Should your SaaS have a blog? Only under one condition
Everyone says you need a blog, and you want to know whether it will actually return anything before you commit a year to it.
Short answer
A blog is worth it only when the posts contain something that cannot be reproduced: your data, your failures, your measurements. Posts that restate what is already on page one compete against dozens of adequate versions from stronger domains and reliably return almost nothing. On the reference dataset, 75 posts produced 2.32% of total traffic at a 0.80% click-through rate.
The measurement, on one real site
Same domain, same author, same window, same authority. Blog pages ran at 0.80% click-through and produced 2.32% of all traffic across 75 posts. The tool pages on the same site ran at 3.56%. The difference is not writing quality; it is that one answers the query by doing the thing and the other by describing it.
Why adequate is worthless
For any topic worth writing about, forty adequate articles exist already, several on domains far stronger than yours. Being one more adequate article gives a search engine no reason to prefer you and gives a reader no reason to finish. Adequate and absent produce the same traffic.
The posts that do work
The ones nobody else could write. A number you measured. A failure you can document with evidence. An internal dataset. These get linked, because linking is cheaper than reproducing them, and links are the constraint on everything else you are trying to do.
The opportunity cost is the real argument
A post is not free. It is a day of writing, plus editing, plus maintenance forever. The question is never "is a blog good" but "is this post the best use of the day", and for most SaaS teams before product-market fit the answer is a page that does a job.
If you do publish, publish fewer
Ten posts that could only come from you beat a hundred that could come from anyone. Volume strategies assume you can outproduce competitors who have more writers than you, which is a bet you lose by definition.
The internal-linking argument, honestly stated
Blog posts are sometimes justified as internal-link fuel for money pages. That works, but a page whose only job is to pass links is a page nobody reads, and it still costs maintenance. If that is the actual reason, be honest about it and build far fewer of them.
The test before you write
Name the query. Name what your post will contain that the current page one does not. If you cannot answer the second in one sentence, the post will join the long tail of pages that exist and do nothing.
Questions
- So blogs do not work?
- Undifferentiated blogs do not. The reference dataset had two posts that worked and seventy-three that did not, and the two contained things nobody else had. That ratio is the argument for being selective, not for abstaining.
- What about topical authority?
- It is real and it is slow, and it does not rescue content that is individually unremarkable. Building authority with forty mediocre posts is a strategy that assumes engines cannot tell them apart.
- How often should I publish?
- As often as you have something only you can say. A fixed cadence guarantees you will publish filler to meet it, and filler is what produced the 2.32%.
Measured, not asserted
ihatepdf.cv, the tool taken from zero to 100,542 users, ran a 75-post blog alongside its tool pages. The blog produced 2.32% of traffic at 0.80%; the tool pages produced the rest at 3.56%. Two pages on that blog carried most of what it did earn, which is the same distribution as everything else on the site.
Free tool for this: Programmatic Set Analyzer. No account, nothing uploaded.
Where this goes deeper
Every number on this page comes from one complete dataset: one product taken from zero to 100K+ users on search alone, with nothing spent on advertising. The full argument is Chapter 28 of the book. Five chapters are free to read.