A POLL SHOULD EARN YOUR TRUST
The OMP Difference
Clear questions. Safeguards behind the count. A process you can inspect.
OMP makes it easier to see what people were asked, how responses are counted and what the result can tell you.
01
Repeat attempts should not become extra votes.
Identity verification is required for every participant in an OMP commissioned poll.
Repeated submissions can distort a poll.
In open community polls, OMP combines signed browser identifiers, database checks and retry protection to help prevent repeat submissions from adding extra votes. People sharing a network can still participate separately.
Open-poll safeguards are browser-based, not identity verification. Clearing browser data or using another browser can bypass them.
02
A result should come with a record.
A percentage on its own is not evidence of how a poll was run.
Our poll pages show the question, recorded response counts, version and publication timing. Recorded sponsors are disclosed before voting and in results. We do not sell a preferred result or invent responses.
These records make OMP inspectable; they are not an independent collection audit.
03
The question should not change under the result.
Wording and answer choices can change how people respond.
Published approved wording is preserved. A revised question uses a separate response dataset, so answers to different versions are not quietly blended together.
Editorial wording checks help flag problems; they do not prove a question is unbiased. Answer-order randomization is not currently active.
Open polls describe the people who choose to participate. They are not representative population surveys.
Research behind these concerns
Pew Research Center has studied bogus respondents, online sample quality and wording effects. These studies do not evaluate OMP.
More about our methods
01
Simple choices. Equal weight.
Editorial review may use AI under OMP publishing standards. We do not claim that a person reviewed every poll. Published polls have two to four choices, including Not sure when present. Choices receive equal visual weight.
02
Repeat-response safeguards
We use browser-based controls to limit repeat responses. These controls have limitations and do not guarantee one-person-one-vote.
03
Your choice stays out of analytics.
No analytics or session replay is installed. Vote choices and private editorial content are not stored in browser storage or logs. A separate theme-only preference remembers Light, Dark, or System without identifying voters. No account is required to vote. Browser-linked voting history is not linked to accounts. The same hashed browser identifier can associate choices across polls and does not establish complete anonymity.
04
Private submissions.
Suggestions and commissioned inquiries go to private human review. Contact email is kept out of public poll pages for editorial review. Submitting does not publish anything; an editor can adapt a question into a separately approved public poll.
05
Retention and browser privacy.
The cookie containing your browser identifier lasts up to 30 days. Vote records and browser identifiers are eligible for deletion from the active database 30 days after poll closure; aggregate totals remain. Private submissions become eligible for deletion after 90 days. Bounded cleanup runs hourly, so deletion can take additional runs. This cleanup keeps poll wording, aggregate totals, and poll editorial history. Opening this site in a shared browser can expose its local session to other users.
06
Commissioned means disclosed.
Commissioned polls identify their sponsor in the feed, before voting, and in results. Payment cannot bypass editorial approval or change recorded responses.
Results reflect the people who chose to participate, not necessarily the wider public. Open polls are self-selected, not random or representative samples. No margin of error applies. US-first describes our editorial focus; participant location and nationality are not verified.
Your question. A closer look.
Identity verification is required for every participant in an OMP commissioned poll. Talk to us about the audience and question you need to study. An inquiry does not place an order or promise a particular result.