How to Decide Which Website Pages Should Survive a Redesign

Most redesign arguments sound like this: “We can’t lose that page, it still gets traffic,” or “Legal says we have to keep the entire archive.” Underneath the noise is a simple decision: which URLs actually deserve a seat on the new site.

Before a redesign, score each URL on five signals—search demand, business criticality, conversion role, link equity, and content quality—then keep, improve, merge, or retire pages based on that score.

You’re not just refreshing layouts. You’re reshaping the portfolio of pages that drives your pipeline, supports customers, and carries your search equity.

In this piece we’ll walk through a practical way to do that, so you can stop arguing page by page and start making fast, evidence-backed decisions.

---

Clarify the real decision: you’re not redesigning pages, you’re redesigning a portfolio

Most teams start a redesign by talking about templates and hero images. The real leverage is one level lower: which individual URLs should survive at all.

We have noticed that once a site crosses a couple hundred pages, no one truly owns the portfolio. Campaigns, microsites, and one-off blog posts accrete over years. Then a redesign shows up and every legacy choice comes due.

This is a later step on your Buyer Maturity Path:

• First, you decide whether the site needs targeted fixes or a full rebuild. If that’s not settled, pause here and read How to Decide Between Website Repair and Redesign as a prerequisite perspective: related guidance on how to decide between website repair and redesign.
• Once you’ve committed to a redesign, the next maturity jump is treating your URLs like a portfolio of assets, not a museum collection.

A portfolio mindset changes the questions you ask:

• From “Can we keep this page?”
• To “If this URL didn’t exist today, would we invest to create it for the new site?”

That’s why we use the compression line internally: redesign your portfolio, not just your pages.

---

The hidden failure mode: treating every existing URL as innocent until proven guilty

The most common failure mode in redesign planning is assuming that every existing page should stay unless someone can prove it’s terrible.

When that happens:

• Low-value content rides along into the new build.
• Navigation gets bloated because no one wants to delete anything.
• Search equity is diluted across dozens of near-duplicate topics.
• Conversion paths get noisy, so visitors bounce instead of acting.

In redesign planning we often see 20–30% of URLs doing almost all of the useful work—driving discovery, helping people evaluate, or capturing leads—while the remaining long tail exists mostly because no one wants the politics of deletion.

Treating every URL as “innocent” also creates bad incentives:

• Stakeholders defend “their” sections with anecdotes instead of data.
• Teams overvalue raw traffic and undervalue quiet but crucial support or pricing pages.
• Old campaign pages that point at retired offers keep capturing the wrong visitors.

The fix is not a heavier meeting schedule. It’s a consistent scoring model that everyone can see and apply.

---

A five-signal model for deciding which pages survive your redesign

To decide which URLs deserve to live on, you don’t need an enterprise content audit platform. You need a clear, shared way to judge each page.

Use this five-signal model:

1. Search demand – Are people still looking for this topic or phrase?
2. Business criticality – Does this page support core offerings, compliance, or customer operations?
3. Conversion role – Does this URL directly convert, or clearly assist, sales or lead generation?
4. Link equity – Does this page receive meaningful internal or external links you can’t casually abandon?
5. Content quality – If you landed here cold, would you trust the company more or less?

You’ll use these signals to categorize each URL into one of four actions:

• Keep – Page is high value as-is; maybe a light UX refresh.
• Improve – Topic and URL matter, but the content or structure needs work.
• Merge – Valuable topic duplicated across multiple weak URLs; consolidate into one strong asset.
• Retire – Low value, low demand, low equity; redirect where appropriate and move on.

A simple spreadsheet with one row per URL and one column per signal is enough. The scoring model doesn’t have to be perfect; it just has to be more consistent than arguing from memory.

---

Step 1: Inventory and group pages by intent before you score them

You can’t score what you can’t see. Before any debates, assemble a basic inventory.

For a typical 300-page WordPress site, a practical inventory process looks like this:

1. Export a URL list from your CMS or analytics tool.
2. Clean obvious noise like media files and system URLs.
3. Add simple labels for page type and owner where you know them.

Then group URLs by intent, not just by folder:

• Core marketing and product – home, About, product/service pages, pricing, key solution explainers.
• Conversion assets – contact, demo, quote, trial, and any form-based thank-you pages.
• Content and education – blog posts, guides, resources.
• Support and operations – documentation, FAQs, status updates, user onboarding.
• Legacy and campaigns – microsites, event pages, promotions, retired product areas.

In a real web team scenario, this is where friction starts:

• Marketing wants to wipe old campaign sections.
• Sales wants to keep old product microsites “just in case.”
• Support cares less about traffic and more about not breaking existing help links.

Grouping by intent first calms the conversation. You’re no longer arguing about 300 individual pages—you’re deciding how aggressive to be with each bucket.

As you group, note whether a URL is an obvious candidate for merge (duplicate topics, parallel pages for similar offers) or retire (campaigns from years ago, expired events).

---

Step 2: Score each page on demand and business criticality

With your inventory grouped, start scoring two signals that are relatively objective: demand and business criticality.

