Ecommerce Customer Interviews: Find Checkout Friction

A shopper can want the product, trust the brand, and still leave before paying. Ecommerce customer interviews help you find the moment when interest turns into hesitation, confusion, or extra work.

Analytics can show that mobile shoppers abandon checkout more often. They can’t always tell you whether delivery dates were unclear, a payment failed, or an account requirement felt unreasonable. A well-run interview connects the behavior to the reason behind it. Start with the shopper’s latest purchase experience, then trace every obstacle in the path.

Table of Contents

Key Takeaways

Ask about a real, recent shopping experience instead of asking what someone might do.
Focus on actions, decisions, workarounds, and moments of hesitation.
Separate purchase friction from general product dissatisfaction.
Rank findings by frequency, severity, and business impact.
Validate interview insights with analytics, session recordings, surveys, usability tests, or A/B tests before making broad claims.

How ecommerce customer interviews reveal purchase friction

Purchase friction is any obstacle that makes buying slower, harder, less certain, or less trustworthy. It can appear before the cart, inside checkout, or after an order requires account support.

A product page may hide important specifications below the fold. Search results may include irrelevant products. The cart may add shipping charges without enough explanation. Checkout may reject an address, require an account, or offer payment methods that don’t fit the customer’s situation.

For B2B stores, friction often sits outside the standard checkout flow. A buyer may need a quote, tax-exempt approval, a purchase order, invoice terms, or permission from another employee. If the site treats that buyer like a casual consumer, the purchase can stall even when the product meets every requirement.

Customer interviews reveal how shoppers experience these obstacles. A participant might say, “I couldn’t tell whether the item would arrive before our installation date.” That statement gives you more direction than a high exit rate alone. You can investigate delivery messaging, date calculations, inventory status, and the handoff between the product page and checkout.

Use the interview to uncover five types of friction:

Understanding friction occurs when shoppers can’t tell what a product includes, whether it fits their needs, or which option to choose.
Effort friction appears when customers repeat information, create an account, or complete unnecessary steps.
Trust friction involves unclear returns, uncertain delivery, unfamiliar payment requests, or missing company details.
Technical friction includes broken fields, payment errors, slow pages, and mobile interaction problems.
Organizational friction affects business buyers who need approvals, invoices, tax handling, or shared account access.

Baymard’s checkout flow audit guide covers recurring checkout issues found through large-scale usability research. Use that research to create a review list, but use interviews to learn which issues affect your customers and why.

The goal isn’t to collect complaints. It’s to understand the work a customer must do before an order can go through.

Recruit the right customers for useful interviews

A broad customer sample can hide the differences between shopping situations. A first-time mobile shopper may face different friction than a repeat B2B buyer placing a monthly order.

Begin with a clear research question. “Why are conversions low?” is too broad to guide recruitment. Better questions include:

Why do first-time visitors abandon after viewing shipping information?
What stops business customers from completing quote requests?
Why do repeat buyers contact support instead of using self-service account tools?
What causes shoppers to switch payment methods during checkout?

Recruit people who recently completed, abandoned, or delayed a purchase. Recent behavior produces better detail because the sequence is easier to remember. Ask about an order or shopping session within the past few weeks when possible.

Include participants who represent different purchase conditions:

First-time and returning customers
Mobile and desktop shoppers
Customers who bought, abandoned, or contacted support
Consumer and business accounts
Customers using different payment methods
Buyers with different delivery locations
People who ordered standard products and products requiring a quote

Avoid recruiting only loyal customers. They may have learned workarounds that hide problems for new shoppers. Also avoid treating customer service staff as substitutes for customers. Support conversations are useful research input, but staff members interpret events through their own role.

A short screening form can ask:

When did you last purchase from this store?
Did you complete the order, abandon it, or postpone it?
Which device did you use?
Did you need help before ordering?
Was this purchase for personal or business use?
Did anyone else need to approve or process the order?

Don’t ask leading questions such as, “Did the shipping fee seem too high?” That wording plants an explanation. Ask, “What did you think when you saw the shipping information?” Then follow the participant’s answer.

For many projects, 8 to 12 interviews can expose repeated patterns within a defined customer segment. That number isn’t a conversion estimate or a market percentage. It is a practical research sample for finding issues that deserve validation.

Plan ecommerce customer interviews around real behavior

A strong interview guide follows the customer’s timeline. It starts before the visit, moves through product discovery and purchase, and ends with what happened afterward.

Use open questions that invite a story. Avoid questions that ask participants to design your store for you. People can describe what they did with reasonable detail. Their predictions about future behavior are less dependable.

Questions about the shopping trigger

Start by learning what created the need:

“Tell me about the last time you needed a product like this.”
“What caused you to start looking on that day?”
“What did you already know about the product before you searched?”
“Which other stores or suppliers did you consider?”
“What did you need to decide before placing the order?”
“Was there a deadline, budget limit, or delivery requirement?”

These questions uncover context that a funnel report cannot provide. A customer may have visited through a product search but needed the item for a specific project. That project may make delivery timing more important than price.

