Over the last three weeks I published 13 technical articles on dev.to to promote a small side project. Twelve of them got between 0 and 10 views combined. One got 54 views, a reaction, and two real comments — more than all the others put together.
I run a small swarm of scripts that writes and publishes these articles automatically, so I actually have the raw numbers for every single one. No guessing, no vanity metrics. Here they are:
| Title (shortened) | Tags | Views |
|---|---|---|
| Your link previews are blank because og:image is missing | python, webdev, tutorial, beginners | 54 |
| Free LLM API tiers change without warning | python, llm, api, opensource | 10 |
| Custom social media share cards with Python | programming, python, tutorial | 10 |
| An autonomous agent burned $40 of API credits overnight | ai, llm, automation, agents | 10 |
| My AI agent script almost burned a $25 budget | (real-number story) | 3 |
| I shipped 9 products and got 0 sales | (real-number story) | 2 |
| Your prompt change broke 3 things you didn't test | python, llm, testing, ai | 2 |
| 6 more articles | mostly llm/agents/retry/devops | 0 |
Same writer, same effort per article, same length. So why did one post outperform the rest by 5-50x?
I went back and compared the winner against the losers line by line. Three differences stood out:
1. The title named a symptom the reader already has, not a concept they'd have to learn.
The winner: "Your link previews are blank because og:image is missing." Anyone who has ever pasted their own link into Slack and seen a blank gray box recognizes that sentence instantly. Compare that to a loser like "Your agent's retry loop is either too dumb or infinite" — technically accurate, but it assumes the reader already has a retry loop and already knows it's broken. That's two assumptions too many.
2. The tags were generic, not niche.
The winner used python, webdev, tutorial, beginners — four of the biggest tag communities on the site. The zero-view posts used llm, api, retry, devops — real tags, but tiny audiences. A post can be excellent and still get 0 views if it's tagged for a community of 200 people instead of 200,000.
3. The fix fit in one copy-paste block.
The winner's entire solution was about 15 lines of plain Python using only the standard library — no install, no signup, paste it and your social cards work. Every low-view post required either background knowledge or a multi-step setup before the payoff.
None of this is a secret formula. It's just: pick a problem almost every reader already has, use tags where that audience actually lives, and make the fix something they can copy in under a minute. The hard part isn't knowing this — it's remembering to check all three before you hit publish, not after you're staring at a 0.
I turned the checklist I now run before every article (plus the title/tag patterns and a few copy-paste CTA snippets) into a free template pack: the Article-to-Checkout Funnel Kit — it's pay-what-you-want, starting at $0.
What's the one post of yours that unexpectedly took off compared to the rest — and once you looked back, could you actually tell why?
Top comments (0)