Wishlists and save-for-later: intent capture that respects the moment

Most wishlists are decorative. A heart icon on the product grid, wired up to nothing, saving to an account the shopper doesn't have and doesn't want. The shopper taps it, gets a registration wall, and leaves. The store logs "wishlist engagement" in a dashboard nobody opens. Everyone's time is wasted.

A wishlist that's actually designed is one of the best intent signals a store can hold. Someone has told you exactly what they want, at full catalogue price, unprompted. That's more information than a paid click will ever give you. The job is to honour it: remember it across sessions, never hold it hostage, and pay it off with alerts the shopper is glad to receive. Here's the pattern set we build into every retail storefront we ship, hardened on projects like Willow & Wren's bookshop and the record browsers at Holloway Records.

Two drawers, not one

"Save for later" and "wishlist" get shipped as the same feature far too often. They're different intents with different clocks.

Save-for-later lives in the cart. The shopper has committed to some purchase this session; this item didn't make the cut. The horizon is days, not months. The right design is one tap from the cart line, prominent "move to cart" on return visits, and inclusion in cart-abandonment messaging rather than its own drip flow.

The wishlist is aspirational storage: gifts half-remembered, a sofa for the renovation that starts in autumn, an out-of-stock jacket. The horizon is months. It deserves its own surface, its own alerts, and no pressure to convert this week.

Conflate them and you get a wishlist full of abandoned-cart items receiving "still thinking about it?" emails, and a save-for-later tray buried three screens deep. Split them and each can behave honestly.

Anonymous save, merge-on-login

The single highest-leverage decision: let people save without an account. Store the wishlist locally, show it perfectly well to a guest, and persist it aggressively across sessions and browsers.

When that guest eventually logs in — usually at a checkout they've independently decided to complete — merge silently when there's no conflict and ask exactly once when there is: "You've saved 4 items in this browser and 9 on your account. Combine them?" Default to combine. Never gate the merge behind onboarding. Never, ever invalidate the local list to force the issue.

We wrote about the same principle at the cart level — the cart is a negotiation, and losing a shopper's remembered state is how you lose the negotiation. The wishlist is the cart's patient cousin; the same manners apply, over a longer period.

The guest wishlist changes your entire funnel maths. Instead of forcing registration for a weak signal, you earn registration at the moment of genuine value — the alert email click, the shared registry, the checkout.

The payoff loop: alerts worth opening

A wishlist without alerts is a notebook in a drawer. The feature earns its keep through two trigger types:

Price drops. Item on wishlist, price decreases below the price at save time (or by a meaningful threshold — we use 10% or $10, whichever is larger, to avoid alerting on rounding). Send once per item per price drop, not daily reminders about the same drop. The email leads with the item, the old price, the new price, a direct "buy now at this price" link, and nothing else. No cross-sell carousel competing with the payoff.

Back in stock. The wishlist is the natural home for restock interest — far warmer than a standalone "notify me" widget, because the shopper already curated the item. One email at restock, one reminder after 72 hours if unopened, then silence. If it sells out again, say so honestly and re-arm the alert automatically. This sits naturally alongside the broader honest inventory patterns we've written up separately.

Cadence discipline is what separates a programme from spam. Our house rules, encoded in the lifecycle architecture we run for clients: cap wishlist-triggered email at two touches per week across all items, batch same-day triggers into one email, respect quiet hours in the recipient's timezone, and make per-item unsubscribe one tap — muting an item is not unsubscribing from the store. A shopper who mutes three items and keeps the account is a retained asset. A shopper who unsubscribes from everything because you mailed them nine times about one lamp is gone.

Gift-registry adjacency

Wishlists that can be shared convert differently — and get purchased by other people, which is the only category of traffic with zero acquisition cost. A read-only share URL, no account required to view, items marked as purchased without revealing the buyer (the registry courtesy rule), and gift-notes flow is a genuinely different product than a private list.

