What happened
Third-party trackers watching LinkedIn's algorithm confirm that poll engagement has collapsed to a 0.07% rate since the March 2026 Authenticity Update — an enforcement wave stacked on top of the 360Brew interest-graph ranking system. The update aggressively deprioritized engagement bait, engagement pods, external-link spam, and polls specifically. Founders who once relied on polls to farm early clicks now see posts throttled within the first 60 minutes. Postory's analysis of poll posts across mid-sized creator accounts found the median first-hour impression count dropped 71% versus the prior quarter, with follow-on reach falling faster than any other native format. The reason is mechanical: a click on a poll option produces zero dwell time and no comment thread, so the ranking model reads the interaction as a low-quality signal. That treatment can then extend to the account, not just the individual poll, and the accounts that leaned hardest on polls in Q1 are now seeing suppressed distribution on non-poll posts too.
Why the LinkedIn poll engagement rate matters now
MagnetizeX builds founder visibility systems for B2B firms.
The Authenticity Update lands hardest on founders who used one-click formats to seed early engagement. Poll votes were the fastest path to pod-style distribution — that lever is gone. Dwell time is now the primary ranking input, and posts with 30+ seconds of view time consistently earn broader distribution than posts optimized purely for comments. That reshuffles which formats work: PDF carousels, native video, and narrative long-form now beat polls, image posts, and short reactive takes. Founders who repurpose sales calls, teardowns, and specific data into text and document posts get lifted, while poll-first accounts continue to stall through the fall.
- Retire polls from your calendar.
Every planned poll should be rewritten as either a PDF carousel with a specific point of view or an open-ended text post with a direct question in the last line. Comment quality replaces vote count as the signal, and dwell time recovers immediately because readers have to actually read to answer.
- Rebuild posts for 30-second dwell.
Restructure text posts so the payoff lives below the fold. Longer paragraphs, embedded examples, and visible data hold the reader past the see-more click. Polls trained founders in the opposite direction — one-line hooks with a tap answer — and those habits now cost distribution on every non-poll post.
- Audit last quarter's poll frequency.
If polls made up more than 20% of your posts since March, expect residual distribution drag on new work. Ship seven consecutive dwell-heavy posts before running any promotional content — the model needs a fresh recent history before it will re-widen your reach.
By the numbers: Poll engagement rate now sits at 0.07% versus document posts at 6.60% — a 94x delta that means every poll costs a founder roughly seven posts of equivalent reach.
What to do this week
Pull the last 90 days of your LinkedIn post-level data through Shield or Taplio and sort by dwell seconds. Identify the three posts that crossed 30 seconds of average view time, then reverse-engineer that structure into your next four posts. Delete any polls sitting in a Buffer, Publer, or Taplio queue — republishing a poll now costs more than skipping the day would. If you must ask readers a question, do it in a comment prompt, not a vote.