mcpField
M
micro-commitment-stacking

micro-commitment-stacking

github.com/louisblythe营销与增长190414 次安装40 周前更新
代码仓库

关于 micro-commitment-stacking

面向销售 Bot 开发者的技能,帮助通过一系列小承诺引导潜在客户逐步作出更大的决定。内容涵盖承诺阶梯、对话流程、请求尺度和进度追踪,并强调尊重拒绝。

主要功能

  1. 01渐进式承诺阶梯
  2. 02认同提问与“如果可以”技巧
  3. 03请求尺度与分阶段 CTA
  4. 04承诺追踪
  5. 05拒绝后的回退流程

适用场景

  1. 01设计销售 Bot 的客户对话
  2. 02选择合适的下一步请求
  3. 03追踪潜在客户的承诺阶段

安装

复制下面的命令,在终端中运行。

$npx skills add https://github.com/louisblythe/sales-skills --skill micro-commitment-stacking

安装后在 Claude Code 中输入/micro-commitment-stacking确认已加载。

详细介绍

微承诺叠加

你是打造销售机器人方面的专家,擅长通过达成小的共识来促成更大的承诺。你的目标是帮助开发者构建系统,运用心理承诺原则,通过一系列小小的肯定,引导潜在客户做出决定。

微承诺为何有效

大要求的问题
First message:
"Would you like to schedule a 60-minute
demo with our sales team?"

Response: *Ignored*

Why: Too much commitment for cold prospect.
No relationship established. Too big a leap.
承诺阶梯
Micro-commitment approach:

Step 1: "Does improving [pain] matter to you?"
→ "Yes" (small agreement)

Step 2: "Would a 2-minute overview help?"
→ "Sure" (small time commitment)

Step 3: "That resonates? Worth a deeper look?"
→ "Yes" (growing interest)

Step 4: "Can we schedule 15 minutes this week?"
→ "Okay" (meeting commitment)

Each yes makes the next easier.

承诺心理学

一致性原则
People want to be consistent with past behavior.

Once someone says "yes" to something small,
they're more likely to say "yes" to something
related—to maintain self-consistency.

"I agreed X matters, so I should act on it."
递进阶梯
Tiny → Small → Medium → Large

Tiny commitments:
- Acknowledge a problem exists
- Agree something is interesting
- Reply to a message

Small commitments:
- Watch a video
- Download a resource
- Answer a question

Medium commitments:
- Take a call
- Complete a survey
- Share information

Large commitments:
- Demo meeting
- Trial signup
- Purchase decision

微承诺技巧

共识式提问
Get them agreeing early:

"Would you agree that [common pain point]
is a challenge for teams like yours?"

Almost everyone says yes → commitment made.

"Is [outcome] something you'd want to achieve?"

Yes → now they're invested in the outcome.
“如果……就……”框架
Instead of: "Want a demo?"
Ask: "If I could show you how to [benefit]
in 15 minutes, would that be worth your time?"

If yes → they've pre-committed to the value.
If no → they've told you their objection.

"If I could prove [claim], would you be open
to exploring this further?"
从小请求开始
Start impossibly small:

"Mind if I send you a 1-page summary?"
→ Very low commitment

Then: "Was that helpful? Happy to expand
on any section."

Then: "Based on what you found interesting,
want to see it in action?"

Build from nothing.
二选一承诺
Don't ask: "Do you want to meet?"
Ask: "Would Tuesday or Thursday work better?"

Assumes commitment, offers choice.

"Would you prefer a quick call or should
I send more info first?"

Both options = some form of engagement.

实施

承诺跟踪
class CommitmentTracker:
    def __init__(self, prospect_id):
        self.prospect_id = prospect_id
        self.commitments = []
        self.commitment_score = 0

    def record_commitment(self, commitment_type, value):
        self.commitments.append({
            "type": commitment_type,
            "value": value,
            "timestamp": datetime.now()
        })

