The Vault — Saved Cart

This is the feature-level case study behind The Vault’s storefront redesign — the buying journey specifically: search through checkout, and the one gap the whole category shared until this shipped.

Role

Sr. Product Designer

Scope

Search, product, cart, checkout

Timeline

Dec 2025 – Jun 2026

Context

The Vault Co · in-house since 2023

1. Search→
2. Product→
3. Save cart→
4. Add to cart→
5. Checkout→
6. Confirmation

Every decision below is argued from these two lists at once — a feature that served one and not the other didn’t ship.

What buyers were trying to do

What the business needed

Collectors comparing cards across sellers hit the same wall everywhere: no way to save a cart and come back to it. TCG buyers don’t impulse-buy — they build a decklist over days or weeks, compare prices across sellers, wait for payday or for a specific card to hit the market. Every competitor forced an immediate decision or the cart was gone.

Layer a category-wide trust problem on top — unverified sellers, no way to confirm a card’s condition before it ships — and an information gap — no real-time pricing, no rarity context — and the pattern across every existing platform was the same: built to move inventory, not to help someone shop.

1. 01

   Objectives & scope

   The questions every later decision had to answer: how buyers search, why carts get abandoned, what it takes to trust a stranger with a payment.

2. 02

   Competitive & heuristic benchmarking

   TCGPlayer, Cardmarket and Hobby Mart on transaction features; Card Kingdom, Star City Games, Yellow Rabbit, MTG Wolf and MTG Mexico as outside reference points.

3. 03

   Pain-point synthesis

   Four recurring breakdowns, ranked by how consistently they showed up across the research above.

4. 04

   Personas & empathy mapping

   The player, the collector and the returning buyer — and what each one thinks, feels and does mid-purchase.

5. 05

   Design decisions

   Persistent Saved Cart, deferred checkout information, transparent seller data — specified against the existing design system.

6. 06

   Success metrics

   The Google Analytics + Shopify funnel this project is read against in Results.

The comparison behind that pain point was concrete. I ran a structured competitive analysis against TCGPlayer, Cardmarket and Hobby Mart across the features that actually decide whether someone buys: advanced search, wishlist, saved cart, multi-seller checkout, seller reputation. Three of the four matched each other almost feature-for-feature. All three shared one identical gap.

The decision: build Saved Cart as a first-class feature — not a session cart that clears on close, but a persistent one tied to the account, positioned as core to the journey rather than bolted on after checkout.

I validated the direction by benchmarking established players outside the direct competitive set — Card Kingdom and Star City Games internationally, Yellow Rabbit, MTG Wolf and MTG Mexico locally — rather than running it through user interviews first. That’s competitive and heuristic validation, not primary research, and it’s worth stating plainly rather than implying otherwise.

Both screens from the sketches — the product page and the cart — taken to information architecture and flow before any visual design. The open questions at this stage were placement and hierarchy: where the save action sits, what a saved cart is inside the account, and what a buyer sees when they return to one.

None of the decision above, or the friction work that followed it, survives contact with a real product without a system underneath it. Saved Cart and checkout run on the exact tokens documented in full in the ecosystem case study — built once, applied here first, and now reused everywhere else The Vault ships.

Black is the product. The canvas is pure black, so card art is the only saturated thing on screen — including inside the cart, where a product thumbnail should read before its price does.

The landing-page redesign tripled traffic in the same window, which makes it tempting to credit the whole funnel to discovery. The stage conversions say otherwise: the share of sessions adding a card to cart and reaching the payment screen barely moved. What changed was the last few seconds — of the people who reached payment, 34% completed checkout before this shipped, 63% after.

That’s the fingerprint of deferring required information to the last step instead of the first. It didn’t pull more people into the funnel or get more of them to add a card to cart — it kept the people already close to buying from dropping off at the finish line.

Two honest limits. The baseline period recorded twelve completed checkouts — small enough that a handful of transactions swings the percentage hard, so the direction is trustworthy and the precision isn’t. And because the landing page and this redesign shipped in the same window, I can tell you the last step tightened, not with certainty how much of that belongs to Saved Cart versus the checkout-flow changes specifically. Per-surface attribution is the first thing I specified on the next feature.