The Brick Drop · Product specification · v1

The Brick Drop app — the spec

The complete plan for the app, written in plain language. Every decision here came from your teardown and the calls we made together — this is the coherent thing we rebuild toward, with the Drop Score engine at its core. Draft for review

2026-07-04 · supersedes the bolted-on Brand Revamp · companion blueprint: coherent-design-v1.html

01What we're building

An easy-to-use app that tells a LEGO fan how a set rates for them, what makes it good, and the full story behind its value — catered to their interests.

Type in (or tap) any LEGO set and get one honest letter grade, the plain-language reasons behind it, and a downloadable card worth sharing. It works instantly for a total stranger, and gets sharper and more personal for anyone who wants that. The grade is powered by The Drop Score — your review-based engine — with the True Value price math folded in as one of its layers.

02The principles

The rules that keep it from bloating back into a maze. When a future decision is unclear, these break the tie.

• One grade, one engine, everywhere. The Result, the Home cards, the Browse tiles, and the share card all read the same grade from the same engine. No second code path — the drift that split the old timeline is designed out.
• Open to the grade, not the catalog. The app's front door is finding and grading a set, never a wall of browsing.
• Personalization is optional and progressive. Full value with zero setup. Tailoring is something you grow into, never a gate.
• Be honest about what we don't know. Show confidence, show receipts, and say "Early read" when the data is thin. Honesty is the brand's edge — never fake certainty.
• Let the photo and the grade do the talking. Image-forward everywhere; cut anything that doesn't earn its place.
• App feel for fans, web reach for everyone. A PWA with public result pages — no install wall on your most casual traffic.
• Plain language only. No engine jargon ever leaks to the user.
• Color means something. Near-black ground, LEGO yellow-forward, red second — the Drop Score design system (this document's own look).

03Platform & the shape

It's a PWA with public result pages

One web-based app. Fans can install it to their home screen and it behaves like a native app; strangers who tap a shared card land on an instant public web page for that set — no install, full result. One codebase, web reach and app feel. (Your current tool is already a PWA, so this builds on what exists.)

Three core screens, plus two supporting

Screen	What it's for	Nav
Home	Find a set (search) + a curated shelf. "For You" once you have a profile; "Notable drops" if you don't.	Tab 1
Result	The Verdict Card — the grade, the why, the receipts, the decision. The heart of the app.	opens from anywhere
Browse	The timeline, demoted: a calm, accurate month-by-month wall of 2026 drops.	Tab 2
Profile	Set up / edit / reconfigure personalization. Home of the (optional) onboarding.	Tab 3
Share card	A downloadable 4:5 image generated from a Result. Not a screen — an artifact.	from Result

Three tabs, not five. Watchlist/Saved is deferred (see §09), so it isn't a tab yet.

04The two engines

One headline grade, powered by two sources that each shine where they're strongest.

• The Drop Score — your holistic, review-based engine (judges, corpus, sentiment). It produces the headline grade, the three reads, and the Hot Take. It's the star when real reviews exist.
• True Value — the price-vs-parts math you love, kept intact but surfaced as the "Is the price fair?" read inside the breakdown. It's the one source that works before a set is even released.

How the grade behaves as a set ages (the living grade)

Because the Drop Score needs reviews and most of the timeline is unreleased, the headline adapts to how much we actually know:

Release state	Headline	What's shown
Unreleased no reviews yet	True Value grade	The grounded parts/price grade leads, with a clearly-labeled predicted Drop Score beside it (from theme, designer, price, piece value, leaks). All under an Early read frame.
Just launched a few reviews	Drop Score (emerging)	The real Drop Score appears with a low-confidence label ("Early read · 2 reviews"). True Value stays as the "price fair?" read. The prediction retires.
Well-reviewed	Drop Score	The Drop Score is the confident headline; True Value is one read among the breakdown. Full receipts.

05Personalization — optional, in three levels

Nobody is forced to set anything up. Tailoring is something a user grows into, and every level is a fine place to stop.

Level	What it is	Setup
0 · Casual	A general Drop Score (how a set rates for most people) + a home of notable, top-rated drops. Search anything, get a grade.	None
1 · Ad-hoc	On any Result, nudge "what matters to you" and that one set re-grades. The card then offers "save these as my defaults."	One tap, no account
2 · Make it yours	A ~30-second onboarding (or saved tweaks). Home becomes "For You" and grades tune everywhere.	Optional, any time

What the profile captures

• Onboarding (tight): why you buy (Display / Play / Collect / Invest) · themes you love · budget comfort. Fast, because friction kills the casual user.
• Per-set dials (deeper), only when relevant: how much minifigs matter, do you have space for this. These flip a con into a pro for specific people (a huge set is a con if you're short on space) — so they live on the Result, not in setup.

Personalization must show its work

When a profile shifts a grade, the app proves it, so tailoring never feels like a gimmick:

06The screens, in detail

Home — casual and "For You" are the same screen

• Search-first: a prominent "search any LEGO set" is always at the top.
• The shelf below adapts: "Notable 2026 drops" (general grades) for casual users; "Rated for you" (personalized) once a profile exists. Same layout, different contents and one label.
• A soft "Make it yours" nudge sits on the casual home — dismissible, and it also lives in Profile so it's never nagging.
• Tightly curated — a few picks, never a noisy feed. The old discovery lanes do not come back.

