Content audits: what I cut first
Before I write a single new word, I want to know what's already there. Usually the answer is too much.
Whenever someone asks me to fix their content, they expect me to start writing. I almost never do. I start by reading everything that already exists, and then I start cutting.
I've run content audits at very different scales, from a single company site to web content that reached 30 million businesses at Meta. The scale changes. What I cut first doesn't.
1. Pages that do the same job
This is almost always the biggest win. Over time, companies publish three versions of the same page. Each one came from another team or another year, with a slightly different message. Readers don't know which one to trust, and search engines don't know which one to rank.
Pick the best one, merge in anything useful from the others and redirect the rest.
Duplicates also create maintenance problems. When pricing or a product detail changes, someone has to remember to update every version. They rarely do, and that's how a site ends up contradicting itself.
2. Content with no reader
Every page should have a specific person it's for and a specific thing it helps them do. If I can't name both, the page probably exists because someone thought it should, whether or not anyone needed it.
These pages are easy to spot once you look. They often have vague titles, very little traffic and no clear next step. Some were written for an internal stakeholder. Others supported a campaign that ended years ago.
3. Anything that's quietly wrong
Old pricing, retired features, outdated statistics, a team page full of people who left. Each one is a small hit to credibility. Together, they make the whole site feel abandoned.
This category matters most in industries where trust is the product. In health care, finance or security, an outdated claim is more than embarrassing. It can create real risk.
“Every page should have a specific person it's for and a specific thing it helps them do.”
4. Copy that talks about us instead of them
This one is more of an edit than a cut. Pages that open with company history, mission statements and "we're passionate about" language. Readers came with a problem. Start there.
A quick test: Count how many sentences on a page start with "we" compared with "you." If the company is the main character, the page probably needs a rewrite.
How I run it
You don't need fancy tools to start. A spreadsheet works:
- List every page with its URL, owner and last update.
- Pull traffic and conversion data where you have it.
- Mark each page: keep, update, merge or cut.
- Note who it's for and what it's supposed to do.
That last column is where most of the decisions get made. Once you write down what a page is for, it's easy to see whether it's doing that job.
Bring stakeholders in early
Content audits can feel personal. Someone wrote every page, and someone usually wanted it to exist. If cuts land as a surprise, you'll spend more time defending decisions than making them.
I share the audit criteria before I share the results. When everyone agrees on what makes a page worth keeping, the individual decisions become much easier to accept. Data helps too. It's hard to argue for a page nobody has visited in two years.
Why cutting comes first
New content on top of a messy foundation makes a bigger mess. Cutting first means everything you write afterward has room to work. It's also the cheapest improvement you'll ever make.
A smaller site is easier to maintain, easier to navigate and easier to keep accurate. It also gives your best content a better chance to be found.
Make it a habit
Content builds up again the moment you stop paying attention, so one cleanup won't hold. I set a regular review cycle for the most important pages and a lighter check for the rest. That keeps the site from sliding back into the same mess.
Measure what the cut did
Cutting content can feel risky, so I track what happens after. Did traffic consolidate onto the pages we kept? Did conversions improve? Did support questions drop?
Most of the time, the numbers improve. When they don't, the data tells you where to look. Either way, measuring the results makes the next audit easier to approve, because you have proof that less content worked.
Sharing those results matters too. When stakeholders see that cutting improved performance, the next round of cuts gets easier, and the team starts treating audits as routine maintenance instead of a painful cleanup project.
The short version
Before writing anything new, cut the duplicates, the orphans and the outdated. Rewrite what's left so it's about the reader, and bring stakeholders in before the cuts land. Then audit on a schedule. Less content doing more almost always beats more content doing less.