Verified against ChatGPT · 2026-08-14
Generate product bundle ideas that are checked against margin and inventory before you fall in love with one
Produces bundle concepts built from actual inventory and margin data, each flagged for viability, instead of a list of plausible-sounding bundle names that ignore whether the combination actually makes financial sense.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Generate bundle ideas from the product catalog below — but every bundle idea needs to survive a margin and inventory sanity check, not just sound like a good pairing. PRODUCTS AVAILABLE TO BUNDLE Ceramic mug ($12, 60% margin), pour-over dripper ($28, 45% margin), bag of coffee beans ($16, 35% margin), travel tumbler ($22, 55% margin) CUSTOMER BEHAVIOR DATA (what's actually bought together, if known) Order data shows the dripper and coffee beans are co-purchased in about 40% of dripper orders; the mug is rarely bought alongside either INVENTORY CONSTRAINTS Coffee beans are a fast-moving, easily restocked item; the pour-over dripper is on a 6-week reorder lead time and currently at 3 weeks of stock left TARGET DISCOUNT OR PRICE ANCHOR Aiming for a 15% discount off combined individual prices to make the bundle feel like a clear deal STEP 1 — GENERATE 5 BUNDLE CONCEPTS Propose 5 bundles, prioritizing ones supported by actual purchase pattern data over ones that just sound logically complementary — a data-backed pairing beats a plausible-sounding one you invented from category logic alone. For each, name the bundle, list the included products, and state in one line the actual reason a customer would want these together (a genuine use-case reason, not "customers who like A also like B" restated). STEP 2 — MARGIN CHECK For each bundle, using the margins I gave you, calculate roughly whether the bundle can absorb the target discount and still hit a reasonable blended margin, or flag it as margin-negative at the stated discount. Do not propose a discount percentage for me without checking it against the actual margins provided. STEP 3 — INVENTORY CHECK Flag any bundle that pairs a healthy-stock item with a constrained or low-stock item, since a bundle can create artificial demand pressure on the scarce component and cause stockouts on the item you can least afford to run out of. STEP 4 — RANK Rank the 5 bundles by overall viability (customer appeal + margin health + inventory safety combined), not just by which sounds most appealing on its own. WHAT NOT TO DO Do not recommend a bundle discount deep enough to be margin-negative just because it would look attractive to a customer — a bundle that loses money per unit sold isn't a promotion, it's a pricing mistake wearing a bundle's clothing. OUTPUT FORMAT Five bundles with the checks above, then a final ranked list with one-line reasoning per rank.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Asked generically for "bundle ideas," GPT-5.1 reliably produces plausible-sounding category pairings (mug and coffee, phone and case) because that's the pattern-matching task the request implies, without any mechanism to check whether the pairing is actually financially or operationally sound — bundle ideation and bundle viability are two different problems, and a prompt that only asks for the first will get confident-sounding recommendations that ignore the second entirely. Requiring the margin check to run against the actual numbers provided, rather than trusting the model's sense of what discount "sounds reasonable," matters because GPT-5.1 has no way to know a bundle's real cost structure unless the margins are given explicitly, and left unconstrained it will happily suggest an attractive-sounding 20-25% discount without any awareness that this could make a bundle unprofitable per unit — a mistake that looks like a promotion but functions as a pricing error once it ships. The inventory check step surfaces a failure mode specific to bundling that a purely creative brainstorm would never catch: pairing a healthy-stock item with a constrained one accelerates demand on exactly the component the merchant can least afford to run out of, potentially creating stockouts that then force the entire bundle offline mid-promotion. Prioritizing purchase-pattern data over category logic in the ranking step matters because two products can sound complementary by category (a mug and a dripper) while actual order data shows customers don't buy them together at any meaningful rate — the model has no way to know this unless told to weight real behavioral evidence over its own plausible-sounding inference.
What you get back
Bundle: Pour-Over Starter Set (dripper + coffee beans) — supported by real data (40% co-purchase rate). Margin check: combined price $44, blended margin roughly 41% before discount; at 15% off ($37.40), blended margin holds around 30% — still healthy, proceed. Inventory check: caution — dripper has only 3 weeks of stock at a 6-week reorder lead time; a successful bundle promotion could sell through remaining dripper stock before restock arrives. Recommend capping bundle promotion to available dripper units or expediting the reorder before launch...
Verified against
ChatGPT GPT-5.1 · 2026-08-14
Changelog
- 2026-08-14 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — Google Ads management, built and maintained for you — that's Scult's day job.
EXPLORE GOOGLE ADS MANAGEMENT
