Scan, survey, analyse

Why a full content audit is hard

Reviewing every page on a website is one of the most useful things a content team can do, and one of the hardest to finish. The difficulty isn’t effort or skill. It’s the shape of the job.

Start with scale: thousands of pages, often across several websites, many of them older than anyone now on the team. Then there’s the time it takes, squeezed between everything else the team has to publish. Consistency is harder still: when several people share the work over weeks, two of them will judge the same kind of page differently, and nobody judges page 3,000 quite as they judged page 30. And the website keeps changing while you work, so the first pages reviewed are out of step by the time the last ones are done.

Underneath all of that is one structural problem. A single pass through the pages is asked to answer three different questions at once: what’s here, should it stay, and how could it be better? Those questions need very different amounts of attention, and the most demanding one, how to improve a page, is wasted on a page that should simply come down. Separating the three is what makes the whole job manageable.

The scan: knowing what you have

You can’t look after content you can’t see. Most organisations know their main pages well, but a website built up over years holds more than anyone can keep in their head: sections that belong to a past campaign, pages from a merged team, a second website that grew up alongside the first.

The scan finds every page on each of your websites, so every decision that follows starts from the full picture rather than the pages people happen to remember. It leaves out addresses that aren’t pages, such as file downloads and log-ins, and you can add rules to leave out anything else that doesn’t belong. Add your website explains how.

And because websites evolve, the scan runs again every month. New pages join your Inventory at each monthly check, and pages that have gone are marked as gone, so the picture stays true without anyone having to rebuild it.

The survey: deciding what earns a place

Not every page needs improving. Some need retiring. Knowing which is which is the most valuable early decision you can make, because it frees your attention, and your plan, for the pages that matter.

The survey reads every page and notes what it is, who it’s for, how long it is, how old it is and whether it has spelling mistakes. From that, it recommends keeping, revising or retiring each page, with a short note of why. Those recommendations follow retention settings you can change: by default, news is kept for six months and job and volunteer adverts for three.

What the survey sees well is the plainly outdated and the trivial: the job advert whose closing date has passed, the event that has happened, the page that says something is still open when it’s over, the page with next to nothing on it. It reads the dates each page gives and checks them against today.

Redundant content is subtler. A page can be accurate and well written and still not be needed, because your strategy has changed, because the issue has moved on, because it was made for reasons that didn’t hold up, or because another page says the same thing better and is better placed to say it. The survey doesn’t know your strategy. What it can do is sort every page by topic and audience segment, so Results → Estate shows how much of your website sits under each one. That makes the question visible: if a fifth of your pages serve an audience segment you no longer focus on, you can see it, and decide.

The value of the survey is that it turns thousands of separate questions into a handful of decisions you can make as a team: how long news should stay up, whether a topic still belongs, which old guides still hold. It recommends; you decide. Planning a big change? covers working through the recommendations. Because the survey runs every month, a page that goes out of date is caught when it happens, not at the next big review.

The in-depth analysis: improving what matters

Once you know which pages you’re keeping, the in-depth analysis tells you how to make them better. Each page is scored against our 15 criteria, from meaningful and readable to up to date and accessible, with examples from the page and what to change. That evidence makes several things possible that are hard to do any other way.

Fixing what matters first. Scores show which pages, and which criteria, are weakest, and the examples point to the exact sentences. A team with limited time can start with the important pages that are furthest from where they should be, rather than with whichever page someone happened to open.

Making the case. “This page is hard to read” is an opinion. A low readability score, with the page’s own words quoted and what to change, is evidence. It helps when the page belongs to another team, or when improving content has to compete for budget.

One standard, applied the same way. Every page is judged against the same criteria, in the same way, every time. That gives a team a shared definition of good content, so writers know what they’re aiming for and reviewers aren’t relying on taste. Once you’ve set out your brand and strategy, the on-brand and strategic criteria judge each page against them too, which is where the question of whether a page still fits your direction gets a clear answer.

Seeing content health improve. Across all the pages you analyse, the results show how healthy your content is as a whole and what to focus on next. Over time, they show whether it’s getting better.

Pages change, and so does the world around them: a fact goes stale, a service changes, someone edits a page in a hurry. Each month, pages that have changed are compared with what was last scored, and analysed again if anything that bears on their scores has moved. So the scores describe your content as it is now, the trend line means something, and when your team improves a page, you see it. Changed a page and don’t want to wait? Re-check it straight away.

To start, the Inventory suggests a list: the pages the survey recommends keeping or revising, most important first. Use it as it is, adjust it or choose your own.

Why the order matters

Each layer makes the next one smaller and better aimed. The scan turns your websites into one complete list. The survey turns the list into decisions, and sets aside the pages that don’t need a close look. The in-depth analysis then spends its attention where it changes what people read. Everything gets looked at; the closest look focuses in on the content that counts.

Questions people ask