Retour
Ecommerce Market Research: A Practical Guide for Stores

Matthieu Saussaye

Ecommerce market research is the process of finding out what online shoppers want, how they find and compare products, and why they buy or leave, using a mix of market data, your own store analytics and direct feedback. Its advantage over offline retail research is that every visit leaves a trace; its trap is believing the trace explains itself.
What good ecommerce research covers:
Demand: is there search and marketplace demand for the product, and at what price points do competitors sell?
Acquisition: how do buyers actually discover you? Post-purchase attribution surveys fill the gaps that tracking loses.
Conversion: why do visitors leave product pages and checkouts? Analytics shows where; on-site surveys and interviews show why.
Retention: what turns a first order into a second, and what drives returns?
Positioning: what language do customers use about your products and competitors? Reviews and open-ended answers are a free copywriting brief.
Ecommerce research methods compared
Most stores already own the data for half of these. The other half costs little more than a well-placed question.
Method | Answers | When to use | Cost and effort | Blind spot |
|---|---|---|---|---|
Keyword and marketplace research | Is there demand, and what do people call it? | Before launching a product or category | Low: search tools, marketplace listings | Shows interest, not purchase intent or margin |
Competitor teardown | How do rivals price, ship, return and position? | Before launch and twice a year | Low: mystery shopping, test orders | Shows offers, not their conversion or profit |
Post-purchase survey | How did buyers find you, and why did they choose you? | Always on, on the order confirmation page | Low; high response because buyers are engaged | Only hears from people who bought |
Exit and abandonment surveys | Why did visitors not buy? | On product pages and checkout, triggered on exit intent | Low to medium; keep to one or two questions | Low response; skews to irritated visitors |
Review and support mining | What delights and annoys customers? | Monthly, across your store and competitors' reviews | Medium without coding tools | Extreme opinions are over-represented |
Panel survey of category shoppers | What does the wider market want, including non-customers? | Category entry, pricing, brand awareness | Medium: paid sample | Stated preferences can differ from behavior |
A/B test | Which version converts better? | After research gives you a hypothesis | Medium; needs traffic | Tells you what won, not why |
Start with the market: demand and competition
Ecommerce is still a minority of retail. The US Census Bureau estimated that e-commerce accounted for 17.1% of total US retail sales in the second quarter of 2026, seasonally adjusted (US Census Bureau). For many categories, the online opportunity depends on how much of the category has already moved online, so check category-level figures before assuming a market is digital.
Then test product demand:
Search demand. Compare monthly searches for the product, its synonyms and "best [product]" queries. A rising trend matters more than an absolute number.
Marketplace signals. Count listings, review volumes and price spread on the marketplaces your buyers use. Many reviews with low ratings is an opening; many reviews with high ratings is a fight.
Price ladder. Map competitor prices including shipping and return costs, because shoppers compare the total, not the sticker price.
Offer teardown. Place a test order with your top three competitors. Note delivery speed, packaging, post-purchase emails and how hard it is to return. This is the experience your customer is comparing you to.
Find out how buyers really discover you
Attribution tracking loses buyers who see you on one device and buy on another, hear about you from a friend, or decline cookies. A one-question post-purchase survey recovers a large part of that picture:
"How did you first hear about us?" (podcast, friend or family, Instagram, TikTok, search, marketplace, press, other with text box)
"What almost stopped you from ordering today?"
"What were you using or buying before you found us?"
"Which other stores did you consider?"
Keep the first question on the confirmation page and make the rest optional. Code the "other" answers each month: new channels often show up there first.
Research why visitors do not buy
Abandonment is the norm. Baymard Institute's average of 50 studies puts the documented online cart abandonment rate at 70.22% (Baymard Institute). In Baymard's own survey of US shoppers who had abandoned an order, setting aside those who were just browsing, the most cited reasons were extra costs being too high (40%), slow delivery (20%), not trusting the site with card details (19%) and being forced to create an account (18%).
Your store's reasons will differ, which is why you ask. Practical setups:
Exit-intent on checkout: "What stopped you from completing your order today?" with five options drawn from your own data and an open "something else".
Product page poll: "Is there anything you are missing on this page to make a decision?" Answers often name a missing size chart, video or delivery date.
Abandoned-cart email: include a single-click reason question alongside the reminder.
Short interviews with recent abandoners who opted in: ten conversations usually reveal objections no dropdown anticipated.
Mine reviews and support tickets
Reviews are unprompted feedback at scale, for you and for every competitor. Pull reviews for your top products and the three closest competitor products, then code them into themes: fit, quality, delivery, packaging, value, customer service. Count how often each theme appears in 1-2 star versus 4-5 star reviews. The gap tells you what to fix and what to advertise.
Support tickets and chat logs add the problems that never become reviews, such as tracking confusion or discount codes that fail. If your team cannot read thousands of entries, use a tool that codes open text into themes; see customer feedback surveys for how to close the loop.
