$15k to $75k
estimated cost of external help
  • Scan and research
  • Prioritise options
  • Confirm options
  • Consult
0-2 weeks
estimated run time
3-5 weeks
estimated lead time
Strong communicator
recommended experience
This is a process of asking participants for their preference from a range of options. This can be as simple as a yes or no answer to a question, to choosing between dozens of options. Results are usually used as evidence of a broad group’s general preference and used as a basis, or justification, for a course of action.

In 5 steps...

  1. Determine what you want to achieve, who you need to represent and how the information will be used.
  2. Determine the right method for selecting and contacting your audience. Selection could be random, or based on specific characteristics to match your target community. For some groups phone polling might be preferable. In other instances you might be able to use online polling tools to get a quick snapshot.
  3. Select your question and answer options, ensuring both are clear and require no interpretation.
  4. Open the poll for responses either until you reach you’re your required participant breakdown, numbers, or date/time.
  5. Communicate poll responses. Many online polls offer an option of showing respondents the results only after the vote, to both encourage them to contribute, and to avoid influencing their input.

When to use it

This is best used when you would like to quickly assess the alignment of a large number of people with a potential decision option. Polling can be used in a light and fun way as a part of an in-person event to complement the agenda, or be highly structured as a part of audience research.


  • Quick way to test the alignment and/or level of division on a subject.
  • May reach those who would not otherwise engage.
  • Prevents single opinions from appearing representative.
  • Can be simultaneously delivered in multiple languages.
Long term
  • Higher-quality, better-aligned solutions.
  • Improved trust of outcomes.


  • Because the outcome of polling is generally not binding or influential, responses are less likely to be considered and be subject to a higher number of fake responses – especially when conducted online.
  • May not take into account or surface why participants are voting a particular way, only what their vote is.
  • Online polling through your own website is likely to be significantly biased. Using complementary applications such as google analytics may help to understand what this bias is likely to be, but they should not be considered statistically meaningful.



  • There are many polling tools out there. Work out the specifications of what you’re looking for first before being overwhelmed by the various opportunities.
  • If you need a more statistically relevant poll, contact a market research provider. They are usually able to add a poll into other surveys for a faction of the cost of a whole survey and will be able to help you match very specific target participant characteristics
  • Consider if you need to provide content in multiple languages.

During the process

  • If conducting online, monitor the response to your promotion of the poll and which promotion seems to be working the best. This may also tell you something about the type of bias that can be built into the results; for example, if most of the voters are coming to your tool from a left-wing news source.


  • You will probably not capture information on respondents. If appropriate you might publish an article on the results, or simple post the outcome in the same place as the poll once it is closed.