4 Types of Surveys To Use in Conversion Rate Optimization
- A low conversion rate shows where visitors leave, not why. A survey asks the visitors themselves, so you test fixes built on their reasons instead of a guess.
- Each survey has three settings: a trigger (exit intent or page load), a segment (new visitors, a traffic source, all visitors) and the pages it runs on.
- The cart page is where an exit survey pays off most, because the visitor has already chosen products and something specific stopped the order.
- Ask open-ended questions first to discover reasons, then turn the most common answers into a single-choice question you can count and compare.
- The NPS survey on the thank you page measures the buying experience; product and delivery need a second NPS survey sent after the order arrives.
Conversion rate optimization involves using multiple tactics that could help an online business grow. One of them is the combination of surveys with A/B testing. Surveys tell you why visitors behave the way they do; A/B tests tell you whether the change you make because of those answers actually works.
For example, a low conversion rate on the product page may point to design elements that don't appeal to shoppers. But a guess won't solve the problem. You need to know exactly why people are leaving the page instead of adding items to their cart, and you can know for sure if you ask. In this article you'll learn four surveys to use for your conversion rate optimization goals: when to run each one, where it appears, what to ask, and what to do with the answers.
Why do surveys belong in conversion rate optimization?
Every A/B test starts with a hypothesis. When that hypothesis comes from an internal opinion, most of the testing budget goes into ideas that customers never cared about. When it comes from what visitors say, the test targets a problem that real people reported.
Surveys and tests work as a loop:
- Analytics shows the page or step where visitors leave.
- A survey on that page asks the visitors why.
- A hypothesis turns the most common answer into a change you can make.
- An A/B test checks whether that change improves the conversion rate.
The four surveys below each cover a different point in that loop, from the first visit to the completed order. For general CRO practices that pair surveys with testing, see this roundup of CRO best practices.
The 4 types of surveys at a glance
| Survey | Trigger | Segment | Pages | What it tells you |
|---|---|---|---|---|
| 1. Visitor satisfaction | Exit intent | New visitors | All pages except cart and thank you page | Whether visitors found what they came for, and what is missing |
| 2. Purchase intention | Page load | Each important traffic source | Landing pages, except paid search campaign pages | Why each traffic source arrives and what it expects |
| 3. Conversion barriers | Exit intent | Each important traffic source | Cart page | Why visitors abandon the order |
| 4. Net Promoter Score | Exit intent | All visitors, or new vs returning | Thank you page | How likely new customers are to recommend you |
The question formats matter as much as the placement. For choosing between open answers, single choice and rating scales, see how to choose a survey scale.
1. Visitor satisfaction survey: did visitors find what they needed?
Trigger: exit intent. Segment: new visitors. Pages: all pages, except the cart page and the thank you page.
This survey does three jobs at once: it measures a simple satisfaction index, it discovers visitor needs, and it captures leads. New visitors are the right segment because they do not know your site yet, so they are the ones most likely to miss a product that exists or to leave because it does not.
Example questions
- Have you found what you were looking for? Yes / No
- If yes: "Thank you!"
- If no: "What were you searching for?" (open answer)
- Then: "If you want to receive personalized offers by email, let us know." (incentive to subscribe to the newsletter)
What to do with the answers
- Track the share of "yes" answers over time. It is your visitor satisfaction index. A drop after a site change is a signal to look closer.
- Group the "what were you searching for" answers. Products you sell but visitors could not find point to navigation, search or category page problems. Products you do not sell point to range gaps.
- Follow up with the leads. A visitor who told you what they wanted and left an email is the easiest personalized offer you will ever send.
2. Purchase intention survey: why did each visitor come?
Trigger: page load. Segment: each important traffic source. Pages: landing pages, except the ones for paid search (Google Ads) campaigns.
Paid search landing pages are left out because visitors there already expressed intent through the keyword they searched, and you pay for every click, so a survey that interrupts them on load is a cost you do not need.
Example questions
- "Welcome to our site! Can you help us by telling us what your intention is?"
- Single-choice options such as: buy something today / compare products or prices / look for inspiration / check an order or get support.
What to do with the answers
- Compare intent by traffic source. If most visitors from a channel are browsing, a hard "buy now" landing page is the wrong page for that traffic.
- Match the page to the intent. Comparison-minded visitors need specifications, reviews and prices side by side; ready-to-buy visitors need a short path to the cart.
- Feed segments. Visitors who say they are comparing are good candidates for a personalized follow-up message or offer on their next page.
3. Conversion barriers survey: why didn't they finish the order?
Trigger: exit intent. Segment: each important traffic source. Pages: cart page.
The cart page is where this survey pays off most. The visitor has already chosen products and shown clear purchase intent, so something specific stopped the order, and that reason is usually something you can fix.
Example question
"Why have you chosen not to finish the order?"
Open answer or single choice?
- Open answer gives you subjective responses and insights, including reasons you would not have listed yourself.
- Single choice lets you see statistical results and compare how different barriers affect customer behavior.
Use the single-choice survey after an open-question survey, never instead of it. The open answers tell you which options to offer; otherwise you only count the reasons you already assumed.
What to do with the answers
- Rank the barriers by how often they appear, separately for each traffic source.
