mcpField
B
buying-signal-amplification

buying-signal-amplification

github.com/louisblytheマーケティング・グロース190410 回インストール40 週間前に更新
リポジトリ

buying-signal-amplification について

購買意向を検出し、応答を調整するセールスボットを設計する開発者向けのスキルです。シグナルのスコアリング、応答方針、対応振り分け、成果追跡を扱います。

主な機能

  1. 01明示・行動・会話のシグナル検出
  2. 02シグナルの組み合わせによる意向評価
  3. 03種類別の応答方針
  4. 04意向レベルに応じた後続対応の振り分け
  5. 05転換率と応答結果の追跡

活用シーン

  1. 01価格やトライアルへの関心の特定
  2. 02提示された時期に応じた応答調整
  3. 03意向の高い見込み客の担当者への引き継ぎ

インストール

コマンドをコピーしてターミナルで実行します。

$npx skills add https://github.com/louisblythe/sales-skills --skill buying-signal-amplification

インストール後、Claude Code で/buying-signal-amplificationを入力して読み込まれたか確認します。

詳細

購買シグナルの増幅

購買シグナルを検出し、戦略的に増幅するセールスボット構築の専門家です。開発者が、関心を示す兆候を認識し、購入意欲を強める方法で応答するシステムを構築できるよう支援することが目的です。

購買シグナルとは?

明示的なシグナル
Direct statements of intent:
- "What's the pricing?"
- "How quickly can we implement?"
- "Who else in our industry uses this?"
- "Can we do a trial?"
- "I need to run this by my team"

These are clear. Prospect is evaluating.
暗黙的なシグナル
Behavioral indicators:
- Multiple email opens (3+ times)
- Pricing page visits
- Case study downloads
- Demo video completion
- Return visits after initial contact
- LinkedIn profile views
- Forwarding emails internally

These require detection and interpretation.
会話中の言語的なシグナル
Language patterns:
- Future tense: "When we implement this..."
- Ownership language: "Our team would use..."
- Comparison questions: "How does this compare to X?"
- Timeline questions: "How long does setup take?"
- Stakeholder mentions: "My boss would want to know..."
- Budget signals: "Is there flexibility on pricing?"

シグナルの検出

キーワード認識
High-intent keywords:
- "pricing", "cost", "investment", "budget"
- "timeline", "implementation", "onboarding"
- "contract", "agreement", "terms"
- "decision", "approval", "sign-off"
- "trial", "pilot", "POC", "demo"
- "compare", "vs", "alternative", "difference"

Medium-intent keywords:
- "features", "capabilities", "integrations"
- "support", "training", "resources"
- "customers", "case studies", "references"
- "security", "compliance", "requirements"

Low-intent (but still positive):
- "interesting", "useful", "helpful"
- "makes sense", "I see", "that's good"
行動スコアリング
def calculate_intent_score(prospect):
    score = 0

    # Email engagement
    if prospect.email_opens >= 3:
        score += 15
    if prospect.link_clicks >= 1:
        score += 20

    # Website behavior
    if visited_pricing_page(prospect):
        score += 30
    if downloaded_content(prospect):
        score += 15
    if watched_demo_video(prospect, completion_rate=0.5):
        score += 25

    # Conversation signals
    if asked_about_pricing(prospect):
        score += 35
    if mentioned_timeline(prospect):
        score += 25
    if mentioned_stakeholders(prospect):
        score += 20

    return score  # 0-100+

# Thresholds
# 0-30: Low intent
# 31-60: Medium intent
# 61+: High intent (prioritize)
パターン認識
Signal clusters indicate high intent:

Cluster 1: Evaluation mode
- Pricing question + competitor mention + timeline ask
→ Actively comparing, decision coming

Cluster 2: Internal selling
- "My team" mentions + "my boss" + feature questions
→ Building internal case

Cluster 3: Ready to buy
- Contract terms + implementation + "next steps"
→ Decision made, working out details

Respond differently to each cluster.

増幅戦略

相手の言葉を反映し、発展させる
When signal detected, reinforce and deepen:

Signal: "What does implementation look like?"

Weak response:
"Implementation takes 2-4 weeks."

Amplified response:
"Great question—sounds like you're thinking about
the practical side. Implementation is typically 2-4 weeks.
Most customers in [their industry] see first value in
week one. Should I walk you through what the first
month typically looks like for a team your size?"

→ Validates their thinking
→ Provides answer
→ Opens deeper conversation
将来を想像させる
Help prospect visualize success:

Signal: "How do other marketing teams use this?"

Response:
"Marketing teams typically start with [use case A],
which usually saves 5-10 hours/week. Then they expand
to [use case B] once they see the time savings. In
6 months, teams your size are usually handling 3x
the campaigns with the same headcount.

What would your team do with an extra 10 hours a week?"

→ Creates mental picture of ownership
→ Asks them to project into future
→ Reinforces value
関係者に配慮する
When they mention others:

Signal: "I'd need to run this by our VP of Sales."

Response:
"Makes total sense—this is definitely a team decision.
What do you think she'd want to know? I can put
together a one-pager that hits her main concerns,
or we could set up a quick call with both of you.
Which would be more helpful?"

→ Validates the process
→ Offers to help them sell internally
→ Advances toward decision-maker
タイムラインを強調する
When urgency surfaces:

Signal: "We need to solve this before Q2."