Result — the Verdict Card, the heart of the app

Image-forward like the share card it produces. The set photo fills the top as a full-bleed hero behind the verdict, with the same gradient overlay; the working controls sit on solid dark below.

• The grade leads, colored from the set's photo, over the image. Shows "for you · overall" when a profile changes it; tap for the reason.
• Confidence is always visible — "Based on 18 reviews · solid" or "Early read · 2 reviews." Never a bare letter.
• The Brick Drop Hot Take — the italic line over the image (this replaces the old "unique fact").
• Three expandable reads: 💰 Is the price fair? (this is True Value), 🤩 Will you love it?, 📈 Holds its value? Tapping a read reveals real review quotes behind it.
• Multi-retailer buy row — LEGO, Amazon, and more. Not LEGO-only.
• A buy-timing line — "Retiring soon · $120 off right now" (see §08).
• One "scores higher for you" alternative — a single tasteful suggestion, not a lane (see §08).
• Per-set tweak — "what matters to you" sliders (Level 1 personalization), which then offer to save as defaults.
• Share — generates the downloadable card, delivered via the bottom-sheet pattern.
• Every Result has a public web URL — this is the page a shared card opens.

Browse — the timeline, calmed and corrected

• Chronological by month, with months as quiet inline dividers (not a sticky jump-rail).
• Grid ↔ List toggle — Grid (image-forward, share-card tiles) is the default; List is the scan-fast density option.
• One simple filter row — All / Coming / Available / Retiring. Status badges only when they matter.
• Same engine, same grades — every tile reads the one true grade; personalized if you have a profile.
• Not the homescreen — it's the second tab, a "nice add-on," never the front door.

Onboarding & Profile — invited, never sprung

• Reached from the nudge or the Profile tab, never forced on a first-timer. Has a clear "Skip for now."
• ~30 seconds: why-you-buy · themes · budget (per §05).
• Reconfigurable any time from Profile — edit or redo it.

07The share card & the Crop Curator

The share card — a downloadable, postable image

• 4:5 portrait, saved as a hi-res image to the user's phone for social. This is its purpose.
• Locked DNA: full-bleed set photo, black gradient overlay, set name in a soft halo, and the grade letter colored from the image.
• Carries the Brick Drop Hot Take (not the old unique fact).
• Links back to the set's public Result page — the card is the top of the growth loop.

The Crop Curator — ship it as a real tool

• Your per-set crop/accent data (173 sets and counting) is the source of truth for how each photo frames on the card.
• Productionize it as a /crops page with a real save endpoint (no copy-paste JSON), an "alt image" cycle for photos that don't fit, and auto-save so work is never lost.

08Decision support — beyond the grade

A grade with no next step is a dead end. Two light touches turn a verdict into a decision — one line each, never the old maze.

• "Is now the time?" — a price/timing line: retiring-soon urgency and price context ("$120 under usual"). Data note this needs price history over time, so it may land a beat after the core grade.
• "Is there something better for me?" — one "scores higher for you" alternative. Comes free it rides on the same recommendation engine as the For-You home.

09What's in v1, and what's later

Your teardown, resolved into scope.

Call	Item
v1	The grade + one engine; Result (Verdict Card) with receipts, for-you/overall, buy-timing, alternative, per-set tweak; Home (casual + For You); Browse (grid/list, month spine, one filter); optional onboarding + Profile; 4:5 share card + image-color grade; Crop Curator as /crops; True Value as a layer; unreleased "Early read" + predicted grade; PWA + public result pages; multi-retailer buy.
Later pillar	Alerts & retention — opt-in "watch this set" for price drops / retiring / new-in-your-themes, which revives Saved/Watchlist. Named now, built after v1. This is the come-back-to-the-app engine.
Later pillar	Price history — the data behind the buy-timing line and price-drop alerts.
Cut	Timeline as homescreen · forced onboarding · calendar countdown hero · editorial discovery lanes · 4-row sticky header · month-jump rail · multi-select chips + sort menus · the separate "giant grade letter" screen · the "Is it worth it?" framing.

10How we'll know it worked, & open questions

Success looks like

• A total stranger can land on a shared card and understand a set's grade in seconds, with no setup.
• The grade a user sees on Browse always matches the grade on the Result. Zero drift.
• Users believe the grade — because the confidence and receipts are right there.
• Enough people cross from casual → "make it yours" that the For-You home earns its keep.
• Share cards get saved and posted — the loop turns.

Open questions to settle before / during the build

• The predicted-grade model — which signals, and how loudly do we label a prediction so it never reads as fact?
• Grade honesty vs. affiliate buy links — the buy row likely earns revenue; we must keep a hard wall so commerce never bends a grade (your "do not lie" rule).
• Coverage — "search any LEGO set" is a promise; where's the honest edge of what the engine can grade well, and how do we set expectations at that edge?
• How much a profile actually moves a grade — tune it so "for you" is meaningful but never feels arbitrary.

---

That's the whole thing: one app, one grade, made for you — the calm reset of everything that drifted, with the Drop Score engine at its heart.