Decision Ladder Navigation (DLN)

What Decision Ladder Navigation is

Decision Ladder Navigation is a way of deciding what each page on a website is for. Instead of filing pages by department or by position in a sales pipeline, it files them by the decision the visitor is making at the moment they arrive. Five rungs cover that decision, and every page sits on exactly one of them.

Surplus information at the wrong rung is friction. A pricing page stuffed with testimonials answers a question the visitor already settled, and leaves the one they came with unanswered. A missing answer stops the climb altogether. The framework exists to make both failures visible before a visitor finds them.

Two operating rules follow, and they are the whole method in practice:

1. One decision role per URL. A page that needs two rungs is two pages.
2. One next step per page. A page with no route onward is a dead end.

The diagnostic question for any existing site is short. Does each main URL have exactly one decision role yet?

The five rungs: Orient, Choose, Prove, Rate, Act

The five rungs of Decision Ladder Navigation each name a question the visitor is holding, a page type that answers it, a role in the structure, and a failure that is common enough to be worth naming. Read the table as a diagnostic. If you cannot place one of your pages in a single row, that page is doing two jobs.

Evidence at Prove has to be specific and falsifiable: a number, a date, a named situation, a method someone could repeat. Testimonials without context are decoration, not proof.

Why Rate means price and not rating

Rate is a noun here, as in a rate card. It means price and cost structure. It does not mean rating, reviewing or evaluating a provider. That work belongs to Prove, which is a separate rung with a separate page type and a separate failure mode.

This is the single most common misreading of the framework in English, and it is worth stating on every asset that names the rungs. A reader who files trust questions under Rate will build a pricing page full of testimonials and no actual price, which is the exact failure the framework exists to prevent.

The split matters because trust and price are different questions. A model that files both under one late stage gives you no way to tell them apart when something breaks.

The four rails that turn pages into a path

Rungs describe pages. Rails describe the links between them, and the rails are what turn a set of pages into a path. Decision Ladder Navigation names four primary rails and one branch, and a site that has the rungs but not the rails is still a pile of pages with better labels on them.

O --> C --> P
      |     |
      |     +--> back to C, or forward to R
      +--> R --> A

The four primary rails are Orient to Choose, Choose to Prove, Choose to Rate, and Rate to Act. Prove is the branch that returns. It is never a destination.

Three rules of link discipline follow:

• Descriptive anchors only. No "click here", no "read more", no "learn more". Target under 20 percent generic anchors across the site.
• Every link has a stated role. A link with no role does not get built.
• In-content links outrank navigation and footer links. Sitewide footer links to money pages are a weak signal and must not be used to push authority. Footer is utility navigation, five to seven links.

A decision journey is a loop, not a chute

The journey that Decision Ladder Navigation describes is a loop. A need is triggered, options are considered, evidence is checked, a purchase is made, and post-purchase questions start the cycle again. Nothing about that shape guarantees that a visitor enters at the bottom and travels upward in order.

Three consequences follow for structure. Every page is an entry point, so the structure has to work from any starting rung. The structure has to offer a way back, not only a way forward. And after Act, the customer's first question is an Orient question about something else, so the ladder restarts rather than ending.

The design test is one sentence. If a visitor is allowed only two clicks, do they land on the right page?

The four layers DLN decides

Decision Ladder Navigation runs through four layers of a website rather than one. The menu is the interface, the information architecture is the structure, and the framework is the decision logic sitting behind both. A site that applies it to only one layer will contradict itself at the others.

The letters O, C, P, R and A are internal vocabulary. They never appear in menu labels or anywhere on the front end. Nobody arrives at a website looking for rung C.

Cluster architecture: pillar, tier one, tier two

A cluster in Decision Ladder Navigation has three layers, and each layer exists to solve a problem created by the layer above it. The pillar carries the decision. Tier one takes the things the pillar cannot fully carry. Tier two explains the terms tier one assumes.

The pillar is the landing page or money page the cluster exists to support. Tier one is the Choose, Prove, Rate and Act articles, each resolving one thing that would otherwise make the pillar unreadable: trust, value, and the friction that stops a decision. Tier two is the Orient articles, the definitions and concepts a tier-one article assumes but cannot stop to explain.

Four link rules govern the cluster, in order of importance:

1. Everything links up. Every tier-one and tier-two article carries a contextual link back to its pillar. No orphans.
2. The pillar links down selectively. Only to articles that increase the pillar's own credibility, placed next to the claim each one supports.
3. Selection is the discipline. A pillar that links to its whole cluster spreads attention thinly and leaks visitors out of the page where the decision happens.
4. Tier two links to its tier-one parent as well as to the pillar. It exists to make that specific article comprehensible.

Build the link map before writing anything. Five columns, and every link fills every column: anchor text, context sentence, from URL, to URL, and rung.

Why Orient sits at the foundation of a cluster, not the top

Orient is the foundation layer of a cluster and not its top, which is the point readers trained on the funnel get backwards most often. In a funnel, top of funnel sits above everything else. In Decision Ladder Navigation, Orient articles are tier two: the widest and most searchable material, existing to make the layer above them comprehensible.

The practical consequence is a rule about what gets written and when. Do not write an Orient article that supports nothing. Tier two exists because a specific tier-one article assumes a concept it cannot stop to explain, so the tier-one article comes first and the definition follows it.

