Before you commission one more post, it is worth knowing which of the ones you already own are actually working. Most devtools teams have no idea. They have a blog with forty or eighty posts on it, a general sense that content is either paying off or not, and no page-by-page read of which pieces earn their place. A content audit fixes that, and it almost always changes what you write next.
I ran exactly this for Grepr, who had 83 posts live and were deciding what to produce next purely on gut feel. The audit turned that into a ranked, specific list, and the next quarter of content got pointed at what the data said was working instead of what felt good. You do not need 83 posts to benefit. If you have more than a dozen and you cannot say which three are carrying the rest, you are overdue.
Score every post on two axes
The simplest useful audit sorts each post by two questions. Does it earn attention, meaning does it rank, pull impressions, or get real engagement. And does it do anything for the business, meaning does it move readers toward a signup or a trial.
That gives you four buckets. Posts that do both are your winners, and they tell you what to make more of. Posts that pull traffic but convert nobody are popular and useless, usually too top-of-funnel or aimed at the wrong reader. Posts that convert well but get little traffic are hidden gems worth promoting and expanding. And posts that do neither are candidates for the chopping block. Sorting your library this way takes an afternoon and immediately tells you where the next effort should go.
Find the decayed winners
A specific thing to look for is the post that used to rank and has slipped. Search results are not static, the topic moved on, a competitor published something better, or your own information went stale. These are some of the best opportunities you have, because the page already proved it can rank and just needs updating rather than building from nothing.
Pull the posts whose rankings or traffic have declined over the last year and check why. Often the fix is an hour of updating rather than a week of building something new, and it recovers lost ground faster than anything you could write fresh. A regular refresh pass on your decayed winners is some of the best-spent time in a content program.
Decide what to kill, and be honest about it
The hardest part of an audit is retiring things. Founders get attached to posts, especially ones they wrote themselves. But a pile of thin, dead pages can quietly hurt how search judges your whole site, so a post that gets no traffic, holds no links, and would embarrass you if a sharp engineer read it today is a liability, not an asset.
Before you delete anything, check for value worth keeping. A dead post with a couple of backlinks should be redirected to a relevant live page so you keep the link equity. Two overlapping mediocre posts on the same topic are usually better merged into one strong one. Only the genuinely worthless, the pieces with no traffic, no links, and nothing to salvage, get removed outright. Done carefully, pruning the dead weight makes the survivors rank better.
Read it for intent, not just traffic
The trap in any audit is leaning on pageviews because they are the easy number. A post can pull ten thousand visits and send nobody toward the product, usually because it ranked for something your actual buyers do not search when they are close to a decision. That page looks like a winner on a traffic chart and is doing less for you than a quiet comparison page that converts a handful of high-intent readers a week.
So score for usefulness and intent. Which posts bring in the people who could actually buy, and which just bring in volume. The measurement setup that lets you see that difference is the one I describe in measuring the PLG funnel, and it is what turns an audit from a traffic report into a real decision.
Turn the audit into a plan
The audit is only worth doing if it changes what you do next. By the end you should have four lists: posts to scale and build clusters around, posts to refresh, posts to merge, and posts to retire. That becomes the first few weeks of your content engine backlog, which means your next quarter starts from evidence instead of a blank calendar.
Run this before you brief a single new post. Most teams discover that a handful of pieces are doing almost all the work, which tells them exactly what kind of content to make more of, and that a long tail of forgotten posts is quietly holding them back. If you want someone to run the audit and act on it, that is the work I do, and you can tell me how big your back catalog is.
