They added it to the cart. Why did they leave?

Imagine a buyer choosing a ₹700 lunch bag during her tea break. She picks a colour, adds it to the cart and reaches checkout. Then a delivery fee appears. She pauses. The phone rings. The bag stays in the cart. This is an imagined scene, but the question it raises is useful: what did your page leave until too late?

To reduce cart abandonment, show costs early, make delivery and returns clear, and test checkout on a real phone. Find the step where people leave before choosing a fix. Some visitors are only comparing options, so an unfinished cart is not always a broken sale.

Find the step that needs attention

Keep cart and checkout counts separate

Cart abandonment means someone added an item but did not finish an order. Checkout abandonment starts later, after they began checkout. Mixing the two can send you to the wrong part of the shop. Write down the counts for each step over the same period, using the same analytics tool.

Watch one complete journey

Ask someone who has not built the store to buy a product in test mode. Say only what they need to buy. Watch where they stop, go back or ask a question. Do not guide their hand. A shop that works only when its owner explains it has more work to do.

Remove the surprise before the pay button

Baymard's abandonment research identifies extra costs, slow delivery, account creation and checkout problems among reasons people leave. Its survey describes US shoppers; it is a useful set of questions to investigate, not an Indian-store benchmark.

Give a useful cost estimate

If delivery depends on the pincode, let the buyer check it before payment. When the final charge needs an address, explain that clearly. Do not write “free delivery” on a banner and quietly add exceptions at the last step. Use the shipping-cost guide to check the offer first.

Put answers beside the decision

Show the arrival estimate and a short returns summary close to the purchase controls. Link to the full terms. “Dispatch in two working days” and “arrives in two working days” mean different things. The wording should match what your packing team and courier can actually do.

Turn a vague complaint into a specific fix

This table is a troubleshooting aid, not a claim that one behaviour has one cause. For example, someone may return to the cart simply to add a second item. Check the recording or ask the tester before changing the page.

Test the fix without changing everything

Write a small test note

Use this format: “We think buyers miss the delivery charge. We will show the pincode estimate in the cart. We will compare checkout completion and support questions.” Save the date and a screenshot. That makes the change easier to judge and undo.

Keep the comparison fair

Do not change price, ads and checkout together if you want to learn which one helped. Note sales, holidays and traffic changes. Ten visits cannot settle a close comparison. Low traffic is a reason to combine observation with the numbers, not a reason to claim a winning percentage.

Baymard's checkout usability guide offers more detailed patterns to inspect. Start with a visible problem on your own store instead of copying every suggestion at once.

Finish one audit before your next campaign

• Open your highest-visited product on a phone and choose a real variant.
• Check the total, delivery estimate and returns summary before payment.
• Try a bad address and a failed payment in a supported test environment.
• Confirm that the buyer sees one clear order result and knows where to get help.
• Record the three most serious problems and fix the first one.

The mobile checkout checklist covers the small-screen part in more detail. For a worksheet you can keep beside the test, use the buyer-confidence kit. Neither exercise needs a bigger ad budget.