What to measure: one metric for each rail

Measure the step, not the sale. Decision Ladder Navigation gives one primary metric per rail: the share of visitors on a rung who take the next rung's action. Four rails produce four numbers, and those four numbers tell you which part of the structure is failing.

• Orient to Choose rate
• Choose to Prove rate
• Choose to Rate rate
• Rate to Act rate

This produces a diagnostic the funnel cannot. When Choose to Rate holds but Rate to Act collapses, the problem is the price structure of the offer, not traffic and not trust. Change one variable at a time and read results 14 to 28 days later.

Four audit checks belong alongside the metrics. No URL with zero inlinks. Generic anchors under 20 percent. No two URLs receiving the same anchor from the same entity keyword. No article without a next step back to its pillar.

How DLN relates to Rasmussen's decision ladder

The term decision ladder already exists in another field, and the debt is stated here plainly rather than left for someone else to raise. Jens Rasmussen introduced the decision ladder in the 1970s as part of cognitive work analysis, a framework for studying how skilled operators reason inside complex technical systems such as power plants and aircraft.

His ladder maps the cognitive states an operator moves through, and its central insight is that experienced operators do not climb every rung. They shortcut across the ladder when a situation is familiar. Decision Ladder Navigation borrows that insight and nothing else.

Rasmussen models an operator's internal cognition inside a control system. This framework models a buyer's decision across a set of web pages. The unit differs, the domain differs, and so does the output: Rasmussen's ladder produces interface designs, and Decision Ladder Navigation produces menus, pillars, clusters and link rules. The two are different entities, not one entity described twice, and this page declares that in its structured data with disambiguatingDescription.

Naming and terminology rules

These naming rules exist so that a copy of the framework identifies itself as a copy. Every one of them is checkable by a reader who has the original in front of them, which is the point. A framework that cannot be told apart from a paraphrase of itself has no priority date worth defending.

Rights: what you may do without asking

Everything in the left column below you may do without asking, and that is the whole point of publishing the rights rather than leaving them implied. All rights are reserved on this page and on every page of this framework. There is no Creative Commons licence anywhere in it. The rights split along one line, and the line is the reason a framework spreads: people spread it by applying it, not by republishing the author's words.

Applying a method does not create a derivative work, which is why paid client work sits in the free column and stays there. Removing that permission would cost nothing in protection and would lose the group with the most reach.

Copyright covers expression, not ideas. Nobody can be stopped from reading this, building their own five-stage model and calling it something else. What is protected is the name, the words and the priority date, and that is true of every possible configuration. It is the same principle that allows this framework to borrow from Rasmussen.

Permission is given readily and in writing. Ask at the contact page and say which pages and which use.

Framework specified by <a href="https://danhnolan.com/en/dln/">Danh Nolan (Decision Ladder Navigation&trade;)</a>. DOI: <a href="https://doi.org/10.5281/zenodo.22843346" target="_blank" rel="noopener noreferrer">10.5281/zenodo.22843346</a>. ORCID: <a href="https://orcid.org/0009-0007-7906-5091" target="_blank" rel="noopener noreferrer">0009-0007-7906-5091</a>.

Frequently asked questions about the framework

These are the questions asked most often about the framework itself rather than about implementing it: whether it replaces the funnel, whether the one-rung rule is really absolute, where the argument against other models belongs, and whether any of it applies to a site that sells nothing.

Is DLN a replacement for the marketing funnel?

It replaces the funnel as a way of structuring a website, and it does not replace it as a way of planning outbound demand generation. The funnel tells you where someone sits in your process. Decision Ladder Navigation tells you what they are deciding, which is the question a page has to answer.

Does every page really need exactly one rung?

Yes, and the test is practical rather than theoretical. If a page needs two rungs it has two opening paragraphs, two audiences and two next steps, and every visitor gets the wrong half first. Splitting it into two pages costs one URL and settles the question permanently.

Where does the comparison against other frameworks belong?

The comparison against other frameworks belongs on the Choose-rung article, not on this page. This page is the definition and it stays neutral, because a definition that argues is harder to cite and easier to dismiss. The argument lives in the Choose-rung article, which sets out selection criteria and states which framework is wrong for whom.

Can I use DLN for a site that sells nothing?

Yes. Rate is the cost rung, and cost is not always money. On a site that publishes rather than sells, the Rate rung answers what the work will cost in effort, time and rework. Skipping the rung because nothing is for sale leaves the reader without the one number they use to decide.

What is the step-up rate in DLN?

The step-up rate is the DLN metric for progress between rungs: clicks to the next rung's page divided by visits to the current rung's page, times 100. It is measured per rail, for example from Orient to Choose or from Rate to Act.

How many conversion events should a DLN page have?

A DLN page should have one primary GA4 key event, such as submit_form, click_call or booking_click. One URL with one primary event keeps each conversion attributable to the page and rung that asked for it.

How are Search Console queries classified in DLN?

In DLN, Search Console queries are classified into five buckets, one for each rung: Orient, Choose, Prove, Rate and Act. The buckets show where a page attracts queries from a different rung than the one it was built for.

What is the difference between DLN and OCPRA?

DLN, Decision Ladder Navigation, is the framework, and OCPRA is the sequence of its five rungs: Orient, Choose, Prove, Rate and Act. Both names are unregistered trademarks of Danh Nolan and carry the TM symbol on first mention.

How should DLN be cited?

DLN should be cited by its DOI, 10.5281/zenodo.22843346, with Danh Nolan as author. It is published under CC BY 4.0, so anyone may apply, teach or quote it with attribution.