Understand visitors
Visitors leave without buying because the offer does not fit their needs, important information is missing, purchasing feels difficult, or they are not ready to decide. Understanding which explanation applies requires more than counting exits: you need evidence about what visitors were trying to do.
Imagine three people leaving the same sofa page.
One needs a narrower model. Another cannot find delivery information. The third is collecting ideas for an apartment they will move into next year.
They produce the same immediate outcome—no purchase—but need three different responses.
An interested visitor can like your business and still find the current offer unsuitable.
Examples include:
These are reasons to investigate demand, not automatically reasons to change your offer.
Ask: “What would make this a better fit for you?”
Then look for repeated, specific needs. Ten requests for “different hours” are less actionable than six requests for Thursday appointments after 6 p.m.
Publishing information does not guarantee people will notice it.
Nielsen Norman Group’s research on online reading finds that people commonly scan pages rather than read every word. Its practical implication is straightforward: essential information needs clear headings and a location that matches the visitor’s task.
Source: Nielsen Norman Group, How People Read Online
If delivery areas appear only in a long FAQ, someone considering a product may miss them. If membership restrictions sit below several promotional sections, visitors may remain uncertain.
Before developing a new offer, ask whether you need to explain the existing one more clearly.
Useful question: “What information would help you decide?”
A visitor may want the offer but struggle to complete the action.
Check the experience yourself:
Use actual errors, support enquiries, and visitor feedback to identify problems. Do not assume that every abandoned journey reflects a price objection.
Useful question: “Was anything difficult while trying to book or buy?”
Some visitors are researching, comparing, or planning.
Trying to force an immediate purchase may be less useful than letting someone express a specific requirement and request an update.
For example:
“I’m looking for a beginner course starting in January.”
That is more useful than treating the person as an unidentified lost visitor. It still does not guarantee a booking.
Use a simple decision table:
| What visitors say | What to investigate |
|---|---|
| “Do you deliver here?” | Whether delivery coverage is clear. |
| “I need a smaller version.” | Demand for specific dimensions. |
| “I can’t complete payment.” | A potential technical problem. |
| “I need this later.” | Timing and a relevant future update. |
Change one meaningful thing, then measure what happens.
No. Someone may have found an address, checked opening hours, or gathered information for a later decision. Interpret exits against the purpose of the page.
No. People who respond are self-selected. Describe findings as patterns among respondents, and test them before making wider assumptions.
Start with one question connected to the page: “What would make this right for you?” is useful when you want to understand an unmet requirement.
Next: Explore 12 questions for visitors who don’t buy.
Try the IfOnly demo to see how a visitor can explain what is missing.
Research basis: Nielsen Norman Group’s findings on scanning and online reading.
Collect feedback
The best question for a visitor who has not purchased is short, neutral, and relevant to what they are considering. Ask about one thing at a time, then clarify only if the answer leaves an important detail unresolved.
Read article 4 min read
Act on demand
Turn customer requests into a useful test by identifying a shared need, creating a small offer that addresses it, inviting the relevant people, and measuring the outcome.
Read article 4 min readTry a short conversation from your visitor’s perspective.
Try the interactive demo