Single-Select vs. Multi-Select Shopify Survey Questions

Use single-select when you need one main answer. Use multi-select when several answers can apply to the same order. The choice changes both what you ask customers to do and what the resulting counts mean.

For example, “What was the main reason you chose us?” asks for a priority. “Which reasons influenced your purchase? Select all that apply” asks for a set of influences. Neither wording can answer the other question on its own.

Choose the format from the decision

Before adding answer choices, write down what you want to decide after reading the results. This keeps the question focused and makes the reporting method easier to choose.

For discovery research, the “How did you hear about us?” guide explains wording and follow-ups. For other goals, use the post-purchase question examples.

Keep choices specific enough to distinguish. “Price” and “Good value” may mean different things, but a buyer could reasonably struggle to choose between them. If that distinction will not change your next action, simplify it. Offer a truthful way to answer when the listed choices do not fit; use a separate text question when you need an explanation.

Multiple selections do not reveal priority

If someone selects Design, Price and Availability, you know those three were selected. You do not know whether Design mattered most, whether each mattered equally, or whether any one of them caused the purchase. The order of labels in a saved answer is not a ranking.

If priority matters, ask a separate single-select question such as “Which of these mattered most in your decision?” Keep that extra question only if you will use it, and configure its choices explicitly.

Question format can also affect what people report. Pew Research Center's experiment found differences between a select-all list and asking a separate yes/no question about each item. That study concerned reported life experiences, not Shopify checkout performance. It is a reason to be careful about comparing formats, not a merchant response-rate benchmark or proof that single-select is always better.

Count answering orders and selections separately

This is an invented example, not OrderSurvey merchant data. Suppose five distinct orders answer one multi-select question: “Which reasons influenced your purchase?” Known test orders and duplicate records have already been removed.

There are five answering orders and eight selections. Each option gets two useful calculations:

For Design, dividing three by five answering orders gives 60%; dividing three by eight selections gives 37.5%. “Design was selected by 60% of answering orders” and “Design accounted for 37.5% of selections” are both correct here. They answer different questions. Label the denominator every time you show the percentage.

The 160% total is expected because several orders contribute to more than one option. SurveyMonkey's checkbox-question documentation explains the same counting principle. It is not a reason to remove valid selections until the percentages total 100%.

For a merchant report, a bar chart labeled “Share of answering orders; multiple selections allowed” is easier to interpret than a pie chart implying mutually exclusive groups. Show counts alongside percentages, especially for small batches of answers.

Define the group before calculating percentages

Keep the same question, wording, answer choices, audience and date window together. Count an order at most once per option for that question after reviewing known tests and duplicate records. Five answering orders are not necessarily five unique customers: a customer can place more than one order.

Use the number of orders that answered this specific question as the denominator for the worked calculation. A conditional follow-up may have fewer answers than the opening question. A skipped question is not evidence that none of its options applied.

Also distinguish combinations from individual options. In the example, “Design; Price” is one combination, selected by one order. Design itself appears in three orders. A chart of complete answer strings will not produce the option-level counts above unless the selections are separated first.

Do not treat these percentages as shares of all store visitors, revenue or incremental sales. Voluntary post-purchase answers describe participating buyers. The channel-report guide covers the additional work and limits involved in joining survey answers to order values.

Review multi-select answers in OrderSurvey

OrderSurvey supports both single-select and multi-select questions. Preview the question on a phone and verify that the instruction matches what customers can select. Use the setup walkthrough to add and verify the Thank you and Order status blocks.

Analytics and Responses are available on Free. The current dashboard groups primary answer text, and the Responses distribution groups saved answer text. A multi-select answer can therefore appear as a combination. These views do not automatically produce the separate-option percentage table shown in this guide.

For a small batch, read the individual answers and tally each option once per answering order. For repeatable spreadsheet analysis, Unlimited provides CSV export, with a row for each question/answer pair. Multiple selections appear together in the Answer field; the export does not create a separate column for each selected option. Total CSV rows are not a count of completed surveys.

Import the file with a CSV-aware spreadsheet, then isolate the question and review its answer labels before separating choices. The displayed multi-select answer joins choices with commas. If a choice itself contains a comma, splitting on every comma can misclassify it. Check those combinations manually rather than assuming every comma separates options. Keep a copy of the original file and document any recoding.

If you change a question from one answer to multiple answers, record a new version and compare the periods separately. Use the survey change log to preserve wording, format and dates.

Start with one useful question

Choose single-select for a main reason or first source. Choose multi-select for several relevant influences, then decide whether your report needs answering-order shares, selection shares or both.

Free includes unlimited surveys, in-app analytics and 100 submitted survey responses across the store per rolling 30 days. A submitted survey containing several selections still uses one response from that allowance. Unlimited is $24 USD/month for unlimited responses, CSV export, configured low-score NPS webhook alerts and priority support. See current pricing.

Install OrderSurvey, or book setup help with Dennis to check your question and both page placements with your permission.

Frequently asked questions