Questions about discovery and product choice

Ask the participant to describe the path, not rate isolated pages:

“Where did you look first, and what did you do next?”
“What search terms did you use?”
“Which products did you compare?”
“How did you decide that one option was suitable?”
“What information did you look for but not find?”
“Did any product detail cause you to pause?”
“What made you remove an item from consideration?”
“Did you use filters, sorting, reviews, or comparison tools? Tell me how.”

Follow up on exact language. If someone says, “The options were confusing,” ask, “Which options were confusing, and what were you trying to compare?” If a participant says, “I wasn’t sure it was compatible,” ask, “What would have helped you confirm compatibility?”

For a B2B catalog, probe product data and purchasing rules:

“How did you confirm that this item matched your company’s requirements?”
“Did you need a specification sheet, compatibility chart, or formal quote?”
“Who else needed to review the product?”
“What information did you copy into your internal purchasing process?”

Questions about cart and checkout

The most useful checkout questions focus on actions and decisions:

“Walk me through what happened after you added the product to your cart.”
“What did you expect to see on the cart page?”
“When did you first see the delivery cost and timing?”
“Did anything in the cart change your decision?”
“Which checkout step took the most effort?”
“Were any fields unclear or difficult to complete?”
“Did you try a payment method that didn’t work?”
“What did you do when the error appeared?”
“Did you consider leaving the site? What led you to stay or leave?”
“What information would you have wanted before submitting the order?”

For an abandoned purchase, don’t ask, “Why didn’t you buy?” Start with the sequence:

“Please tell me everything you remember doing after you reached the checkout page.”

Then ask, “What happened next?” until the participant reaches the stopping point. This approach often reveals details such as checking a competitor’s delivery promise, waiting for a manager’s approval, or abandoning after a card verification failure.

Questions about post-purchase support

Purchase friction can appear after payment:

“Did you need to contact support before or after ordering?”
“What were you trying to find in your account?”
“How did you check the order status?”
“Did the invoice contain what your business needed?”
“Have you needed to reorder this product? Tell me what you did.”
“What part of the order process would you avoid next time?”

A customer who calls support for a tax-exemption question may complete the purchase. That still indicates friction, because the site shifted work to the customer and the support team.

Conduct the session without leading the customer

A 30 to 45-minute session gives most participants enough time to reconstruct a recent purchase. Send a clear consent request, explain how recordings will be used, and remove personal payment information from the session.

The interviewer should speak less than the participant. Let silence work. People often add the most useful detail after a short pause.

Use neutral prompts:

“What were you expecting there?”
“What did that label mean to you?”
“What did you do next?”
“How did you decide that?”
“What made you look for help?”
“Can you show me where that happened?”
“What would you have done if that option wasn’t available?”

Avoid defending the site. A response such as “That information is actually on the page” ends the discovery. Ask instead, “Where would you expect to find it?” The issue may involve placement, wording, contrast, or the order in which information appears.

Don’t turn the session into a usability test unless that is the research method. Interviews examine remembered behavior and context. A usability session observes someone completing a task in the moment. Ecommerce usability testing can follow interviews when you need to watch errors, hesitation, and task completion directly.

During each session, capture four things:

The customer’s goal
The action they took
The obstacle they encountered
The workaround or consequence

Write down the participant’s exact words when possible. “Shipping was expensive” differs from “I couldn’t tell whether the shipping charge included Saturday delivery.” The second statement points to a clearer design question.

Turn interview notes into ranked friction findings

Unstructured notes become difficult to use when every observation receives equal weight. Use a simple finding format:

Customer context -> observed behavior -> friction point -> consequence -> evidence -> possible response

For example:

A repeat B2B buyer needed a purchase order. They searched the account area, found no purchase-order option, and emailed sales. The order took an extra day and required manual processing. Three of five business participants described the same path.

Keep the observation separate from the interpretation. “The participant opened the shipping accordion twice” is an observation. “Shipping information is too hard to find” is an interpretation that needs support.

A practical prioritization model uses frequency, severity, and business impact. Score each from 1 to 3:

Factor	1	2	3
Frequency	One participant	Several participants	Common within the segment
Severity	Minor delay	Confusion or workaround	Abandonment, failed order, or lost trust
Business impact	Little measurable effect	Support work or slower purchase	Lost revenue, margin, or account risk

Multiply the three scores for a simple priority score. A problem mentioned by two people may outrank a common minor complaint if it causes payment failure or purchase abandonment.

Use the score as a discussion aid, not as scientific measurement. Interview samples show patterns within the people you spoke with. They don’t prove that a certain percentage of all customers experience the same issue.

A finding log might include these columns:

Finding	Segment	Frequency	Severity	Business impact	Evidence to check	Owner
Delivery date appears late	Mobile first-time buyers	2/6	3	Abandonment risk	Checkout exits by device	Product
PO instructions are unclear	B2B accounts	4/5	3	Manual sales work	Support tags and order path	Ecommerce
Size guide is hard to compare	Apparel shoppers	3/8	2	Returns and hesitation	Size-guide clicks, returns	UX

Group similar statements before counting them. “I couldn’t find delivery timing” and “The arrival date was hidden” may describe one issue. Keep distinct problems separate when they require different fixes.

