Audit Basics

Setting the Scope of a Large Site Audit

How to sample and segment a big site by page type, section and importance so a large audit stays manageable and still tells you something reliable.

On this page
  1. Why scope matters more on large sites
  2. Start with the question
  3. Think in templates, not pages
  4. Segment the site
  5. Sample sensibly
  6. Use counts to find the big issues
  7. Decide depth for each segment
  8. Document what is out of scope
  9. Keep the scope stable

Why scope matters more on large sites

On a site of thirty pages you can look at everything. On a site of thirty thousand you cannot, and trying to will produce either an endless project or a shallow pass that misses patterns. Setting scope means deciding in advance what you will examine, in what depth, and what you are deliberately leaving out. It turns an overwhelming task into several small, finite ones.

Scope is also a form of honesty. A report that says “we examined these four segments using samples of about this size, and did not examine these two” is more trustworthy than one that implies complete coverage it did not achieve.

Start with the question

Before dividing the site, be clear about why you are auditing. A site-wide drop in visits, a planned redesign, a suspected technical problem and a routine health check call for different scopes. A technical question might need a broad but shallow look at every template. A content question might need a deep look at one section. Write the question at the top of your notes and let it decide what is in and out.

Think in templates, not pages

Large sites are usually built from a limited number of page types: home page, category pages, detail pages (products, articles, listings), search or filter pages (often tied to url parameters), and utility pages such as contact or policy pages. Pages of the same type are generated from the same template, so they tend to share the same strengths and flaws.

This is the key to manageable audits. A problem in a template shows up on every page built from it. You can find that problem by looking at a few examples, then estimate how widely it applies by counting how many pages use the template. Listing the page types and roughly how many pages each covers is a good first step.

Segment the site

Once you have page types, consider other ways to divide the site:

  • By section. News, support, shop and blog often have different owners, goals and problems.
  • By importance. Pages that bring visits or support the site’s main purpose deserve more attention than archives nobody visits.
  • By age. Older content may have different issues, such as outdated information or old formatting, from recent content.
  • By language or region, where the site has several versions, as with multi-location sites.

Give each segment a short name and a note on why it is in scope. Large sites often have areas that are best treated as separate audits.

Sample sensibly

For each segment, choose a sample to examine closely. There is no magic number. For a segment with a few hundred pages similar to each other, a dozen or so varied examples may reveal the main patterns. For one with many thousands, larger or repeated samples may be needed. The point is variety, not size.

Good sampling habits include:

  • Include the best-performing pages, typical pages and some of the weakest, not just the ones that are easy to reach.
  • Pick some pages at random so you are not choosing only what you already expect to see.
  • Include unusual cases, such as the oldest pages, the longest pages and those with special layouts.
  • Write down how you chose the sample, so someone else could repeat it.

If your first sample keeps turning up the same problems, you have probably found the pattern. If every page reveals something new, widen the sample or split the segment further.

Use counts to find the big issues

Alongside manual review, large audits often rely on lists generated by a crawl of your own site or exports from your content system. These let you count things: how many pages lack titles, how many return errors, how many look like duplicate content. Counts help you see whether an issue is isolated or widespread. Remember that tools see only what they can reach and are configured to see, and that any single run reflects one moment in time.

Because large sites can run into limits on how much search engines will crawl, it is also worth knowing the basics of crawl budget, which tends to matter far more for very large sites than for small ones. Even so, it is usually one factor among several, not the first suspect.

Decide depth for each segment

Not every segment needs the same attention. A simple way to plan is to give each one a level:

  1. Deep: read pages, check technical details and review links.
  2. Standard: check templates and a modest sample.
  3. Light: confirm there are no obvious blockers and move on.

Put your most important segments at the deep level, and be willing to treat low-value areas lightly. If a section has thousands of near-empty pages, the right finding may be that the section itself needs a decision, such as content pruning, not that each page needs a fix.

Document what is out of scope

Keep a short list of what you did not examine and why; it belongs in your audit reporting as much as the findings do. Include any assumptions, such as “we assumed templates in the support section match those in the help center.” This protects you from false confidence and gives a ready-made agenda for a later round.

Keep the scope stable

During an audit it is tempting to chase every interesting thing you notice. Add such discoveries to a “later” list instead of expanding the scope mid-project. If something looks serious enough to change the plan, make that choice deliberately and note it. A clear scope, even an imperfect one, keeps the work finite, and the findings you do deliver will be easier to trust and act on.

The complete guide to audit basics

How to Conduct a Website Evaluation for SEO
Read the guide

Keep exploring

Guides from other topics

All topics →