Content Clusters SEO: A Build-and-Ship Runbook for Link Contracts (Steps, Mistakes, Checklist)
Stop designing clusters and start shipping them. A production-line procedure for hub selection, spoke quotas, reciprocal linking and quarterly repair - with South African cost context.
Why Most Content Clusters SEO Projects Stall After the Outline Is Approved
The outline gets signed off. The spreadsheet has 40 rows. Six months later, four pages exist and nobody can explain what happened. The failure is almost never strategic - it is operational. A cluster is not a folder of briefs; it is a link contract. If the links between pages are not specified before writing starts, they do not get built, and an unlinked set of pages delivers none of the three things a cluster is actually for: fewer competing URLs per query, shorter click depth to money pages, and a higher percentage of queries with an assigned owning URL.
Who This Build-and-Ship Procedure Is For (and When Not to Use It)
This runbook suits an SEO lead or content ops manager who inherited a half-built cluster, answers to an editor who controls sign-off, and needs a ship schedule rather than another architecture deck. It assumes you can publish 4-8 pages a month. If you are pre-launch with no indexed pages, build your first ten service or product pages instead - clusters amplify existing relevance, they do not create it.
Before Step 1: Four Prerequisites You Need in Place
- Crawl export of every live URL with status code and inbound internal link count.
- Query-to-URL inventory pulled from the GSC Performance report, pages tab, 3-month window.
- A named editorial sign-off route - one person, one turnaround commitment in days.
- A cluster-level tracking filter (URL path or regex) built before the first page ships, so you can report the cluster separately from day one.
Step 1: Pick the Hub Using Search Console Data - Not Intuition
A hub must pass three tests: it already earns impressions in the query-page report, it carries commercial intent, and it is broad enough to link to every spoke without behaving like a directory index. If no existing page qualifies, the hub is a new page and must ship first. Google's guidance on site structure supports this: keep important pages within a few clicks of the root and link them by meaning.
Hub-selection rule: if the candidate hub links to more than ~15 spokes, it is a directory, not a hub. Split it or pick a tighter parent.
Step 2: Write the Cluster's Link Contract Before You Brief a Single Spoke
The contract is three clauses: the hub links to every spoke with descriptive anchor text; every spoke carries at least one contextual link back to the hub and at least two sibling links; and no linking lives only inside a bullet list at the foot of the page. Topical authority is built the same way - contextual, in-prose links carry more information scent than a link farm - NN/g's work on information scent explains why the surrounding sentence matters as much as the anchor itself. Write the contract into the brief as a table of three mandatory link slots per page.
Step 3: Quota the Spokes Into Waves (With a Realistic Throughput Estimate)
One writer-editor pair typically ships around 4-8 substantive spokes per month alongside maintenance - an estimate, not a benchmark. That means a 40-spoke cluster needs a 6-10 month runway. Promise Wave 1 as 8-12 spokes, not the whole hub.
| Wave | Spokes | Monthly output | Elapsed |
|---|---|---|---|
| Wave 1 | 8-12 | 4-6 | 2-3 months |
| Wave 2 | 12-16 | 6-8 | 2 months |
| Wave 3+ | Remaining | 4-8 | 3-5 months |
