One honest decision per URL, keep, refresh, merge, retire, or create, with the demand that already works protected on the way through.

A site can carry a huge page inventory and still leave important questions unanswered. Finding keywords is the easy part. The difficult decision is which pages still have a job.

We decide what to keep, refresh, merge, retire, or create using page-level evidence and the value each page carries for your business.

Two Zeo figures sorting a tall stack of webpage cards into trays marked keep, merge, refresh, and retire.

Some of the 500+ brands we've worked with

See all references
  • Bayer
  • Tazedirekt
  • Isuzu
  • Gedik Yatırım
  • Bundle
  • Sportive

We treat the page inventory like a living product.

  1. Build the current page inventory

    We reconcile crawl, CMS, canonical, analytics, and ownership data into one list (including the pages your navigation has forgotten).

    One current inventory where every URL has a template, locale, owner, purpose, and lifecycle state.

    AI assist
    Reconciling crawl, CMS, canonical, and analytics exports into one page list and surfacing the orphaned URLs navigation forgot.
    Human gate
    A specialist assigns the owner, locale, and lifecycle state that the reconciled data can't infer on its own. Inventory ownership assigned per URL
  2. Connect queries with the pages people reach

    We join Search Console, analytics, conversion signals, and current result pages at the same URL and time window. This keeps us from blaming a traffic decline on the wrong change.

    A map showing the demand each page currently owns, with uncertain matches clearly identified.

    AI assist
    AI compares each page's current ranking position with the pages winning the same query and flags pages that may be competing for identical demand.
    Human gate
    The content lead separates genuine cannibalization from the uncertain matches that do not justify touching anything. Demand match confidence checked
  3. Choose a job for every page

    A specialist checks intent, unique evidence, authority, internal links, performance, and risk before assigning keep, refresh, merge, redirect, retire, reframe, or create.

    A page-level decision list with a reason beside every call.

    AI assist
    Pre-scoring each page against the decision checklist, intent, evidence, authority, links, so review starts sorted instead of blank.
    Human gate
    A specialist makes the actual keep, refresh, merge, retire, or create call, weighing risk the score can't see. Page action decided with a reason
  4. Turn the decision into a usable brief

    We turn each accepted decision into an owned edit, merge, redirect, or new-page brief with factual sources, dependencies, and a way back if it underperforms.

    A backlog editors and developers can ship without reopening the audit.

    AI assist
    Drafting the edit, merge, or redirect brief from the decision's source facts and dependencies so editors start from a document.
    Human gate
    A specialist sets the rollback condition and checks the brief against the decision before it reaches the queue. Brief approved for the queue
  5. Watch changed pages as cohorts

    We compare kept, refreshed, merged, redirected, retired, and new pages against the same baseline, and keep useful losses visible instead of hiding them inside totals.

    A clear keep, adjust, restore, or stop decision for each type of content action.

    AI assist
    Aggregating each changed-page cohort's performance against the same baseline automatically, cohort by cohort.
    Human gate
    A specialist decides keep, adjust, restore, or stop for each cohort so a losing page can't hide inside a winning total. Cohort verdict recorded per action

AI reconciles and pre-sorts; a specialist makes the call.

AI reconciles crawl, CMS, canonical, and analytics exports into one page list, surfaces the orphaned URLs navigation forgot, compares ranking positions on shared queries, pre-scores pages against the decision checklist, drafts the edit, merge, or redirect brief from the decision's source facts, and aggregates each cohort against the same baseline. Nothing useful is deleted, redirected, canonicalized, or noindexed without the source owner's approval, evidence of the traffic being protected, and a recoverable path.

The next editorial cycle gets easier to run because of what comes out of this one.

  • Content calendar

    Page action inventory

    Accepted when

    Every priority URL has one action, one owner, and one reason (no orphan rows or mystery duplicates).

  • Architecture map

    Intent & coverage map

    Accepted when

    The map shows which page owns each search task, where pages genuinely overlap, and where the site still lacks a useful answer.

  • Prioritized backlog

    Consolidation & creation backlog

    Accepted when

    Source facts, the affected page, the outcome being protected, and the condition for restoring it, on each item.

  • Tracking plan

    Editorial validation plan

    Accepted when

    Changed-page cohorts are measured separately, preventing a large total from hiding a useful page that has lost performance.

We call it done when: The action inventory, coverage map, backlog, and validation plan are done when every priority URL has one action, one owner, and one reason, and every merge or retirement names the outcome it protects and the condition that restores it.

This is for teams ready to make page-level decisions.

A good fit when

  • Your content library has grown so quickly that no one can clearly explain the purpose of every page.
  • Several pages compete for the same searches, but you need to avoid merging the wrong one and losing useful demand.
  • You need a defensible plan for publishing and consolidation, grounded in product truth, observed queries, and the editorial capacity you have available.

Better handled as other work when

  • You only need a keyword list or a batch of AI rewrites, without assigning each page an owner or a clear purpose.
  • You want to delete, redirect, or noindex useful pages without protecting the traffic, facts, and customer journeys they already support.

If one of these is closer to your situation, start here instead: SEO Audit

We call it done when: Every priority URL carries a clear action and an owner, every merge or retirement has a safe destination and recovery path, and every new page has a distinct job. The clutter comes out without sacrificing the useful demand you've already earned.

  • Ahrefs

    shows which pages carry links and rank beyond recent analytics

  • Semrush

    compares audited pages with current competitors and missing topic coverage

  • Screaming Frog

    builds the working URL inventory with templates and canonical states

  • Keyword Cupid

    groups overlapping queries before merge or new-page decisions

  • Copyscape

    confirms substantial copy overlap among suspected duplicate pages

  • Google Search Console

    maps the queries and search demand each page currently owns

  • Google Analytics

    adds entrances, conversions, and journey value to page decisions

Bring your sitemap, crawl, or CMS export (whatever you've got. We'll help you find the overlap, protect the good pages, and decide what deserves the next editorial hour).
Audit what you've published

No. A low-traffic page may still support a high-value journey, own useful links, answer a specific customer question, or protect another page. Traffic is evidence. A page's job decides its future.