Research methods
Ecommerce market research: decide what your store should sell next
Use store searches, orders, returns, Google Trends and competing offers to decide whether to add a product range—and which questions need interviews.
Ecommerce market research should help you decide what to stock, whom to sell it to and why someone would choose your store. Start with what shoppers already search for, buy and return. Compare that evidence with search interest and competing offers, then interview people about the questions the records cannot answer.
For an existing store, a useful research question is specific: “Should we add this product range?” A report on the size of online retail will not settle that decision.
Start with one buying decision
Suppose an online luggage store is considering adding replacement suitcase wheels. This is a hypothetical example. Searches and support questions might suggest an opportunity, but the owner needs to know which suitcase models people are repairing, whether compatible parts exist and whether selling them would leave enough money after support and returns.
Write down the decision before opening a research tool: “Choose between stocking a limited range of replacement wheels, improving repair information or leaving this category alone.” Keeping all three options open makes the research more useful.
Read your store's evidence first
Review orders, returns, customer questions and site searches for the same period. For the wheel idea, separate requests for a replacement part from complaints about a recently purchased suitcase. One suggests a possible new sale; the other may require fixing the original product.
Shopify's behavior reports distinguish searches that returned no results from searches that returned results but received no clicks. The difference changes your next step:
- No results: check whether the product is missing or whether shoppers use a name your search does not recognize.
- Results but no clicks: inspect what appeared. The results may be irrelevant, or shoppers may need clearer pictures or compatibility information.
- Purchases followed by returns: read the reasons before treating sales as evidence of a successful product.
Shopify says its search reports can lag by up to 72 hours. Leave that gap at the end of a comparison instead of treating the newest incomplete days as a decline.
For another platform, look for equivalent reports. If you have no store history, begin with public sources below and keep the missing purchase evidence explicit.
Choose outside tools by the question they answer
These sources cover different parts of the decision:
| Source | What to look for | What it leaves unanswered |
|---|---|---|
| Google Trends | Recurring seasonal interest and differences between product terms in the same market and period | How many people will buy from your store |
| Google Keyword Planner, if you already have access | Search estimates for specific product and problem phrases | Whether searchers found a repair guide, bought a part or gave up |
| Competing stores' product and policy pages | Compatibility, delivered price, delivery promise and return conditions | Actual sales, margins and customer satisfaction |
| Your orders, returns and support records | What your customers bought and where the purchase went wrong | Demand among people who never reached your store |
Google explains that Trends uses sampled searches normalized to the chosen time and geography on a 0–100 scale. A peak is relative search interest, not a sales estimate. Compare terms together under consistent settings; an isolated spike deserves investigation before an inventory order.
Keyword Planner's definitions also matter: average monthly searches include close variants, and “competition” measures advertisers. A low label does not establish easy organic rankings or few stores selling the product.
For the wheel example, compare “replacement suitcase wheels” with repair-oriented phrases. Then inspect a few competing listings for the same suitcase type. A cheap wheel that does not fit is not an equivalent offer. Record the date and delivery destination beside each comparison.
Use interviews to explain the gaps
Now the owner has questions worth asking. Talk to people who recently repaired a suitcase, replaced one instead or abandoned a repair. Ask them to walk through the last attempt: how they identified the model, where they looked, what they compared and what they eventually did. Consumer research interviews covers recruitment and question wording.
An experienced repair-shop owner or ecommerce operations manager can answer a different question: what makes replacement parts difficult to sell and support? Ask about compatibility checks, information customers need before ordering and common return causes. Request general experience they can share, rather than another company's private sales records.
Instant Expert can help you find professionals for that operational perspective and invite them to a paid call. Start with retail operations professionals, then check each person's experience against your question. Those conversations complement the accounts of actual shoppers.
Turn the findings into a limited test
For this example, the research could lead to a small range with explicit compatibility guidance, better repair instructions for existing customers, or a decision to stop. Choose the action that addresses the evidence you found.
Before testing a range, set the maximum stock commitment and the return, support and cost conditions that would make you stop. Record purchases alongside refunds and support effort; a popular item can still be difficult to sell profitably.
Your next step is to choose one proposed range, review its closest existing evidence and write the unresolved question. That question determines whether you need another report, a shopper conversation or someone who has operated the category.