Business Language Generation Skill: Letting Data Tell Compelling Stories

Data doesn't speak for itself.

The same analysis results, different ways of expression, the effect is vastly different:

Expression A (technical):
"This month's sales 5.2M, QoQ growth 6.5%, YoY decline 2.8%.
Gross margin 32%, QoQ decline 3pp. Customer churn rate 8%, QoQ increase 2pp."

Expression B (storytelling):
"This month overall positive, but hidden concerns.
Good news: Sales back to 5.2M, second highest in past 6 months.
Bad news: Customers are quietly leaving - monthly churned customers
increased from 35 in last quarter to 45.
If no action taken, Q2's revenue foundation will be shaken."

Expression A is data. Expression B is decision.

AskTable's Business Language Generation Skill does one thing: letting data tell compelling stories - not just "what happened", but "what it means, what you should do".

---

I. From Data to Story: Four-Step Transformation

1.1 Step 1: Conclusions First

Business people don't need to see the derivation process, they first want to know "are results good or bad".

❌ Data-style opening:
"This month's sales 5.2M, traffic 1M, conversion rate 5.2%..."

✅ Story-style opening:
"This month overall positive, sales hit second highest in 6 months,
but customer churn acceleration is the biggest hidden concern."

1.2 Step 2: Explain Complex Metrics with Analogies

Technical terminology is a communication barrier, analogies are bridges to understanding.

Technical metric → Analogy expression

"LTV/CAC = 2.8"
→ "For every 1 yuan spent on customer acquisition, we earn back 2.8 yuan"

"Inventory turnover days 38 days"
→ "Your money from becoming inventory to becoming cash, waits an average of 38 days"

"Customer churn rate 8%"
→ "Out of every 100 customers, 8 choose to leave each month"

"NPS 42"
→ "Out of every 10 customers, about 5 are willing to recommend us to friends"

1.3 Step 3: Use Comparisons to Enhance Impact

Isolated numbers have no impact, comparisons create perception.

"Sales 5.2M"
→ Compare to self: "Second highest in past 6 months"
→ Compare to target: "Achieved 83% of monthly target, still 1M short"
→ Compare impact: "Equivalent to revenue from opening 2 new stores"

"45 customers churned"
→ Quantify impact: "These 45 customers contributed ~3M annually"
→ Compare trend: "50% more than same period last year's 30"
→ Visualize: "Equivalent to 1-2 customers who partnered for 2 years leaving monthly"

1.4 Step 4: End with Suggestions Pointing to Action

Good data stories don't end with data, but with action.

❌ Data-style ending:
"Above is this month's data analysis."

✅ Story-style ending:
"To summarize, this month's revenue is positive, but customer churn is accelerating.
Suggestion: Prioritize starting churned customer recovery plan -
for every 10 customers recovered, that's locking in ~60K monthly revenue.
At the same time, maintain new customer acquisition momentum to build foundation for Q2 growth."

---

II. How Business Language Generation Skill Works

2.1 Audience Adaptation

AskTable adjusts expression depth and style based on target audience:

Same data, different audience expressions:

【For CEO】
"This month's sales positive, but customer churn accelerating.
Suggestion: Prioritize retention focus, then chase growth."

【For Business Manager】
"This month's sales 5.2M, QoQ +6.5%, but customer churn rate
rose from 6% to 8%. East China contributed most to churn (40%),
suggestion: Focus on investigating East China customer satisfaction."

【For Frontline Sales】
"This month our team completed 5.2M, second best in 6 months!
Special recognition to East China region Xiao Wang, contributed 350K personally.
Next month's target is 5.5M, let's go!"

2.2 Narrative Structure

AskTable uses classic "Pyramid" narrative structure:

Pyramid tip (1 sentence): Core conclusion
  ↓
Pyramid body (3 findings): Key data + interpretation
  ↓
Pyramid base (2-3 suggestions): Specific actionable next steps

This structure ensures: Understand conclusions in 30 seconds, understand full picture in 3 minutes, 3 suggestions can be acted on immediately.

2.3 Emotional Tuning

Good data stories aren't cold numbers but also mobilize appropriate emotion:

---

III. Typical Application Scenarios

Scenario 1: Reporting to Boss

User asks: "Write me a paragraph to report this month's operations to the boss."

