StackBlueprint

Turns "what's the best way to buy this" into one auto-computed, shareable answer — pulling live portal rates and your own wallet's card-category bonuses into the stacking math OfferStack already has, instead of leaving you to look it up by hand.

Problem

OfferStack's Portal Stack Builder calculator already runs the correct math for stacking a shopping-portal rate, a card's category bonus, and a coupon — but a user still has to type in the portal rate and the card rate by hand, every time, for every retailer. Meanwhile the portal-rate sync pipeline refreshes a live portal_rates table every six hours, and the wallet already knows which cards a user holds — neither one talks to the calculator. A user shopping at a specific retailer today still opens Rakuten in one tab, checks their card issuer's current quarterly category in another, and does the arithmetic themselves, which is exactly the fragmented, manual process OfferStack exists to replace.

Target user

Primary: the Deal Hunter persona from STRATEGY.md §2 — arrives from a search like "best card for [retailer]" wanting one clean, trustworthy answer, not a 1,500-word blog post. Secondary: the Active Optimizer, who already has a full wallet in OfferStack and wants the stack resolved automatically against their own cards instead of running the calculator by hand before every purchase.

MVP scope

• A /tools/stack-blueprint page: pick a merchant, enter a purchase amount, and see the live portal rate pulled straight from the portal_rates table (already synced every six hours by the existing portal-rate pipeline) instead of typed in by hand.
• Reuse the locked Portal Stack Builder formula from VALUATION_FORMULA_SPEC.md unchanged — same inputs, same "Show math" audit panel pattern already shipped for the four calculators, just auto-populated instead of blank.
• A small data/card_category_bestfor.json rule pack mapping merchant categories to which card in the catalog currently earns the best rate there, refreshed manually until a real card-category feed exists — the recommendation must show its source card and category, not a black-box pick.
• Wallet-aware mode (auth-gated): cross-reference the user's own saved cards against that rule pack so the recommended card is one they actually hold, not a generic best-in-catalog pick.
• A one-tap optional field for a targeted Amex Offer or Chase Offer percentage the user already has, folded into the same stack math.
• A shareable result card — merchant, portal, card, and the resulting effective discount rate as a static image — matching the "shareable stack blueprints" mechanic the product strategy already names but has never built.
• Free tier: a public, un-personalized blueprint for roughly 25 high-traffic merchants already in the portal-rate table, generic best-in-catalog card, no auth. Pro: the wallet-personalized version, the Amex/Chase Offer field, and a saved history of blueprints.

Monetization

none (internal tool) — feeds the existing OfferStack freemium model. The public per-merchant blueprint lookup is the SEO hook (keyword cluster: "best card for [retailer] 2026", "rakuten stacking guide", "portal stacking calculator"), matching the calculator-driven SEO strategy already laid out in STRATEGY.md §4.1. The wallet-personalized resolver, the offer field, and saved blueprint history sit behind the existing $5–8/mo Pro tier from STRATEGY.md §5.

Why now

Independent 2026 guides on cashback stacking are multiplying — WalletGrower's "15-25% back" guide, Kudos's portal-stacking guide, and PerkStacked's beginner guide all cover the same ground this year, and all of them are static articles that leave the reader to do their own lookup and arithmetic per purchase. That's the exact gap STRATEGY.md §3 says OfferStack is built to close against content sites like The Points Guy — a tool instead of another guide. The pieces to build this already exist and sit unconnected: the portal-rate sync pipeline (portal_rates, live since June 2026) and the Portal Stack Builder calculator (locked math, shipped) never reference each other, and the "shareable stack blueprints" viral mechanic named in STRATEGY.md §4.3 and required in PROJECT_REQUIREMENTS.md §4.11 has no backlog item building it yet.

Risks & open questions

• Demand-side: sharing a stacking result is a nice-to-have, not an obvious Pro-tier reason to pay. Validate whether the free public blueprint actually gets shared before the wallet-personalized resolver, or a bookmarks-style feature, would convert better.
• Build-side: rotating quarterly category bonuses (Chase Freedom Flex, Discover it) mean card_category_bestfor.json goes stale within 90 days without an owner and a refresh cadence — the recommendation needs a visible "as of" date or it quietly becomes wrong.
• Affiliate-bias optics: the card recommendation touches affiliate revenue directly. It has to be pinned to the math-first rule in STRATEGY.md §6.3 — never surface a higher-payout card over the one the math actually favors.
• Coverage gap: the portal_rates table only holds merchants the existing sync pipeline already tracks; a searched merchant with no row needs to show "no live portal data" plainly, not silently default to 0%.
• Share-image infra: rendering a static shareable result image is new infrastructure for a solo build — worth scoping as a plain server-rendered card before reaching for anything more elaborate.

Next step

Promote to a weekly prototype — build /tools/stack-blueprint as a free-tier, public, 25-merchant blueprint page reading live portal_rates data with a generic best-in-catalog card recommendation, no auth and no share-image yet, to validate the SEO and share hook before scoping the wallet-personalized resolver and the image render as separate LIF-### items.

Sources

• https://walletgrower.com/blog/cashback-stacking-how-to-earn-15-25-back-on-every-purchase — 2026 guide on layering portal, card, and coupon cashback to 15-25% back; confirms the stacking pattern this tool automates
• https://www.joinkudos.com/blog/credit-card-shopping-portals-stacking-guide-how-to-double-or-triple-rewards — 2026 shopping-portal stacking guide; representative of the static-content competition OfferStack differentiates against
• https://thepointsguy.com/loyalty-programs/how-to-use-rakuten/ — Rakuten + card stacking mechanics, confirms the manual click-through-then-pay workflow this tool shortcuts
• D:/AI_Agents/Projects/Mikes_AI_Lab/Repos/Live_Apps/life-hacker/docs/STRATEGY.md — personas (§2 Deal Hunter, Active Optimizer), SEO strategy (§4.1), the named-but-unbuilt "stack blueprint" viral mechanic (§4.3), freemium pricing (§5), trust-first positioning (§6.3)
• D:/AI_Agents/Projects/Mikes_AI_Lab/Repos/Live_Apps/life-hacker/docs/PROJECT_REQUIREMENTS.md — §4.11 viral & share mechanics requirement for shareable stack blueprints, still unbuilt
• D:/AI_Agents/Projects/Mikes_AI_Lab/Repos/Live_Apps/life-hacker/docs/EXECUTION_BACKLOG.md — confirmed LIF-035 (Portal Stack Builder calculator) and LIF-046 (portal_rates sync pipeline) are both shipped and unconnected; no open LIF-### item covers a stack blueprint or share-image feature
• D:/AI_Agents/Projects/Mikes_AI_Lab/Repos/Live_Apps/life-hacker/docs/VALUATION_FORMULA_SPEC.md — Calculator 2 (Portal Stack Builder) formula, reused unchanged