"Alfa-Market" case

UX research: multi-seller cart

A comparative test of two cart concepts — seller tabs vs. a single combined list. Full methodology, 9 tested hypotheses, key insights, and recommendations.

5participants

5in-depth tests of both concepts

9hypotheses tested

10problem areas found

Research goal

Determine which cart concept better helps users manage items and understand that orders from different sellers are checked out separately.

What we tested

Option A — tabs

Each seller's items live on their own tab. Seeing the whole cart means switching between stores.

Option B — single list

All items are visible on one screen, grouped by seller. Each block has its own checkout button.

Methodology

Method

In-depth interviews and moderated testing of an interactive prototype

Format

Online, one-on-one, with screen sharing

Sample

11 active marketplace users

Comparative test

5 participants tested both cart options

Rotation

Three started with tabs, two started with the single list

Completion date

June 17, 2026

Team

Designer — Maria Vorobyeva, product owner — Alexander Kimaev

Hypotheses and results

#	Hypothesis	Result	What we saw in testing
H1	Tabs make the cart's structure clearer and help users perceive each order separately	Rejected	Three participants missed the other tab or lost an item. Grouping by seller was hard to read.
H2	Tabs reduce visual load by showing one order at a time	Rejected	The hidden part of the cart felt like a drawback. Switching between tabs added extra load.
H3	Users notice other orders and understand each seller must be checked out separately	Partial	Separate checkout was understood mainly thanks to each seller's own button, not the tabs.
H4	Users correctly understand the total cost and what it's made up of	Confirmed	Per-seller totals were read correctly, but participants missed a total for the whole cart.
H5	Users notice and correctly interpret discounts and savings	Partial	After a follow-up question the savings were understood, but spontaneous attention to discounts was low.
H6	Basic item actions are easy to find	Confirmed	Favoriting, removing, and changing quantity were found on the first try. Selecting items for a partial order was missing.
H7	Users correctly relate price, discount, and actions to a specific order	Partial	The connection was clearer in the single list. In the tabbed version, hidden content and unclear grouping caused confusion.
H8	The single list lets users see the whole cart and all sellers at once	Confirmed	All sellers and items were visible without switching. No items were lost in this version.
H9	Visual separation of orders in the list is enough without tabs	Confirmed	Block borders and separate checkout buttons helped users perceive items as distinct orders.

Key insights

Tabs hide part of the cart.
Three participants concluded that a third item hadn't been added, because it was sitting on another seller's inactive tab. They went back to the product card and tried adding it again.Direction: drop the tabs; the cart counter should match what's actually visible.

The cart doubles as a wishlist.
Four comparative-test participants kept items in the cart "for later" and bought them in batches. When asked to check out a single item, they looked for checkboxes. The workaround via favorites and removal was frustrating.Direction: add item selection and a "Check out selected" action.

Split checkout goes against the familiar model.
Participants expected a "one cart — one order" scenario, even when items belonged to different sellers. A separate button in each block helped explain the new mechanic, but didn't remove the friction entirely.Direction: show the seller, order number, per-order total, and a dedicated CTA clearly and up front.

Seller isn't always read as the grouping principle.
Some participants explained the item split by category instead — e.g., "electronics" vs. "clothing." For most, seller wasn't the main mental anchor when picking a product.Direction: strengthen the visual boundaries of each block and the link between seller, items, total, and checkout button.

Final decision: use the single list as the base and carry over the tabbed version's main advantage — a summary block with the cart's total amount and total savings.

What already works

Showing the unit price when quantity changes helps prevent mistakes.
The order total recalculates immediately after a quantity change.
Favoriting, removing, and changing quantity feel familiar.
A separate checkout button per seller explains the split checkout.
A summary block with the total and total savings helps users assess the whole cart.

Recommendations

Must — before launch

Drop the tabs and show all orders as a single list grouped by seller.
Add item-selection checkboxes and a "Check out selected" action.

Nice to have — a noticeable improvement

Show an estimated delivery window or date on the product card and in the seller block.
Consider search and filtering inside the cart for users with many items.
On the checkout screen, show the pickup-point address as a list selection, revisit the "Recipient" field, and move the delivery date higher.
Show savings in rubles without an aggressive "before / after" comparison.
Separately improve the home screen: add search and balance the amount of ads and promo blocks.

Product impact

The research helped us choose the single list as the foundation for the multi-seller cart and shape the requirements for refining it. This approach keeps every item visible, lowers the risk of duplicate adds, and better explains the link between seller, total, and separate checkout — while still giving users control over which items to buy now and which to leave for later.