Collect survey responses and compare votes

Imagine a bakery tasting three new pastries before choosing one for the counter. “I liked the lemon one” is pleasant feedback, but it does not tell the baker whether the sweetness needs changing or whether people would actually order it. A favorite vote and a few consistent ratings answer different parts of that decision.

A survey app fits when you want structured responses and a way to compare them without transcribing a stack of notes. Decide what you will do with the results first. Each question should help you make that decision, and people should be able to skip items they have not tried.

The short version

Choose the decision your survey should help you make. Ask for a ballot with defined scales and a results view with response counts. Submit a sample answer and compare the charts with your choices. Check skipped-answer handling and record permissions before sharing your survey.

Write the decision into the brief

For a tasting, keep ratings separate from the final favorite. That lets you see a pastry that scores well across several qualities even when another wins the single vote.

Build a tasting survey for an illustrative bakery testing three pastries:
lemon bun, chocolate twist, and almond tart.
Let tasters mark which items they tried. For each tried item, ask for
flavor and texture ratings from 1 to 5, a sweetness choice,
and whether they would order it.
Ask for one overall favorite and an optional short comment.
Add a results view with response counts, average ratings,
and favorite-vote totals. Exclude untasted items from rating averages.
Use clearly labeled sample responses for the draft.

Replace the pastries with your actual choices and define the scale before asking anyone to respond. For app creation, publication, and permissions for answers and results, follow Build a live app on your canvas.

Try the tasting example

The bakery example separates a tasting ballot from a results view. It illustrates the workflow rather than promising that the request above will reproduce the same survey or charts.

Try these two actions:

1. In Tasting Ballot, mark the pastries you tasted, enter ratings, choose a favorite, and submit a sample response. Leave an untasted item excluded.
2. Open Results & Charts and inspect Flavor & Crumb Scores and Tasters’ #1 Favorite Vote. Compare the rating summary with the favorite tally.

The preview uses disposable sample data. Your changes reset on reload, and its results represent example responses rather than real customer research. You can inspect both sides here; access to records in your real app comes from its collection permissions.

Improve how you interpret the results

Try one revision that makes the numbers easier to read:

Show the number of tasters beside each pastry's average rating.
Explain that skipped pastries are excluded from that average.
Keep existing responses and show “No ratings yet” instead of zero
when an item has not been rated.

Before using your own survey, submit a response with a skipped item and confirm the denominator in the results. Decide whether people can answer more than once and test the rule you request; the sample preview is not proof of voter identity or a one-person-one-vote policy. Choose who should see individual answers and who should see summaries. A hidden name field alone does not establish anonymity: check the stored fields and collection access before describing the survey that way.