Patterns become easier to discuss when you connect each finding to a page, event, customer segment, and business outcome. An ecommerce conversion checklist can help teams compare interview findings with checkout starts, payment failures, internal searches, and support contacts.

Validate qualitative findings before changing the store

Interviews tell you what customers experienced and how they explain it. They don’t measure the size of the problem across your entire customer base.

Pair each high-priority finding with a validation method. If interviewees describe delivery uncertainty, compare checkout exits before and after the delivery message appears. Segment the result by device, country, product type, and customer status.

Session recordings can show whether more shoppers pause at the same field. On-site surveys can test whether a larger customer group reports the same concern. A usability test can reveal whether people complete the revised flow without help.

Analytics questions should be concrete:

Do shoppers who view shipping information convert at a different rate?
Where do mobile users exit checkout?
How often do payment attempts fail by method?
Do B2B buyers start quotes but fail to submit them?
How many customers contact support before placing an order?
Does account creation correlate with checkout abandonment?

When the issue is clear but the best solution is uncertain, run an A/B test. Test one meaningful change, such as showing an estimated delivery date beside the add-to-cart button. Avoid testing several unrelated redesigns at once, because the result won’t show which change affected behavior.

A survey can help measure prevalence after interviews identify the language and response choices. For example, a survey might ask recent abandoners to select every reason that affected their decision, then offer an open text field.

Use the impact of UX design on conversion rates as a broader reference for connecting usability issues with conversion metrics. Still, don’t claim that interviews caused a conversion change. Compare a defined baseline with a defined test period and account for traffic and product changes.

A useful research conclusion sounds like this: “Five of eight interviewed B2B buyers couldn’t find purchase-order instructions. Support data shows related contacts increased during the same period. We should test clearer account and checkout messaging.” That is actionable without pretending to be a population estimate.

Convert friction findings into focused improvements

A research finding needs a product decision. Assign one owner, one proposed change, and one measure for each priority issue.

Suppose interviews show that shoppers don’t trust the delivery promise. A focused response could place the estimated arrival date near the product price, repeat it in the cart, and explain how the date changes by location. The measure might include checkout completion, delivery-information exits, and support contacts about arrival dates.

For a B2B account, the response may need a different flow. Add visible links to request a quote, show invoice terms, explain tax-exemption steps, and let approved users submit purchase orders. Track quote submission, approval time, order completion, and manual support contacts.

Use the customer’s language in the interface when it is clear and accurate. If participants ask, “Will this arrive by Friday?” a delivery section titled “Estimated arrival” may work better than an internal label such as “Fulfillment SLA.”

Prioritize changes that remove work at the point where the customer needs the answer. A long help article won’t fix a missing delivery date on the product page. A general checkout redesign won’t fix a rejected address if the error message gives no recovery path.

After release, review the same evidence that supported the finding. Did payment failures fall? Did customers stop calling about invoices? Did mobile shoppers reach the confirmation page more often? Keep the original interview notes beside the result so the team can compare behavior, explanation, and outcome.

Repeat interviews when the customer segment, product mix, or buying process changes. A checkout that works for consumer orders may still frustrate procurement teams. The right question depends on the decision the customer is trying to make.

A reusable ecommerce customer interview template

Use this compact guide for a 35-minute session:

Opening, 3 minutes

“Thanks for speaking with me. We’re studying how people shop and place orders. There are no right answers, and we’re interested in what you actually did.”

Recent purchase context, 7 minutes

“Tell me about your most recent purchase from this store.”
“What created the need?”
“What did you compare before deciding?”
“Who else was involved?”

Purchase path, 15 minutes

“Where did you begin?”
“What did you do after that?”
“What information did you look for?”
“When did you feel confident enough to continue?”
“Did anything slow you down or make you reconsider?”
“What did you do when you couldn’t find an answer?”

Checkout and account process, 7 minutes

“Which checkout step required the most effort?”
“Did the payment, delivery, tax, or account process cause a problem?”
“How did you recover?”
“What would have made the order easier to complete?”

Closing, 3 minutes

“What is the first thing you’d change about that experience?”
“Is there anything you expected to happen that didn’t?”
“May we contact you if we need to clarify part of your story?”

Review notes soon after each session. Record the exact path, the customer’s words, and the evidence you need to check. Then rank findings by frequency, severity, and business impact before choosing a design response.

Conclusion

A customer interview becomes useful when it stays close to real behavior. Ask about the last purchase, follow the customer’s sequence, and record the point where effort, doubt, or delay entered the process.

Ecommerce customer interviews won’t tell you how every shopper behaves. They will show you which questions your store fails to answer and which workarounds customers use to keep moving. Combine those insights with behavioral data and controlled tests, then fix the friction that blocks the highest-value purchase decisions.

Spread the love