- name
SKILL.md
Tệp skill vẫn nằm trong kho mã của nhà phát hành.
Dán lệnh đã sao chép vào terminal. Lệnh sẽ cài đặt tệp SKILL.md này vào agent đó.
npx skills add deanpeters/Product-Manager-Skills --skill activation-design -a cursor -yDán lệnh đã sao chép vào terminal. Lệnh sẽ cài đặt tệp SKILL.md này vào agent đó.
npx skills add deanpeters/Product-Manager-Skills --skill activation-design -a claude-code -yDán lệnh đã sao chép vào terminal. Lệnh sẽ cài đặt tệp SKILL.md này vào agent đó.
npx skills add deanpeters/Product-Manager-Skills --skill activation-design -a codex -yDán lệnh đã sao chép vào terminal. Lệnh sẽ cài đặt tệp SKILL.md này vào agent đó.
npx skills add deanpeters/Product-Manager-Skills --skill activation-design -a grok -yGửi tin nhắn đã sao chép trong cuộc trò chuyện Grok Bot. Bot sẽ lấy tệp SKILL.md này từ GitHub và thêm vào bộ skill mà mọi Bot của bạn dùng chung.
Install this skill: https://github.com/deanpeters/Product-Manager-Skills/tree/HEAD/plugins/pm-deliver/skills/activation-designTừ nhà phát hành
Define the moment a new customer first gets the value they bought for, and design the shortest reliable path to it. The output is an activation definition, a time-to-value measure, the blockers on the path, an onboarding path with owners and checkpoints, and the metric the team will track.
The sentence to keep in your head: a customer who hasn't reached first value hasn't really bought yet. Sign-ups, deliveries, and logins are activity. Activation is the first time the customer's own problem is solved. Customers who don't reach it quietly leave, and the churn shows up months later in a number that can't be traced to its cause.
It connects the launch (launch-plan), the funnel
(usage-data-investigation), and retention
(retention-diagnostic-advisor).
Works best with: The product, the customer, and what first real value looks like to them.
Also useful: Where customers stall today (funnel or interview evidence), the current onboarding steps and owners, time-to-value if you know it, and who touches the customer (sales, onboarding, support, partners).
Anything supplied with the invocation itself (text after the skill name, a pasted context dump, or
an appended ARGUMENTS: line) counts as answers already given. Use it and skip whatever it covers;
don't re-ask.
Arriving empty-handed? That works too. The skill asks what the customer is trying to get done
and what "it worked" looks like to them, proposes an activation moment as a hypothesis, and marks
every number it lacks [NEEDS SOURCE]. It will not invent a benchmark.
Example invocations:
Activation for Route Optimizer. A dispatcher optimizing a real day is the aha moment, I think.Commissioning plan for NFA-500. Units are installed but not running the connected service.The first time the customer gets the outcome they bought for, in their own terms. It is a behavior or a result, not a page view or a delivery. Good tests:
Measure the time from purchase or sign-up to activation, and watch the median and the slow tail. The tail is where customers disappear. Shorten the path by removing steps, pre-filling data, providing defaults, and doing work for the customer, in that order. In hardware, time to value includes shipping, installation, and commissioning, and the delays are often scheduling, not the product.
List the steps from purchase to activation, then for each ask: is it necessary, who owns it, and what stops customers here? Typical blockers: missing data the customer must supply, a decision that needs someone else's approval, a configuration nobody explained, an install that waits on a site visit, a handoff between sales and onboarding where context is lost. Use funnel data and a few conversations to find the real one, not the one you assume.
Build the onboarding as the shortest path to the activation moment, not a tour of features. Put one action per checkpoint, name the owner of each (the product, an onboarding specialist, a partner, the customer), and decide what happens when a customer stalls (a nudge, a call, an offer to do it for them) and after how many days. A path without stall handling is a hope.
The customer told sales what they want. If onboarding asks again, the customer notices. Write what
sales must capture (the outcome, the people, the constraints) and pass it on. Partner-led
onboarding needs the same handoff, plus training so the partner can reach activation, not just
installation. See buying-journey-enablement.
A self-serve add-on needs a short path and a nudge. A complex installation with several roles needs a plan with owners and dates. Don't build a 30-day program for something customers finish in 10 minutes, or a 10-minute tour for something that needs a site visit. Recommend a size and let the user move it.
