Value Engineering Brief

How to use it

1. 1Paste the prompt into ChatGPT, Claude, or Gemini.
2. 2Fill the bracketed inputs (product, cost pressure, cost structure, quality anchors).
3. 3Answer its clarifying questions about constraints and non-negotiables.
4. 4Iterate: ask it to sequence the levers into savings waves.

See it in action

Cost hotspot map for Hearthside (directional — confirm against your costing):
- Recipe: the cheese blend is almost certainly the largest single cost; blend ratio and coverage pattern are the levers.
- Pack: printed inner liner and box grade look over-spec for a freezer environment.
- Process: par-bake time and yield loss are worth an ops review.
- Logistics: box cube efficiency per pallet.

Noticeability ranking (least to most noticeable):
1. Box grade and liner simplification — invisible in use.
2. Cheese coverage pattern (same edge coverage, adjusted center density) — low risk; test blind.
3. Cheese blend ratio shift — medium risk; the mozzarella pull is a Hearthside signature.
4. Topping count reduction — high risk; visible in the pack photo and at first bake. Do not lead with this.

Do-not-touch list: rising-crust texture, browning, cheese pull. These carry the promise.

Test plan: paired blind bake-offs for levers 2-3; an in-home use test before any visible change ships.

Margin story: frame internally as protecting price position and avoiding a shelf-price increase — not as cheapening.

Want me to draft the blind bake-off protocol for the cheese levers?

The prompt

Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.

# Role
You are a senior value-engineering lead for food manufacturers. You protect the attributes consumers actually buy the product for and hunt cost everywhere else. You refuse to present a cost-down list that has not been ranked by consumer noticeability, because that list always ends in a quality crisis.

# Context I'll provide
- Product and category: [PRODUCT]
- Cost pressure and target: [COST PRESSURE e.g. input inflation on cheese, meaningful COGS relief this year]
- What I know about the cost structure: [COST STRUCTURE — rough splits help; write 'unknown' if unknown]
- Quality anchors: [QUALITY ANCHORS — what consumers praise or would riot over]
- Constraints: [CONSTRAINTS e.g. no price increase, no visible downsize, clean label]
- Past changes and consumer reactions (optional): [HISTORY]

# Your task

Frequently asked questions

What is value engineering in food and beverage?

Value engineering is systematic cost reduction on an existing product that protects what consumers buy it for. It examines recipe, packaging, process, and logistics for savings, ranks each candidate change by how noticeable it would be to consumers, and tests before committing. Done badly it is quiet cheapening; done well, shoppers never notice and margin recovers without a price increase.

How is this different from the Reformulation Brief skill?

Same toolbox, opposite trigger. The Reformulation Brief is driven by health and nutrition goals — cutting sugar or sodium, cleaning up a label. The Value Engineering Brief is driven by cost, with quality parity as the hard constraint. If your project must deliver both, run this one first for the cost map, then fold its levers into the reformulation work.

What data should I feed it?

Whatever you have. A rough COGS split, spec sheets, and pack details sharpen the cost map considerably; consumer review themes and complaint logs sharpen the do-not-touch list. With nothing but a product description it still produces a directional plan — it labels its cost assumptions and tells you which ones to confirm against real costing first.

What AI model should I use for this skill?

Any capable chat model works — ChatGPT, Claude, or Google Gemini; the prompt is model-agnostic. Because cost programs run for months, many teams save it as a Custom GPT or a Claude Skill and rerun it each wave, feeding back what testing showed so the next round of levers starts from evidence rather than memory.