Are long-tail keywords worth it? Yes, with one caveat
You have been told long-tail converts better. You want to know whether that is true before you build sixty pages on it.
They rank better, which is the real argument
A narrow phrase has fewer, weaker competitors and a page that answers it exactly beats a general page that mentions it. That advantage is structural and it does not require authority you do not have. This is the whole case and it is enough.
The conversion claim is weaker than advertised
The received wisdom is that longer queries convert dramatically better because intent is more specific. On this dataset they clicked at about the same rate as shorter ones, and the five-to-six word band converted worst of the four. Believe the ranking argument, not the conversion one.
The tail is what compounding looks like
Nobody builds a page for a query that will produce two clicks. But hundreds of queries producing two clicks each, accumulated over months and never targeted individually, is a large share of a mature site's traffic.
Long tail is where the clicks were
Across 1,228,913 impressions on this site, the traffic that converted into 42,320 clicks was overwhelmingly specific: queries with intent already baked in, where the searcher knew what they wanted and there were three results instead of thirty. The broad terms delivered impressions. The specific ones delivered users.
The trap: long tail is not the same as low effort
The failure mode is generating a thousand near-identical pages against a thousand near-identical phrases. On this site's own programmatic set, 6 of 85 pages took 82.9% of the clicks and 22 took none at all. Long tail works when each page genuinely answers a different question. It fails the moment the pages are the same page with a word swapped.
Measured, not asserted
Across 224 days this site's traffic came from over a thousand distinct queries, of which only a handful were ever deliberately targeted. Brand queries were 42.3% of clicks and had to be excluded before any of this was legible.
Free tool for this: Search Console 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 31 of the book. Five chapters are free to read.