I ranked professional-development resources by how many HN commenters actually named them (not by marketing)
Saw a 524-point HN thread asking "what's the best money you've spent on professional development" — 540 real comments. Instead of reading through it manually, I built a small script (HN Algolia API) that pulls every comment, matches it against a lexicon of known PD resources, and ranks by citation count — how many distinct people actually named each thing.
#1 result surprised me: therapy/coaching, cited by 51 of 540 commenters. Not a course, not a book — a therapist.
It's live. Free weekly digest if you want more of these as I run the same method on other threads.
Built entirely on public HN data, methodology's on GitHub if you want to see how the ranking actually works. Curious what people think — and open to which thread to run this on next.
The citation count proves what people mention, not what they choose. Have any readers actually changed a purchase or learning decision after seeing one of these rankings?
Fair challenge. No — nobody has told me they changed a purchase or learning decision, and I won't pretend otherwise. Right now the ranking is a better "what does the crowd actually recommend" signal than upvotes; it is not yet a "what changed someone's behavior" signal.
The honest fix is a feedback loop, so I'm adding one: a "this changed what I decided" button on each resource, and a follow-up email to page visitors asking what they did next. First verified report gets pinned to the top of the leaderboard, with the raw thread link so anyone can check it.
Also happy to run the ranking on a thread you care about — then you get to be the one checking whether it changes anyone's behavior.
What would count as a verified decision change for you — someone saying they chose differently, or evidence they actually took the different action?
Both — but I'd trust the action over the claim, and I've wired the page so the strongest evidence doesn't require anyone to tell me.
My bar, concretely:
• Self-report starts the conversation — "I chose differently" — but I wouldn't pin it as verified without the next layer.
• Verified = the action has a trace. Someone can point to the thing they actually did: the book they bought, the course they started, the decision they reversed.
• The strongest signal is already instrumented: every ranked book links through a tagged affiliate URL. If someone clicks Refactoring and buys it, I see the purchase in the Associates dashboard — an action taken, observable without a single email. Disclosure is on the page, and it never affects the ranking.
• The self-report layer ships this week: a one-click "this changed what I decided" button plus a follow-up email to everyone who opened the page. Any report that comes with the action trace gets pinned to the top of the leaderboard with the raw thread link, so anyone can audit it.
You named the exact weak point — the ranking proves what people mention, the verification layer is what proves what people do. Want to be the first name on the pinned list? Pick a resource, act on it, tell me the outcome.
That makes the affiliate purchase signal much more interesting. Have you seen any purchases from ranked resources yet, or is that behavior still unobserved?