Response:
"That's tight but doable. For Q2 readiness, we'd
want to kick off by mid-February. I've seen similar
timelines work when we do a focused pilot first.
Want me to sketch out what a February start
would look like?"

→ Confirms timeline is possible
→ Creates slight urgency
→ Proposes concrete next step

シグナルの種類別応答テンプレート

価格に関するシグナル
Signal: "What does this cost?"

"Happy to walk through pricing. To give you an
accurate picture—are you thinking about this for
just your team, or would other departments
potentially use it too? That affects which tier
makes sense."

Why: Reframes pricing as value discussion, gathers
qualifying info.
比較に関するシグナル
Signal: "How are you different from [Competitor]?"

"Good question—I'd imagine you're looking at them.
The main difference customers tell us: [key
differentiator]. That said, the best fit depends
on your priorities. What's most important to you
in this decision?"

Why: Addresses concern, redirects to their needs,
avoids trash-talking competitor.
試用に関するシグナル
Signal: "Can we try it before committing?"

"Absolutely—that's how most customers start.
We do a 14-day pilot where you'd use it with
real scenarios. By day 5 you'll know if it fits.
Want me to set that up? I'd just need to know
[qualifying question]."

Why: Enthusiastic yes, sets expectations, moves
toward action.
緊急性に関するシグナル
Signal: "We need to move fast on this."

"Understood—let's make sure we don't slow you down.
What's driving the timeline? If you need to be
live by [date], here's what the next steps would
look like: [concrete plan]. Should I send over
a timeline we can work backward from?"

Why: Matches their energy, shows competence,
provides structure.

増幅時に避けること

売り込みすぎない
BAD:
Signal: "This looks interesting."
Response: "It's the best solution on the market!
Let me schedule a demo, send pricing, and introduce
you to our CEO."

Too much, too fast. Kills the signal.

GOOD:
"Glad it resonates. What specifically caught your
eye? I want to make sure I point you to the most
relevant info."

Match their energy. Build incrementally.
シグナルを無視して先に進まない
BAD:
Signal: "What about security certifications?"
Response: "Great! Now let me show you our
reporting features..."

Ignored their concern. Lost trust.

GOOD:
"Security is critical—good that you're asking.
We're SOC 2 Type II certified, and I can send
the report. Any specific requirements I should
know about for your compliance team?"

Addressed concern. Explored further.
成約したと思い込まない
BAD:
Signal: "I'm meeting with my team next week."
Response: "Perfect, I'll send over the contract
to have ready for that meeting."

Presumptuous. Pressure.

GOOD:
"That's a good sign—sounds like it's worth
discussing internally. Would a summary deck
be helpful for that meeting, or is there
specific info they'll want that I can prep?"

Supportive. Helpful. Patient.

シグナルのスコアリングシステム

リアルタイムスコアリング
class SignalTracker:
    def __init__(self):
        self.signals = []
        self.score = 0

    def record_signal(self, signal_type, details):
        weight = SIGNAL_WEIGHTS.get(signal_type, 5)
        recency_bonus = 1.2 if is_recent(details) else 1.0

        self.signals.append({
            "type": signal_type,
            "details": details,
            "timestamp": now(),
            "weight": weight * recency_bonus
        })

        self.score = self.calculate_score()

    def calculate_score(self):
        # Recent signals weighted higher
        # Signal clusters get bonus
        # Decay over time
        return weighted_sum(self.signals)

    def get_intent_level(self):
        if self.score >= 70:
            return "HOT"
        elif self.score >= 40:
            return "WARM"
        else:
            return "COOL"
シグナルの重み
High weight (25-40 points):
- Asked about pricing
- Requested contract/terms
- Mentioned specific timeline
- Asked about implementation
- Requested references

Medium weight (10-24 points):
- Asked comparison questions
- Mentioned stakeholders
- Downloaded case study
- Revisited pricing page
- Watched demo video

Low weight (5-9 points):
- Multiple email opens
- General feature questions
- LinkedIn profile view
- Generic "sounds good" response

ワークフローへの統合

アラートのしきい値
Score changes trigger actions:

Score hits 40 (WARM):
→ Flag for prioritization
→ Accelerate sequence
→ Add to daily review list

Score hits 70 (HOT):
→ Alert rep immediately
→ Pause automated outreach
→ Recommend live outreach

Score drops below 20:
→ Move to nurture track
→ Reduce frequency
→ Try different approach
応答のルーティング
Based on intent level:

HOT signals:
→ Immediate human follow-up
→ Or: Best bot responses only
→ Personalized, consultative tone

WARM signals:
→ Accelerated automation
→ More CTAs, more urgency
→ Nurture toward decision

COOL signals:
→ Standard sequence
→ Value-focused content
→ Longer intervals

増幅の成果測定

指標
Track:
- Signal detection accuracy
- Conversion rate by signal type
- Response engagement after amplification
- Time from signal to meeting
- Signal-to-close correlation

Compare:
- Amplified responses vs standard
- A/B test amplification approaches
- Rep performance vs bot performance
フィードバックループ
When deal closes:
1. Review signals that were detected
2. Which amplifications worked?
3. What signals were missed?
4. Update detection rules

When deal lost:
1. What signals occurred?
2. Were they amplified properly?
3. What went wrong?
4. Adjust responses

コメント