If gifting is a meaningful slice of your calendar, the full treatment is in our gift commerce flows piece. The short version: a shareable wishlist is a registry with the formality removed, and it deserves its own measurement because its buyer is not its curer.

Measuring wishlist-to-purchase without attribution fantasy

Here's where wishlist programmes die in the boardroom: the measurement. The naive report — "wishlist users buy 4.1× more" — is survivorship in a trench coat. People who save items were already higher-intent. The wishlist didn't cause it.

The honest approach:

1. Define a conversion window — purchase of a wishlisted item within 30 days of an alert touch counts toward the programme; organic purchases of wishlisted items with no alert count as list value, not alert value. Splitting these stops you claiming credit for what would have happened anyway.
2. Run matched cohorts — compare wishlisters who received alerts against wishlisters whose alerts were withheld (a small holdout) for a quarter. The lift number that survives a holdout is the one you put on a slide. We went deep on this style of reasoning in honest attribution.
3. Track saves per returning session — the wishlist's deeper job is giving shoppers a reason to come back to your site instead of searching again. Saves per returning session is a leading indicator of that habit; alert revenue is the lagging one.

Expect the honest number to be a third of the flattering one. It's still excellent, and now nobody can torpedo it with a spreadsheet.

Prune and archive, like a garden

Wishlists rot. Items get discontinued, prices double, the renovation gets cancelled. A wishlist full of ghosts looks abandoned and, worse, its alerts bounce into dead product pages.

Design the endings:

• Unavailable items stay visible as clearly archived entries ("This edition has sold out"), with one well-chosen alternative — not a delete. Shoppers saved that item as a record of taste; respect it. A hard delete reads as the store forgetting them.
• An annual "spring clean" email listing items untouched for 12 months, with one tap to keep-all, prune, or convert stale saves into a fresh alert on newer equivalents. This email runs at reliably high open rates precisely because it isn't selling anything.
• Never expire the list itself unless the shopper asks. The list is the reason they return. Treating it as session-scoped data is the original sin of the default heart icon.

A pruned, alert-armed, shareable wishlist is a tiny CRM the customer maintains for you. That's the real feature — the heart icon was only ever the label.

Key takeaways

• Save-for-later (cart, days) and wishlist (aspirational, months) are different intents; ship them as separate surfaces with separate messaging.
• Anonymous wishlists with merge-on-login out-convert registration walls by an embarrassing margin — earn the account at checkout, not at the heart icon.
• Alerts are the payoff: one email per genuine price drop, one at restock plus one reminder, capped globally per week, with per-item mute.
• Measure alert lift with a holdout cohort and split organic wishlist purchases from triggered ones — the honest number is smaller and bulletproof.
• Archive dead items with alternatives instead of deleting; send an annual review email; never expire a shopper's list.

FAQ

Should wishlists require an account? No. Store them locally for guests and merge on login. Forced registration converts a warm intent signal into an abandoned session. You get the account later, at checkout or the alert signup, when the value is obvious.

How often can we email wishlist activity? Two triggered touches per shopper per week, batched per day. One email per price drop, one plus a single reminder per restock. Frequency is the whole game — the wishlist alert has the highest open rate in your programme because it's specific, and every extra send spends that trust.

What's a healthy wishlist-to-purchase rate? Across retail projects we've measured, 8–15% of wishlisted items are purchased within 90 days when alerts are on, roughly half that without. But read it as a curve by category, not a target: considered purchases (furniture, hi-fi) sit low and slow; consumables and gifts convert fast. Compare like with like.

Do shared wishlists / registries need their own checkout? No — they need their own view. A public, read-only list page where items can be marked purchased (without exposing the buyer) and gifted at the normal checkout. One extra page template; meaningful revenue during gifting seasons.

Should we delete discontinued items from wishlists? No. Show them as archived, explain briefly, and offer one alternative. Deleting erases the shopper's memory of their own taste — the list is theirs, you're just the filing cabinet.