        # Update commitment score
        commitment_weights = {
            "acknowledged_problem": 10,
            "agreed_with_statement": 10,
            "replied_to_message": 15,
            "answered_question": 15,
            "watched_video": 20,
            "downloaded_content": 20,
            "took_call": 40,
            "attended_demo": 50,
            "started_trial": 60
        }

        self.commitment_score += commitment_weights.get(commitment_type, 10)

    def get_next_commitment_level(self):
        if self.commitment_score < 20:
            return "tiny"
        elif self.commitment_score < 50:
            return "small"
        elif self.commitment_score < 80:
            return "medium"
        else:
            return "large"
请求规模
def get_appropriate_ask(prospect, commitment_level):
    asks = {
        "tiny": [
            "Does this resonate with you?",
            "Is [problem] something you deal with?",
            "Mind if I share a quick insight?"
        ],
        "small": [
            "Can I send you a brief overview?",
            "Worth a 2-minute look?",
            "Any questions I can answer?"
        ],
        "medium": [
            "Would a 15-minute call be helpful?",
            "Can I show you a quick demo?",
            "Want to try it free for a week?"
        ],
        "large": [
            "Ready to discuss next steps?",
            "Should I send over a proposal?",
            "Can we schedule an implementation call?"
        ]
    }

    return asks.get(commitment_level, asks["tiny"])
渐进式 CTA
def select_cta_for_stage(prospect):
    tracker = get_commitment_tracker(prospect.id)
    level = tracker.get_next_commitment_level()

    # Never jump more than one level
    current_max = tracker.commitment_score

    if current_max < 15:
        # Haven't agreed to anything yet
        return CTAs["acknowledge_problem"]

    elif current_max < 30:
        # Have acknowledged problem
        return CTAs["share_content"]

    elif current_max < 50:
        # Have engaged with content
        return CTAs["quick_conversation"]

    elif current_max < 70:
        # Have had conversation
        return CTAs["demo_or_trial"]

    else:
        # Ready for serious discussion
        return CTAs["proposal_or_next_steps"]

对话流程

早期阶段流程
Message 1:
"Is [pain point] something your team
deals with?"

Response: "Yes, definitely"
→ Record commitment: acknowledged_problem

Message 2:
"I've got a quick framework that addresses
exactly that. Mind if I share?"

Response: "Sure"
→ Record commitment: agreed_to_content

Message 3:
"Here it is: [content]. Does any of this
match what you're experiencing?"

Response: "The part about X resonates"
→ Record commitment: engaged_with_content

Message 4:
"Want to see how it works in practice?
Just takes 10 minutes."

→ Building toward meeting commitment
对方拒绝后的应对
If they say no to an ask:

Don't: Push harder on same ask
Do: Step back one level

"No problem—would you prefer I just send
some info for when the timing's better?"

Or: "Totally understand. If [problem]
becomes more urgent, feel free to reach out."

Preserve relationship for future commitment.

承诺话术

促成肯定的表达
Low pressure starters:
- "Would you agree that..."
- "Is it fair to say..."
- "Most people find..."
- "Would it be helpful if..."
- "Makes sense?"

Assumptive continuations:
- "Since you mentioned X..."
- "Given that you're dealing with..."
- "Based on what you said..."
- "Now that you've seen..."
锁定承诺的表达
After they agree:
- "Great, so we're aligned that..."
- "Good—so the key thing is..."
- "Perfect, let's build on that..."
- "Exactly—and the next step is..."

Reference previous commitments:
- "You mentioned [pain] matters..."
- "Since you found [content] useful..."
- "Given you're interested in [benefit]..."

反面模式

不该怎么做
Don't:
- Ask for big commitment first
- Jump multiple levels at once
- Ignore their previous agreements
- Push after a "no"
- Make false scarcity claims

Do:
- Start small, build up
- Progress one level at a time
- Reference their past agreements
- Step back gracefully on "no"
- Create genuine value at each step
避免给人操纵感
The line between persuasion and manipulation:

Persuasion (good):
- Genuine value at each step
- Honest about what you're asking
- Respect "no" answers
- Build real relationship

Manipulation (bad):
- Tricks without value
- Hidden agendas
- Ignoring objections
- Purely extractive

指标

承诺漏斗
Track progression:
- Acknowledged problem: X%
- Engaged with content: Y%
- Took call: Z%
- Demo/trial: A%
- Closed: B%

Identify:
- Where do people drop off?
- Which commitments predict success?
- What's the optimal path?
承诺速度
Track speed:
- Time from first yes to meeting
- Number of commitments to close
- Correlation between early commits and close

Optimize:
- Right number of steps
- Right size of asks
- Right pacing

评论

支持 Markdown