How to Review a Multi-Page Learning Journey
A collection is not automatically a journey. A learning journey works when a reader can enter on any page, understand where they are, decide whether order matters, take one useful action, and recover the parent route without relying on browser history.
Use this review for courses, documentation series, concept clusters, onboarding lessons, and Playbook subtrees. Review the rendered set as a system; a page-by-page prose audit cannot prove the transitions.
Declare the reader transformation
Before opening the browser, write:
Reader:
Starting problem or belief:
Useful change after the journey:
Pages in scope:
Required order, if any:
Practice or decision produced:
Observable proof:
If the pages have no shared transformation, treat them as a collection and provide a chooser. Do not imply a sequence merely because files have a sidebar order.
Map the route
For every page, record five things:
| Route element | Review question |
|---|---|
| Entry | Can a direct-entry reader identify the collection, page job, and intended outcome? |
| Position | If order matters, is the current step and total visible in words, not color alone? |
| Transition | Does the next link explain why that page is useful now? |
| Return | Is there a real link to the parent route near the top and after long content? |
| Practice | What can the reader do, decide, teach back, or measure before continuing? |
W3C guidance supports the underlying controls: links should communicate their purpose in context; repeated navigation should remain predictably ordered; and people should have more than one way to locate pages in a set. See link purpose, consistent navigation, and multiple ways.
Choose sequence or choice
Use an ordered journey only when one page supplies knowledge or output required by the next. Number the steps, preserve the same order in the hub, sidebar, breadcrumbs, and previous/next controls, and state what each transition carries forward.
Use a chooser when pages solve different situations. Name the situation beside each link. A recommended route may still exist, but say that it is optional. This prevents navigation chrome from inventing a dependency the teaching does not have.
Do not mix the two models silently. “Step 3 of 4” and “read in any order” can coexist only when the page explains that the number is a recommended route rather than a prerequisite.
Run the browser review
Start at the parent hub, then repeat the run from each deep URL as if it arrived from search or an agent citation.
- At desktop width, capture the hub and first viewport of every page.
- At 320px and 375px, verify the same route, labels, and reading order remain available.
- Follow the intended next link instead of typing the next URL.
- At every transition, state what you expect before clicking and what the destination actually provides.
- Scroll to the end of long pages and verify the next action and parent return path repeat.
- Check headings, landmarks, link names, keyboard order, focus visibility, console errors, and horizontal overflow.
- For wide tables, diagrams, or code, preserve the relationship on mobile. If horizontal scrolling is necessary, provide a visible cue and keep row or column labels understandable while scrolling.
Capture evidence while moving through the route. A screenshot proves a visible state; a reproducible click path proves a transition; an accessibility-tree snapshot proves names and structure. Source intent alone proves none of these.
Gauge the journey
Return READY, HOLD, or STOP against these controls:
| Control | Pass condition |
|---|---|
| Transformation | The set names one reader change, or honestly identifies itself as a collection. |
| Route coherence | Hub, sidebar, breadcrumbs, and previous/next controls express the same sequence or choice model. |
| Direct entry | Every deep page names its collection, job, and parent return route. |
| Transition scent | Link text says what the next page helps the reader do, not only its title or “Next”. |
| Practice | Every teaching leg changes a decision, produces an artifact, or supplies a teach-back check. |
| Mobile continuity | Order, labels, actions, comparisons, and return paths survive recomposition. |
| Accessibility | Page structure and repeated navigation are semantic, descriptive, and predictable. |
| Retention | The journey ends by naming what evidence changes the next question, control, or setpoint. |
STOP when the route hides private material, misrepresents evidence, or traps the reader. HOLD when a required transition, parent route, or mobile/accessibility check is unproven. READY requires direct-entry and end-to-end browser evidence, not five individually plausible pages.
Copy-and-paste review prompt
Review these pages as one multi-page learning journey.
Reader and intended transformation:
URLs in the intended order:
Required order or optional chooser:
Expected practice or decision:
Use browser evidence at desktop, 320px, and 375px. Start once from the hub and
once from every deep URL. For each page record entry context, position,
transition scent, parent return path, practice, mobile continuity, and
accessibility structure. Compare hub, sidebar, breadcrumbs, and previous/next
order. Report reproducible issues by severity, then return READY, HOLD, or STOP.
Do not call the journey coherent from source inspection alone.
Failure modes
- Collection masquerading as sequence — sidebar order implies prerequisites that the lessons do not require.
- Five good pages, one broken journey — local quality hides missing or contradictory transitions.
- Title-only transition — the link names a destination but not why it is useful next.
- Browser-back dependency — a deep-link reader has no real parent route.
- Desktop-only course — the learning route or comparison disappears behind mobile overflow.
- Completion without retention — the last page ends, but no practice, evidence, or improved next question survives.
Context
- depends-on How to Audit a Page — verify rendering, hierarchy, accessibility, and responsive behavior for each page.
- pairs-with Product Design — apply the shared hierarchy, interaction, accessibility, and performance thresholds.
- applies-to Evolution — an example of a hub and four instruments that must work both as direct-entry lessons and as a recommended route.
Questions
What must a reader carry from one page into the next for this to be a journey rather than a folder?
- Which transition currently has the weakest information scent?
- Where does navigation imply an order the teaching does not justify?
- What observable action or teach-back proves the reader is ready to continue?