Methodology

H-W6 Authority Gate

The discipline that says no article ships without three citations from methodology, research, and product evidence.

5 min read

Most agency content recycles. Three of every five “data-driven marketing” posts on the Indonesian web cite the same secondary source loop — Statista on top of HubSpot on top of a thinkpiece. That recycling is why agency content does not compound: nothing in the article is a primary signal that only the publishing entity can produce.

H-W6 is the editorial discipline that prevents ANYÉ from joining the recycling loop.

What it is

A gate every article must pass before it goes from draft: true to draft: false. The rule is simple: at least three ANYÉ-original citations, with at least one citation each from these three buckets:

  1. Methodology citation. A pointer to a published methodology page like Process Map v6, Rating Engine, Five Lens, or the Investment Priority Matrix. Tells the reader the article rests on a documented framework, not a vibe.
  2. Research citation. A pointer to an ANYÉ-published research piece — the deep reports under /riset/. Tells the reader the article rests on data we have collected and re-analyzed, not data we paraphrased.
  3. Product citation. A pointer to a published case study at /studi-kasus/ or to a public deliverable artifact (template, simulator, downloadable spreadsheet). Tells the reader we have done this work — not just written about it.

Most articles plant more than three. The minimum exists so an article that fails to plant three triggers a halt-and-rework, not a publish.

Why “H-W6”

H is Handy (the principal) and W6 was Week 6 of the original sprint when this discipline was first locked. The name stuck because the discipline is principal-enforced, not delegated to an automated linter. A reviewer (currently the principal, eventually a senior editor) reads the draft, counts the three citation types, and either signs off or sends it back.

Why three buckets, not just three citations

Three citations from one bucket is easier to game. An article could cite three methodology pages and pass on count, while never proving the writer has done the work the article describes. The bucket rule forces the article to display three signals — instrument, data, output — that together demonstrate full-stack capability.

This is the same logic that makes Stripe Press writing trustworthy: every chapter has framework + data + product evidence side by side. A reader cannot tell which one is the load-bearing one because all three are.

What this looks like in a real article

Take the first published ANYÉ article, strategi digital marketing. Inside its 2,400 words: one citation pointing to Process Map v6 Group H (the visibility data points behind the channel-allocation argument), one citation pointing to research on Indonesian SME ad spend (the BPS 2024 dataset re-analyzed), one citation pointing to the WhatsApp gate live on the same site (the conversion-mechanism the article asks readers to use). All three planted, gate passed, published.

What it deliberately does not do

It does not police external citations. ANYÉ cites BPS, Kominfo, GoTo, Sea Group earnings reports, and so on freely — those are primary external sources we want readers to trust. The gate only counts ANYÉ-original citations because those are the ones that prove ANYÉ produces the artifact, not just paraphrases it.

It also does not police topic. Any topic that fits the editorial guidelines is fair game for an article, as long as the H-W6 gate passes. Topic choice is upstream; H-W6 is downstream.

Where it sits

Inside the publishing pipeline. After an SEO article passes L4 quality scoring, before it can flip draft: false and merge to main, it goes through the H-W6 gate as a final read. Failure routes the draft back to L3 (the strategy stage) for citation augmentation; success ships.