it links down to, and a defined thing it links up to — so the last article in the series can point at real, live pages instead of placeholders, which is the entire reason this one was held until eight others were confirmed live.

Where does this kind of content strategy actually get used for a client, not just for Charles’s own brand? Inside the Copywriting & Content capability of the Agency AI Operating System — the same discipline behind this suite (required elements, voice-fork by audience, a link map built ahead of time) runs for client blogs, not just this one.

🔧 Build Log Drafted direct from brief by Claude Sonnet 5 · reviewed against CM voice guardrails in the same session · ~2,150 words / ~2,950 tokens output · Est. elapsed 7m · Est. agency equivalent: $500–700

If turning one piece of real work into a content engine that compounds instead of repeating itself is useful to you, that’s the Copywriting & Content capability inside the Agency AI Operating System — see how Claude and I build it, then apply to work with us.