We took over a B2B SaaS blog last year with 340 published posts and roughly 60,000 monthly organic visits. Sounds healthy until you look closer: 19 posts drove 80% of that traffic. The other 321 were sitting at single-digit monthly visits, some at zero, some indexed and quietly embarrassing.
The instinct in most content teams is to keep publishing. More posts, more coverage, more chances to rank. We did the opposite first: we pruned.
The prune came before the plan
We merged eleven overlapping "what is X" posts into two comprehensive ones. We deleted 40-something posts that were thin, outdated, or answering a question nobody was searching for anymore. We redirected the URLs with any residual link equity into the pages that were actually working.
Traffic didn't drop. It went up 14% in the following quarter, because we'd stopped diluting authority across pages competing with each other and started consolidating it into fewer, stronger ones.
Then we looked at who the 19 winning posts were actually for
The posts that worked weren't the ones optimized hardest for keyword density. They were the ones written by someone who'd actually done the job the reader was trying to do — a former practitioner, not a generalist freelancer working from a brief. That's a staffing decision as much as a content one. If your writers have never done the thing they're writing about, readers can tell, and so, increasingly, can search engines evaluating expertise signals.
What actually goes on the calendar now
We plan content backward from sales conversations, not forward from keyword volume. Every quarter, we ask the sales team: what are the three questions prospects ask on every single call? Those become pillar content. Then we build supporting pieces that link into them, rather than 20 disconnected posts hoping one of them sticks.
A blog with 40 posts that each get real traffic beats a blog with 340 posts where 90% are dead weight — for readers, and for how a search engine evaluates the site as a whole. If your blog has years of backlog, the fastest way to grow it usually isn't publishing more. It's honestly grading what's already there and being willing to delete the majority of it.