- Turn the top barrier into a hypothesis. If visitors say shipping costs appear too late, test showing them earlier; if they say they are unsure about returns, test a returns message next to the order button.
- Test one barrier at a time, so you know which change moved the result.
Ask why visitors abandon the cart, then test the fix on the same page and the same segment.
See surveys in Omniconvert Explore →4. Net Promoter Score survey: will new customers recommend you?
Trigger: exit intent. Segment: all visitors, or new and returning visitors separately. Pages: thank you page.
Example questions
- "How likely are you to recommend our website to your friends and colleagues?" (0 to 10)
- Follow-up: "What is the main reason for your score?" (open answer)
The Net Promoter Score matters for growth because word of mouth is one of the strongest ways to become visible online, and NPS measures how willing your customers are to provide it. For follow-up wording by respondent group, see these NPS question examples.
What to do with the answers
- Read the detractors' reasons first. Complaints about the checkout, payment or order confirmation are problems you can test on-site.
- Compare new vs returning customers. A lower score for returning customers means something after the first order is not working.
- Add a post-delivery NPS survey by email. The thank you page score reflects the buying experience only. The product and the delivery can be judged only after the order arrives, which is what an NPS survey template sent after delivery measures.
If you want to measure satisfaction with a specific interaction rather than likelihood to recommend, use CSAT instead. More survey formats for customers are on the customer satisfaction surveys page.
How do you turn survey answers into A/B tests?
-
Pick the page with the biggest drop-offUse analytics to choose where to ask. For most stores the cart page and the product pages come first.
-
Run the matching survey with an open questionChoose the trigger, segment and pages from the table above. Keep it to one or two questions, and use frequency capping so the same visitor is not asked repeatedly.
-
Group the answers into themesRead the responses and tag each one. Stop when new answers only repeat themes you already have.
-
Measure the themes with a single-choice surveyOffer the top themes as options, plus "other". The counts show which reason affects the most visitors.
-
Write a hypothesisState the change, the audience and the expected result: "Because cart visitors say X, showing Y will increase completed orders." For structure, see how to build an A/B testing plan.
-
A/B test the change on the same segmentRun the test until it reaches statistical significance, then implement the winner and run the survey again to confirm the barrier is gone.
To use this combination of surveys and A/B testing you would normally need two different products, one for testing and another for surveys. Omniconvert Explore is a complete conversion rate optimization tool that does both. Surveys in Explore can trigger on page load, on scroll, on click or on exit, target specific segments, run as a pop-up or widget, and include lead capture forms. Explore then runs the A/B test, personalization or overlay built on what visitors told you.
Frequently Asked Questions
Four on-site surveys cover most CRO needs. A visitor satisfaction survey on exit tells you whether people found what they came for. A purchase intention survey on landing pages tells you why each traffic source arrives. A conversion barriers survey on the cart page tells you why people abandon the order. An NPS survey on the thank you page tells you how likely new customers are to recommend you.
Analytics shows where visitors drop off, but not why. A survey asks the visitors themselves, so the reason comes from them rather than from a guess. The answers become hypotheses, and an A/B test then proves whether the change built on that hypothesis actually lifts conversions. Surveys without tests produce opinions; tests without surveys test guesses.
Show an exit-intent survey when the visitor's behavior signals they are about to leave the page, such as moving the cursor toward the browser bar or tab. It is the last moment you can still ask them something. Use it on pages where leaving is a loss, such as product and category pages or the cart, and keep it to one or two questions.
Start with an open-ended question. It gives subjective answers in the customer's own words and shows you reasons you had not thought of. Once you have enough of those answers, group them into the most common reasons and run a single-choice version of the same question. The single-choice survey gives you counts you can compare, so you can see which barrier affects the most visitors.
On-site, place the NPS survey on the thank you page, where every respondent has just completed a purchase. That score reflects the buying experience. To measure the product and the delivery as well, send a second NPS survey by email after the order arrives, because the customer cannot judge those at checkout.
Keep an on-site survey to one or two questions. The visitor is in the middle of a task and did not come to your site to give feedback, so every extra question costs answers. Use a follow-up question only when the first answer needs a reason, for example asking what the visitor was looking for after they answer no.
There is no fixed number. Keep collecting open answers until new responses stop producing new reasons and the same themes repeat. For a single-choice survey you compare between options, so collect enough answers that the ranking of the top reasons stays stable from week to week. Surveys generate hypotheses; the A/B test is what needs statistical significance.
Yes. Omniconvert Explore includes on-site surveys alongside A/B testing, personalization, overlays and segmentation. Surveys can trigger on page load, on scroll, on click or on exit, target specific segments, and use frequency capping, so you can ask a question and then test the fix on the same audience without a second product.
If you run only one of these surveys, run the conversion barriers survey on the cart page. Start with the open question, read every answer for two weeks, and group what people tell you. Then switch to a single-choice version to find out which barrier is the biggest. Pick that barrier, write a hypothesis, and A/B test the fix. Add the satisfaction, purchase intention and NPS surveys once the first loop from question to test is working.
Ask your visitors, then test the answer
Omniconvert Explore runs on-site surveys and A/B tests in one platform. Trigger a survey on load, scroll, click or exit, target the segment you care about, and test the fix on the same audience.