value-curve — скилл для Claude Code: анализ ценности
Скилл value-curve для Claude Code строит кривую ценности продукта для одного сегмента покупателей. Сравнение показывает, какие потребности закрыть лучше конкурентов и какую цену предложить.
MIT ★ 741 обновлён 2026-11-14
/plugin marketplace add phuryn/product-engineering
/plugin install product@product-engineeringЧто делает value-curve
Скилл помогает сравнить продукт с альтернативами по потребностям одного сегмента покупателей. Он предлагает оценить важность каждой потребности и удовлетворённость текущими решениями, затем вычислить возможность добавить ценность: важность × (1 − удовлетворённость). Если оценки основаны не на интервью, их нужно обозначить как предположения и назвать источник.
В итог входят таблица оценок, график и разбор: где продукту нужно опережать конкурентов, где достаточно паритета и на чём сегменту неважно конкурировать. Учитываются конкуренты, которыми сегмент действительно пользуется, и хотя бы одна замена — например, ручной процесс или таблица. Кривую строят максимум по семи факторам; цена тоже входит в сравнение, причём более низкая цена получает более высокую оценку.
Какие задачи решает
- сравнить продукт с конкурентами по потребностям сегмента
- выбрать потребности, по которым продукту нужно опережать альтернативы
- отделить обязательные для сегмента факторы от тех, на которых не нужно конкурировать
- сформулировать ценностное отличие продукта и его цену
Примеры запросов агенту
- Сравни наш продукт с альтернативами для сегмента, которому важны спортивные трансляции.
- Построй кривую ценности для выбранного сегмента и отметь предположения, которые нужно проверить.
- Сравни наши оценки с конкурентами и покажи, где достаточно паритета.
Как установить value-curve
1Через маркетплейс плагинов
/plugin marketplace add phuryn/product-engineering/plugin install product@product-engineeringКоманда ставит плагин product целиком, со всеми его скиллами
2Через Skills CLI (Vercel)
npx skills add phuryn/product-engineering --skill value-curve -a claude-code3Вручную: положить файлы
~/.claude/skills/value-curve/SKILL.mdПапка скилла целиком; для одного проекта — .claude/skills/value-curve/
1Через Skills CLI (Vercel)
npx skills add phuryn/product-engineering --skill value-curve -a codex2Вручную: положить файлы
.agents/skills/value-curve/SKILL.mdДля всех проектов — ~/.agents/skills/
1Через Skills CLI (Vercel)
npx skills add phuryn/product-engineering --skill value-curve -a cursor2Вручную: положить файлы
.cursor/skills/value-curve/SKILL.mdCursor читает и .agents/skills/, и .claude/skills/
1GitHub CLI
gh skill install phuryn/product-engineering product/skills/value-curve2Вручную: положить файлы
.github/skills/value-curve/SKILL.md1Через Skills CLI (Vercel)
npx skills add phuryn/product-engineering --skill value-curve -a gemini-cli2Командой платформы
gemini skills install https://github.com/phuryn/product-engineering --path product/skills/value-curveВручную: положить файлы
.devin/skills/value-curve/SKILL.mdСтарый путь .windsurf/skills/ тоже читается
Claude.ai
Упакуйте папку в zip и загрузите в Settings → Capabilities → Skills
По шагам, с командами для каждой платформы: Как установить скилл в Claude Code: пять способов
Вопросы о value-curve
Как установить value-curve в Claude Code?
Добавь маркетплейс командой /plugin marketplace add phuryn/product-engineering, затем установи плагин командой /plugin install product@product-engineering.
Что делает value-curve?
Скилл помогает построить кривую ценности для одного сегмента: сравнить потребности покупателей, продукт, конкурентов и заменители, а затем описать стратегические отличия.
Почему для каждого сегмента нужна отдельная кривая?
Потребности различаются между сегментами. Если свести их в одну кривую, приоритеты размоются.
