Poll the whole lecture hall. No accounts, no cap, no cost.

The free-tier ceiling problem

Live polling tools are cheap to try and expensive to actually use, because their free tiers are metered on the one number a lecturer can't control: how many people are in the room.

Every polling tool demos beautifully. Then you run it on a real class. Free tiers in this category typically stop somewhere between 10 and 100 participants, cap you at two or three polls per session, or — the harshest pattern — meter participants monthly, so a single 50-person lecture exhausts an entire month's allowance on day one.

The result is a tool that works in a seminar and fails in a survey course, which is exactly backwards: the 200-seat hall is where you have the least idea whether anyone is following you. So the check-for-understanding either becomes a line item in the department budget, or it quietly stops happening.

The second tax is on students. Most tools want them to install something, make an account, or hand over an email — a two-minute detour, times 200 people, every single time, before a single question is answered.

How Live Poll is built differently

Live Poll has no participant cap on any plan and no student-side account, because it never creates a student identity in the first place. Answers are written as anonymous aggregate counts, so scale costs almost nothing to serve.

There is no roster, no join list, no per-student row, and no login on the student side. A student's phone posts a single anonymous count against one option and gets nothing back but the distribution you choose to reveal. The correct answer, when you set one, is never sent to a student device at all.

That architecture is the whole reason the free tier can be generous. Because a participant is not an account to provision or a connection to hold open, the 201st student costs essentially the same as the 21st. The only limit is a spam guard that stops counting at 3,000 responses per question — roughly fifteen times the largest lecture hall most people will ever teach in.

Typed multiple-choice polls are unlimited and free forever. You build as many questions as your lesson needs, run them in any order, and set a timer between 15 and 300 seconds per question.

Anonymous by architecture, not by policy

Live Poll stores no student identity, so there is no per-student record to leak, subpoena, or accidentally show the class. Answers exist only as aggregate counts, and ended sessions are deleted automatically after 24 hours.

"Anonymous" usually means a vendor promises not to look. Here it means the data does not exist: no roster, no names, no emails, no device fingerprint tying an answer to a person. A student's response is a count in a column, indistinguishable from the 213 others.

This matters for honesty of the signal, not just for privacy. Students who suspect a poll is graded, tracked, or visible to peers answer strategically — they pick what they think you want to hear. Anonymity is what makes a formative check actually formative, and it is the reason a wrong-answer cluster is worth reading.

Ended sessions, their questions, and their answers are purged on a 24-hour schedule. If you want the summary, screenshot it. If you want per-student comprehension history across a term, that is a different job — use Pulse Check, which is built for exactly that and keeps the roster on your side of the room.

Misconception Radar

When a wrong answer attracts a large cluster, Misconception Radar reads the distribution and explains the reasoning error behind it, then turns that into a reteach prompt you can use immediately.

A distribution tells you that 41% of the room picked B. It does not tell you why, and the why is the entire teaching decision. A cluster on one distractor is almost never random — it is a shared mental model that is wrong in a specific, nameable way.

Radar takes the question, the options, the correct answer you marked, and the actual response shape, and names the likely confusion — for instance, that students are conflating a mechanism with its most visible outcome. Then it gives you a prompt to reteach against, while the class is still in the room.

Radar is included with Neural Pro and Power User plans with no per-use charge. It needs a correct answer marked on the question, which is optional when you build one — polls with no right answer, like opinion checks, simply skip it.

Where Live Poll fits the rest of the suite

Live Poll answers "does the room understand this right now?" Pulse Check answers "who specifically is falling behind over the term?" They are deliberately different tools with different privacy models.

Reach for Live Poll in a lecture hall, a workshop, or any room where students have phones and you want an honest, unattributable read mid-lesson. It scales to any size and needs nothing from you but a projector.

Reach for Pulse Check in a class you know by name, where phones are restricted and you want a per-student record that builds into a comprehension timeline. You mark the roster yourself; students never touch a device.

From a wrong-answer cluster to tomorrow's lesson

A Live Poll finds the gap in real time. The rest of the NeuralRun suite turns that gap into review material without retyping the topic anywhere.

Once Radar names the misconception, hand the topic to NeuralRun Quizzes for source-grounded review questions, print them as a handout through Mind Prints, or split the room into support pairs with Smart Lists.

New to formative checks? The classroom exit tickets guide covers when to check for understanding and what to do with the answer.