Create three simple levels for each signal: High (3), Medium (2), Low (1). You’re not building a PhD model; you’re creating a fast, comparable view across hundreds of URLs.

Scoring search demand

You don’t need perfect keyword data to tell whether a page’s topic is alive.

Ask:

• Does this topic show up regularly in your organic search queries or site search?
• Does sales still hear this question on calls?
• Does the industry still use this language, or has it shifted?

Examples:

• A service page describing a flagship offer in the exact language your buyers use: High demand (3).
• A blog post about a one-off event from four years ago: Low demand (1).
• A legacy product page for a still-relevant problem, but using outdated terminology: Medium demand (2) – likely an Improve or Merge candidate.

Scoring business criticality

This is where a lot of traffic-obsessed decisions go wrong. Some high-traffic pages aren’t critical. Some low-traffic pages absolutely are.

Ask:

• If this URL disappeared, would sales, support, legal, or operations feel pain?
• Is this page necessary to explain or qualify your core services?
• Does any external contract, compliance requirement, or customer onboarding flow depend on it?

Examples:

• A core pricing or plan comparison page that sales sends in nearly every deal: High criticality (3), even if the traffic isn’t huge.
• A blog post that pulls search traffic but attracts the wrong type of visitor: Low or Medium criticality (1–2).
• A detailed integration guide that only existing customers use: often High criticality (3), despite niche demand.

At this stage, you might already see patterns:

• High demand + High criticality: clear Keep/Improve.
• Low demand + Low criticality: strong Retire candidates.

Flag high/low mismatches (high demand but low criticality, or vice versa) for a deeper look in the next steps.

---

Step 3: Score each page’s conversion role and link equity

Now you add the signals most teams underestimate: conversion assistance and link equity. These are the pieces that connect your site into a working Content Neural Network rather than a pile of disconnected pages.

Scoring conversion role

Two different roles matter here:

1. Direct converters – pages where a lead or sale happens (forms, trials, bookings).
2. Assist pages – pages people consistently view in the sessions that lead to conversions.

In support work, we often see teams over-index on raw traffic to top-of-funnel blog posts and underweight unglamorous mid-funnel explainers and “how it works” pages that quietly do the closing.

Score like this:

• High (3) – Direct conversion pages and critical mid-funnel content that appears often in converting journeys.
• Medium (2) – Supporting content that sometimes appears in converting sessions and clearly nudges decisions.
• Low (1) – Pages with negligible presence in converting paths.

This is also where you distinguish between traffic-leading and conversion-assist pages:

• A broad, beginner blog post might be your traffic leader but contribute very little to leads.
• A comparison page between two approaches might get modest traffic but show up in most closed deals.

Your redesign should protect and improve the latter even if the vanity metrics favor the former.

Scoring link equity

Link equity isn’t just an SEO ranking issue. It’s how authority and context flow through your site.

Consider two sources:

• External links – Are industry sites, partners, or media linking here?
• Internal links – Does this page act as a hub that many other pages rely on?

Ask:

• Would removing or radically changing this URL break a lot of inbound links?
• Is this page a crucial connector in your internal linking, like a central guide or hub?

Score:

• High (3) – Pages with notable external links or central internal hub status.
• Medium (2) – Some meaningful links, but not irreplaceable.
• Low (1) – Orphaned or barely linked pages.

Pages with high link equity don’t automatically earn “Keep as-is” status—but they do demand careful planning: rewrite instead of delete, or consolidate with clean redirects.

---

Step 4: Score content quality and decide keep, improve, merge, or retire

The last signal is the one stakeholders feel instinctively but rarely articulate: is this content actually good enough for the site you want to own next year?

Scoring content quality

Judge quality with observable criteria:

• Clarity – Is the promise obvious in the first screen?
• Accuracy – Is anything outdated, misleading, or inconsistent with how you sell today?
• Depth – Does it answer the real questions buyers have, or just repeat surface-level marketing copy?
• UX and readability – Is it easy to scan, on mobile and desktop?
• Brand fit – Does the tone and structure match where your company is headed, not where it started?

Give each page a quality score:

• High (3) – You’d be proud to send it to a new prospect today.
• Medium (2) – Solid foundation but needs updates to match current offers or standards.
• Low (1) – Embarrassing, confusing, or clearly off-strategy.

Turning scores into actions

Now add up the five signals for a simple total score out of 15. Then apply consistent rules to categorize each URL.

Here’s a workable decision grid:

• Keep

  • Total score: 12–15.
  • Characteristics: High demand, high or medium criticality, strong conversion role or link equity, solid quality.
  • Action: Carry forward into the redesign, focus on UX and visual refresh.

• Improve

  • Total score: 9–11.
  • Characteristics: Right topic and role but uneven quality or outdated details.
  • Action: Plan content rewrites, structural improvements, or better CTAs as part of the redesign scope.

• Merge

  • Total score: 7–10, usually where several URLs cover overlapping topics.
  • Characteristics: Valuable topic, but the value is split between multiple thin or partially redundant pages.
  • Action: Choose a primary URL to survive, consolidate content there, and redirect the others.