Track the share of new customers who reach activation within N days (N set by your reality, not a
benchmark), with the count beside it. Also watch the median time to activate and where customers
stall. Pair it with a guardrail (support contacts per new customer) so you don't speed activation
by dumping work on support. See launch-metrics.
Use workshop-facilitation as the interaction protocol when
running this conversationally.
Default entry mode: Guided · set in pm-skills.config.yaml (library-wide default or a per-skill override); the user's own choice in the session always wins. On the first turn, offer all three modes, then open in the default without waiting for a pick: Guided asks the first question, Context dump invites the paste, Best guess asks for a one-sentence minimum and drafts with every assumption labeled.
Use template.md for the scaffold.
Output convention. Artifact first: the activation definition, time-to-value, the path with
blockers, the onboarding design (owners, checkpoints, stall handling), the metric. Commentary short,
after. Every figure is from supplied data or [NEEDS SOURCE]. Label whether the activation moment
is evidenced (customers who reach it retain better) or a hypothesis.
If the PM brings an onboarding plan, critique first. Is "activated" a customer outcome or activity? Is time to value measured, with the tail? Are blockers evidenced? Does each step have an owner? Is stall handling defined? Is the handoff from sales explicit? Name gaps; offer rewrites last.
What the customer is trying to get done and what "it worked" looks like to them. Propose the activation moment and label it evidenced or hypothesis.
Steps from purchase to activation, who owns each, how long each takes (from data or marked
[NEEDS SOURCE]), and where customers stop.
For each slow or leaky step: necessary? owned? what stops customers? Rank by how many customers it costs.
Shortest path, one action per checkpoint, an owner each, stall handling, and the sales-to-onboarding handoff.
Activation rate within N days with counts, median time, and a guardrail. Short summary (supplied,
inferred, unknown), then a numbered ladder, evidence-favored first, labeled (Recommended): check
whether activated customers retain better; remove the worst blocker first; add stall handling at
day 3; or interview five customers who stalled.
See examples/sample.md (Fieldlight self-serve onboarding, SaaS) and
examples/sample-industrial.md (Northfield commissioning and
connected-service activation, industrial).
Symptom: "Activated" means an account exists. Consequence: The metric rises while customers get nothing. Fix: Define it as the first real outcome, in the customer's terms.
Symptom: The moment is declared, never checked against retention. Consequence: The team optimizes a step that doesn't predict anything. Fix: Label it a hypothesis until activated customers are shown to retain better.
Symptom: Onboarding walks through screens. Consequence: Customers finish the tour and never reach value. Fix: Build the path to the outcome, one action per checkpoint.
Symptom: "Onboarding" owns everything. Consequence: Steps nobody owns are the steps that stall. Fix: Name the owner of each checkpoint.
Symptom: A customer goes quiet on day 4 and nobody notices. Consequence: The slow tail becomes churn. Fix: Define the nudge, the call, and the day each happens.
Symptom: Onboarding asks what sales already learned. Consequence: The customer feels unheard. Fix: Define what sales captures and passes on.
template.md: activation definition, path, blockers, onboarding design, metricexamples/sample.md, examples/sample-industrial.mdusage-data-investigation: find where customers stallretention-diagnostic-advisor: downstream; why customers leavelaunch-plan and launch-metrics: the launch around it, and its measurescustomer-journey-map: the customer's experiencebuying-journey-enablement: the handoff from sellingworkshop-facilitation: interaction protocolDistilled from practitioner experience designing onboarding for software and connected-hardware products.
Define the moment a new customer first gets real value, then design the shortest path to it. Use when trials stall, installs sit idle, or onboarding is a checklist nobody finishes.
---name: activation-designdescription: Define the moment a new customer first gets real value, then design the shortest path to it. Use when trials…---# plugins/pm-deliver/skills/activation-design/SKILL.mdCâu hỏi thường gặp
Tải plugins/pm-deliver/skills/activation-design/SKILL.md từ nhà phát hành, hoặc sao chép đường dẫn và mở kho mã. PlugMCP chỉ liệt kê skill và không lưu trữ tệp này.
Nhà phát hành duy trì nó tại https://github.com/deanpeters/Product-Manager-Skills.
Claude, Cursor, Codex và Grok Build có thể đọc tệp SKILL.md mà bạn thêm từ kho mã của nhà phát hành. Grok Bot cài đặt skill khi bạn gửi liên kết thư mục GitHub của skill trong cuộc trò chuyện.