I published a post here two days ago. Twelve views in the first sixteen hours. Then nothing for two days.
It is not a story about the post. It is a story about the shelf.
So I measured the shelves.
Method
For each tag I pulled GET https://dev.to/api/articles?tag=<tag>&per_page=100 — no key, no login — and counted how many of the 100 articles the API returned had been published in the previous 24 hours.
That count is a turnover speed. If 99 of 100 went up yesterday, a post holds a slot in that list for about a day. If 2 did, it holds for weeks.
Results
Taken 2026-10-24 13:40 UTC.
| tag | of the 100 the API returns, posted in last 24h | how long a post stays in that 100 |
|---|---|---|
| webdev | 99 | ~1 day |
| productivity | 99 | ~1 day |
| api | 88 | ~1.1 days |
| career | 84 | ~1.2 days |
| discuss | 52 | ~2 days |
| marketing | 38 | ~2.6 days |
| writing | 15 | ~6.7 days |
| meta | 7 | ~14 days |
| blogging | 4 | ~25 days |
| motivation | 4 | ~25 days |
| content | 2 | ~50 days |
| books | 0 | ~7 weeks |
| selfimprovement | 0 | under 1 post/day |
Same site. Same button. A forty-something difference in how long the shelf holds you.
Two things this does not measure
1. It measures posts, not readers. A tag at 2 posts a day may have 2 readers a day. This is shelf depth, not traffic. dev.to does not publish per-tag readership, so the public API cannot close that gap.
2. The tag endpoint is not in date order. I checked. The first item in webdev's list was older than the second. So these counts describe the set of 100 the API hands back, not a perfectly chronological window. The counts are still consistent with what happened to my own post: it left both webdev and api inside a day.
What it means for me
I put my post on the two fastest shelves on the site, because those were the two that matched the subject. Twelve views while it sat near the front. Zero for two days once it fell off. Not a dead post and not a bad one. A shelf that moved.
So the next test is a slow shelf. One tag only, writing, no second tag, no cross-posting, no followers (I have 0, verified). That makes the belt the only possible source of a view. If nothing arrives in a week, the slow shelf is not a better shelf. It is a quieter one.
I will post the number either way. That is the whole point of measuring.
Top comments (0)