• Retire

  • Total score: 5–8, typically low on demand, criticality, and conversion role, with minimal link equity.
  • Characteristics: Old campaigns, outdated features, or content that attracts the wrong audience.
  • Action: Set up redirects to the closest relevant page or a more current alternative; document the decision.

Don’t obsess over one-point differences. The goal is to surface obvious actions and a short list of “review manually” cases, not to prove mathematical precision.

By the end of this step, you should have:

• A spreadsheet of URLs, signals, and total scores.
• A clear action column (Keep / Improve / Merge / Retire).
• A realistic sense of how much content work must be budgeted into the redesign.

This is where redesign success starts to look like a business decision, not just a visual refresh.

---

Don’t ignore multi-stakeholder and legacy pages: how to handle politics with a scoring model

On a 300-page site, the hardest pages to decide on are rarely the high-scoring ones. They’re the legacy, cross-department, or “pet” pages with loud defenders.

Typical examples:

• A sales leader’s favorite microsite for an outdated product.
• A compliance-heavy page no one has updated in years.
• A thought-leadership series with internal fans but no measurable demand or conversions.

The point of the scoring model isn’t to erase politics; it’s to give you a neutral surface to have those conversations.

Use it this way:

1. Make the scoring visible. Review each signal in a working session so stakeholders see the same data.
2. Separate the topic from the URL. It’s often correct to keep the topic but retire or merge the current page.
3. Offer a path for sentiment. If someone insists “This matters for reputation,” treat that as an input to business criticality or quality, not an override.
4. Document exceptions. When you keep a low-scoring URL, record why and set a review date.

When a redesign has too many reasonable stakeholders all making good points, we’ve seen that teams make more progress by agreeing on the decision logic first, then applying it, rather than debating each page in isolation. If you recognize that pattern, there’s an expanded discussion of stakeholder alignment here: related guidance on when a website redesign has too many reasonable stakeholders decide this first.

The hidden cost of avoiding these conversations is quiet: you launch the new site, but the URL portfolio is still bloated and incoherent. Six months later, content governance feels just as messy as before.

---

How this decision model fits into your broader redesign plan

Your keep/improve/merge/retire decisions are not just a content exercise; they drive scope, navigation, and technical work.

Once your spreadsheet is stable for a first pass:

• Navigation and IA – Use the “Keep” and “Improve” lists as the backbone of your new sitemap, and ensure “Merge” decisions are reflected in clearer, fewer destinations.
• Content production plan – Turn “Improve” and “Merge” actions into concrete writing and editing tasks with owners and deadlines.
• Redirect map – For “Retire” and merged URLs, build a redirect plan early so your link equity and user paths survive the launch.
• Design priorities – Identify which high-value pages deserve more design attention, custom layouts, or interactive elements.

This is also where the Content Neural Network idea becomes practical: your high-scoring pages are the key nodes you cannot afford to break. Their internal links, cross-references, and redirects are what keep visitors flowing from discovery to decision.

If you’ve already been thinking about whether you even need a redesign or just targeted optimization, this URL model sits downstream of that earlier decision. For more context on that tradeoff, the article How to Decide Between a Fix and a Redesign offers an expanded contrast between incremental change and full rebuilds: related guidance on how to decide between a fix and a redesign.

To see how page-level work fits into the rest of a redesign—from timing to scope to ownership—browse the broader collection of Website Redesign articles that expand on planning, governance, and measurement: Website Redesign articles.

---

When to bring in outside help to apply this model

By this point you should have a clear sense of whether your team can run this process alone or needs a partner.

You probably can self-manage if:

• Your site is under ~150 pages and the number of stakeholders is small.
• You already have clean analytics and basic tagging in place.
• One person clearly owns final decisions on content and navigation.

You’re more likely to need help if:

• Your inventory reveals hundreds of URLs no one can confidently own.
• Stakeholder politics are stalling decisions, especially around legacy sections.
• You don’t have time or in-house expertise to translate the spreadsheet into redirects, navigation changes, and content rewrites before launch.

Delaying this decision has real costs. If you push ahead with design and development without a clear keep/merge/retire plan:

• Weak legacy pages will get new paint instead of being retired.
• Navigation will stay bloated, and visitors will still wander.
• High-value pages may lose their internal support network and quietly drop in performance.

A redesign that ignores the URL portfolio often feels like an expensive reskin—your team spends money, but lead quality, clarity, and governance barely move.

If you’d like this model applied inside a real build, the Web Design & Development team at Best Website can turn your messy inventory into a working portfolio: Web Design & Development. In a typical engagement, we help you:

• Build and refine the five-signal scoring spreadsheet.
• Facilitate working sessions to resolve keep/merge/retire debates.
• Translate decisions into a concrete sitemap, content plan, and redirect map for your developers.

If you’re looking at your current site and realize you can’t confidently say which pages deserve to survive a redesign, it’s worth a focused conversation about scope and ownership; you can start that by sharing your situation and goals through this short contact channel: start a redesign planning conversation.

Before the next website change, document and approve the ownership decision this article has outlined. Leaving that decision unresolved creates avoidable delay, rework, and production risk.