Example ecommerce survey questions
Grouped by the decision they inform. Use one or two per touchpoint, never all at once.
Product and category research (panel or customers)
"When you last bought [category] online, what made you choose that product over others?"
"Which of these features matter most to you? Pick the most and least important." (MaxDiff)
"What price would you consider so low that you would doubt the quality?"
"Where do you usually start looking when you need [category]?"
Conversion and checkout
"What stopped you from completing your order today?"
"Was the delivery cost what you expected?" (yes / higher / lower)
"How easy was it to find the information you needed?" (1 to 5)
Post-delivery and retention
"Did the product match what you expected from the page?" (yes, better, worse, with a reason)
"How likely are you to order from us again in the next three months?"
"If you returned the item, what was the main reason?"
"How likely are you to recommend us to a friend? Why?" (see NPS done right)
Pricing, shipping and returns research
Online shoppers see the full price at checkout, so research the whole offer, not the product price alone. Three studies pay off quickly:
Free-shipping threshold test. Ask category shoppers what order value they expect for free delivery, then compare with your average order value. A threshold just above typical basket size can lift AOV; one far above it reads as a hidden fee.
Delivery promise research. Ask how many days buyers will wait for this category, and whether they would pay for faster delivery. Answers differ sharply between a gift bought last-minute and a restock purchase.
Return policy research. Ask whether the policy influenced the purchase and what would make them hesitate. Generous returns can raise conversion and return rates at the same time, so pair the survey with your return data before changing anything.
For price points themselves, Van Westendorp or Gabor-Granger questions with a panel of category shoppers give a range to test, and an A/B price test on live traffic confirms it.
Ecommerce KPIs research should explain
KPI | What it measures | Research that explains a change |
|---|---|---|
Conversion rate | Orders divided by sessions | Exit surveys, product page polls, usability tests |
Cart abandonment rate | Carts created but not purchased | Checkout surveys, abandonment interviews |
Average order value | Revenue per order | Bundle and price tests, MaxDiff |
Repeat purchase rate | Share of customers ordering again | Post-delivery surveys, cohort interviews |
Return rate | Share of items sent back | Return reason codes and follow-up questions |
Customer acquisition cost | Marketing spend per new customer | Post-purchase attribution survey |
Customer lifetime value | Gross margin per customer over time | Retention and satisfaction research |
Constraints and common mistakes
Consent and privacy. On-site surveys that tie answers to a customer record need a lawful basis under GDPR or state privacy laws. Say what you collect and why, and do not ask for more than you use.
Survey fatigue. Pop-ups on every page hurt conversion. Trigger surveys on specific events, cap frequency per visitor, and keep each one under 30 seconds.
Only hearing from buyers. Post-purchase surveys are easy and biased toward people you already convinced. Balance them with panel research among category shoppers who have never bought from you; our guide to finding survey respondents covers panel options.
Testing without a hypothesis. A/B tests on button colors burn traffic. Research first, then test the change research suggests.
Ignoring the open text. The "other" box and the "why" follow-up hold the insights; the closed options only hold what you already guessed.
Ask why, at the moment it matters
Ecommerce teams have more behavioral data than any other retailer and still struggle with the why: why this visitor left, why that customer returned the item.
SmartInterview runs short AI-moderated surveys by voice or text that follow up on vague answers ("too expensive" becomes "shipping was more than the item"), in the shopper's language, and codes every open answer into themes so you see the top reasons at a glance instead of reading thousands of comments.
Frequently Asked Questions
What is ecommerce market research?
Ecommerce market research is the study of online shoppers, competitors and demand to guide product, pricing, marketing and site decisions. It combines market data such as search and marketplace demand with your own analytics and direct feedback from surveys, reviews and interviews.
How do I do market research for an online store?
Check demand with keyword and marketplace research, tear down your top competitors with test orders, add a one-question post-purchase attribution survey, place a short exit survey on checkout, and mine reviews for recurring themes. Then turn the findings into A/B tests.
What is the average cart abandonment rate?
Baymard Institute's average across 50 studies puts the documented online cart abandonment rate at 70.22%. In Baymard's shopper survey, excluding people who were just browsing, extra costs being too high was the most common reason given, cited by 40%.
What questions should I ask in an ecommerce survey?
Ask how buyers first heard about you, what almost stopped them ordering, what stopped abandoners from completing checkout, whether the product matched the page, and how likely they are to order again. Always include an open follow-up asking why.
Is a post-purchase survey better than analytics for attribution?
It is a complement, not a replacement. Analytics captures tracked clicks precisely but misses word of mouth, podcasts, cross-device journeys and visitors who decline cookies. A post-purchase "how did you hear about us?" question fills those gaps, with some recall error.
How many responses do I need for an ecommerce survey?
For a directional read on top reasons, 100 to 200 answers per question is usually enough to rank the main themes. For comparing segments, such as new versus returning customers, aim for at least 100 per segment.


