Two years ago I started a blog content platform called NextBlog. The pitch was simple: help SaaS products get organic traffic through consistently published, well-researched blog posts. I'd been reading for years that "content is king," that "SEO is a long game," that "just keep publishing and you'll rank."

After publishing over 500 articles across a dozen different SaaS blogs, I've learned that most of that advice is dangerously wrong.

Here's what the data actually shows — and what I wish someone had told me before I started.

From June 2024 through June 2026, we published articles across 6 product blogs in the B2B SaaS space. Topics ranged from lead generation to video production to social media automation. We tracked every post through Google Search Console, ranked by keywords, impressions, clicks, and position.

Total: 512 published articles. Total keywords tracked: 8,742. Total monthly organic traffic (June 2026): ~127,000 visits across all properties.

Post frequency. In 2024, we published 3–4 articles per week. In 2025, we dropped to 1–2. In 2026, sometimes we'd go two weeks without publishing anything. Traffic kept growing through every slowdown. The volume of publishing had zero correlation with ranking improvements. If anything, the slower pace produced better-performing posts because we spent more time on each one.

Word count. The SEO playbook says 2,000+ words for competitive keywords. Our best-performing post (5,800+ monthly organic visits) is 847 words. Our longest post (3,400 words) gets 112 monthly visits. Length didn't help.

Perfect formatting. I spent hours on proper heading hierarchy, internal links, schema markup, meta descriptions. Meanwhile, a post I wrote in 20 minutes with zero formatting — just straight paragraphs — is our third-highest-traffic article. Google doesn't care about your heading structure.

We found that >80% of a post's eventual traffic was determined before we wrote a single word. The topic itself — whether people were actually searching for it — dwarfed every other variable.

The winning formula: find a question that real humans ask, that has search volume, but that existing results don't answer well. Sounds obvious, but most SEO advice is about optimizing the post, not about picking the right question to answer.

Posts where I wrote from direct experience — "we tried this, here's what happened" — consistently outranked posts about the same topics written from research alone. Google's helpful content system isn't a myth. It genuinely seems to prefer content that shows the author has done the thing they're writing about.

Our "I analyzed X and found Y" posts had a 4.2x higher click-through rate than generic "ultimate guide" posts.

Posts that opened with a specific, surprising data point held readers 3x longer (as measured by scroll depth) than posts that started with "In today's digital landscape..." or similar filler. And since Google uses engagement signals in its ranking, that matters.

The most common mistake in our early posts: trying to cover too much. A post about "how to get leads" would get buried. A post about "how I got 147 B2B leads from a single Reddit post" would rank. Specificity beats breadth every time.

Stop chasing keywords. In 2024, we built entire content calendars around keyword clusters. We'd take a head term, map out 15 supporting keywords, and write posts for each one. Traffic from those clusters was marginal. The ROI of deep keyword research was negative — the time spent finding keywords would have been better spent writing one genuinely useful post.

No more "pillar pages." The SEO industry convinced everyone they need 3,000-word pillar pages with 15 supporting cluster posts. We tried it twice. Both pillar pages get fewer than 50 visits/month. Meanwhile, a 600-word "here's a thing I built" post about a failed feature experiment still gets 1,200 visits/month.

Write for humans, not algorithms. This is the cliché advice that everyone nods at and then ignores. But the data backs it up. Posts written for humans (first-person, specific anecdotes, imperfect but honest) outperformed posts written for SEO (keyword-rich, formulaic, comprehensive) in every single metric: impressions, clicks, time on page, and — crucially — ranking improvement over time.

All of this came out of building NextBlog, a platform that helps SaaS companies research, write, and publish blog content that actually ranks. I started it because I was frustrated with SEO agencies that charge $5k/month for generic articles that never see search traffic.

The platform handles research, drafting, and SEO optimization — but the most valuable feature turns out to be the part that tells you not to write something because nobody's searching for it.

For those of you running content programs: what's the most surprising thing you've learned from your own data? I'm starting to think everything I was taught about SEO was optimized for a Google that no longer exists, and I'd love to hear what others are seeing.