Исходный текст
SKILL.md · MIT
--- name: value-curve description: "Comparing a product with competitors or choosing what to be best at: a value curve on one segment's top needs, plus price." --- # Value curve **Compare offerings on the factors one customer segment values.** The value curve comes from Blue Ocean Strategy (W. Chan Kim and Renée Mauborgne). It asks: does our offering meet this segment's specific needs far better than anything else they can use? **A good strategy competes to be unique, not the best at everything.** The curve is a starting point, not the whole answer. It feeds the Alternatives part of the value proposition (the `value-proposition` skill), which needs all six parts. ## Before you start - **One segment per curve.** Needs differ between segments, so mixing them shows no focus. If the segment isn't chosen, find clusters of customers with similar underserved needs first (`market-segments`, then `beachhead-segment`). - **The factors are the segment's needs, not your features.** Phrase each as a solution-agnostic outcome the customer wants, as jobs to be done are phrased: "find live sports events", "content for kids". - **Evidence comes from customers.** For every need, interviews ask its importance and satisfaction with what customers use today. Ask which competitors' products they use and what they would pay relative to those. A Van Westendorp price survey is one way. About 10-15 interviews per segment. - **With no interviews yet, label every score an assumption and name its source.** A curve can still use only websites, reviews and the owner's own knowledge. Its first output is then what to test with customers. ## Steps 1. **Score every need for the segment.** For each, *opportunity to add value = importance × (1 − satisfaction)*, with both on a 0-1 scale. This is Dan Olsen's formula from *The Lean Product Playbook*. It is simple and agrees with the jobs-to-be-done opportunity score where it matters most: high importance, low satisfaction. 2. **Pick 5-7 factors.** Include: - The needs with the highest opportunity scores. - Must-haves: needs this segment rates highly important even though today's products already satisfy them. Their low opportunity score doesn't make them optional; dropping one loses the customer. - Leave out needs this segment doesn't rate as important, even if competitors compete hard on them. - Relative price, scored so a lower price scores higher. Name factors with customer direction, such as "Low price for one host"; higher always means better for the customer. 3. **List the offerings.** Include your product (as planned, or as it is), the 2-4 competitors the segment actually uses, and at least one substitute. Substitutes include the spreadsheet, the agency, the manual workaround and doing nothing. 4. **Score each offering 0-100 on each factor.** Use what customers said and what reviews show, not feature lists. A feature that exists but doesn't serve the need scores low. 5. **Draw it.** Put factors on the horizontal axis, ordered by importance to the segment, and scores on the vertical axis. Draw one line per offering, with a legend. Keep the score table beside the chart, with each score's source. 6. **Read the curve and write its meaning in a few sentences.** Cover: - Where we must be far ahead: the segment's top-opportunity needs. If our line isn't clearly above everyone else's there, there is no reason to switch. - Where parity is enough: the must-haves. - What we deliberately don't compete on: factors where we sit at or below competitors because this segment doesn't care. Saying it out loud is the strategy. - Lines that look the same: the same shape as a competitor's means competing on degree, not difference. Blue Ocean's four actions help reshape it: which factors to eliminate, reduce, raise or create. **Every displayed score names its scale beside the number or in the subtitle.** **Word each action in the customer's direction.** For example: "Raise: one host pays less". Label an unbuilt factor "to create" while its score remains low. ## Output **Provide the segment, score table, chart, reading and assumptions to test.** Include: - The segment the curve is for, in one line. - The table: factors × offerings, each score with its source (an interview count, a review, or "assumption"). - The chart. - The reading (step 6), and the Alternatives line it gives the value proposition: our unique value against each competitor and substitute, and the relative price. - The weakest assumptions behind it, as what to test next. ## Example A streaming service for one segment: people who care about live sports events and concerts, and whose needs today's services underserve. The scores are illustrative, on a 0–100 scale. | Factor | Our product | Competitor A | Competitor B | |---|---|---|---| | Sports events | 95 | 40 | 30 | | Concerts | 90 | 35 | 45 | | Content for kids (must-have) | 70 | 75 | 70 | | Movies | 30 | 90 | 85 | | Relatively low price | 60 | 50 | 65 | - We aim to be the best at what this segment needs most: sports events and concerts. - Content for kids stays at parity. Customers are already satisfied with it, but it matters to them. - Movies are left low on purpose: this segment doesn't rate them as important. ## Mistakes - **Score needs rather than features.** Scoring features turns the curve into a feature comparison. - **Keep the curve to seven factors or fewer.** Past seven, the shape stops saying anything. - **Leave out factors the segment doesn't care about.** Don't keep them because competitors fight over them. - **Don't try to beat every competitor on every factor.** That is the opposite of a strategy, and usually unaffordable. - **Don't use one curve for several segments.** - **Don't treat assumed scores as findings.** Unlabelled guesses look exactly like evidence on a chart. Each rule's origin, the case that taught it, is in `receipts.md` beside this file; read it when a rule's edge is unclear. ## How you can tell it worked The curve is for one segment, and every score has a source or is labelled an assumption. Its reading says where the product must be far ahead and what it deliberately doesn't compete on.
© Paweł Huryn · MIT · phuryn/product-engineering