📊 This Month's Operations Report (CEO version)

Boss, this month's operations overall positive, but one signal needs attention:

Good news:
- Sales 5.2M, QoQ +6.5%, second highest in 6 months
- Net profit 850K, exceeded expectations by 12%, cost control effective
- New customer growth 25%, best in past 6 months

Needs attention:
- Customer churn rate rose from 6% to 8%, increasing for 3 consecutive months
- Without intervention, Q2 expected to have ~3M revenue gap

My suggestions:
- Short-term: Start churned customer recovery plan, focus on Top 20 churned customers
- Medium-term: Investigate customer satisfaction, find root causes of churn
- Maintain new customer acquisition momentum, Q2 growth foundation is good

Scenario 2: Compelling Data Presentation

User asks: "Make me a compelling quarterly summary."

📊 Q1 Quarterly Summary

If summarized in one sentence:
"Facing challenges, standing firm in uncertainty."

Three numbers to understand Q1:

📈 14.8M
This is Q1 total sales.
Though 4.4% below target, after removing seasonality
actual trend grew 4.2%.
We're not retreating, we're climbing a steeper hill.

👥 230
This is Q1 new customers, YoY +25%.
Every contract represents customers' trust in us.
This trust is our biggest confidence for Q2.

⚠️ 45
This is Q1 churned customers, 15 more than last year.
Every customer who left is a lesson we need to reflect on.
Q2, we must bring churn rate back down.

Q2's direction is clear:
Hold onto growth, retain customers, turn Q1 momentum into Q2 victory.

Scenario 3: Same Report for Different Audiences

User asks: "Rewrite this analysis report for frontline staff."

📊 This Month's Operations Brief (Team version)

Hi everyone! First, share some good news:

This month we completed 5.2M in sales,
second best performance in 6 months! 🎉

Special recognition:
- East China team: Grew 15% against trend
- Customer service team: 92% satisfaction, highest in company
- Product team: New product broke sales record in first week

Next month's target: 5.5M
Gap isn't large, as long as we maintain this momentum, we can achieve it.

One thing to watch together:
Recently some customers chose to leave, we need to think about how
to do better in service, make customers willing to stay.

Let's go! Q2 together!

---

IV. Hands-On: How to Use Business Language Generation Skill

4.1 Natural Language Trigger

"Write me a paragraph to report to the boss"
"Make a compelling data presentation"
"Rewrite this analysis for the business team"
"Write me a quarterly summary"
"Explain what happened this month in the simplest terms"

4.2 Specify Audience

"Write in language the boss can understand"
"Write a version for frontline sales"
"Write a report for investors"

4.3 Specify Style

"Write more formally, suitable for operations analysis meeting"
"Write more concisely, suitable for WeChat group"
"Write more compelling, suitable for all-hands meeting"

---

V. How Business Language Generation Links with Other Skills

Anomaly detection/attribution/prediction (produce analysis results)
    ↓
Metric interpretation (translate technical metrics)
    ↓
Business language generation (assemble into compelling narrative) ← Core
    ↓
Report orchestration (final output)

Business language generation is the "last mile" of all analysis results - no matter how brilliant the analysis, if expressed poorly, it can't move decision-makers or drive action.

---

VI. Customer Case

A Certain Tech Company: From "Data Reports Nobody Reads" to "Everyone Looks Forward to Weekly Report"

Pain point: Data analysis team sent detailed weekly reports weekly, but open rate was less than 30%. Management feedback was "can't understand, no time to read, don't know what the key points are".

Solution: Enable Business Language Generation Skill, change weekly reports to "conclusions first + 3 findings + 2 suggestions" storytelling structure.

Effects:

• •Weekly report open rate increased from 30% to 85%
• •Management average reading time decreased from 15 minutes to 3 minutes
• •Decision execution rate based on reports increased from 20% to 65%
• •Data analysis team's sense of value significantly improved (because people read and use it)

"Same data, different way of speaking, completely different effect. Before was 'your data team sent another report', now is 'weekly report is here, let's see how this month went'. This isn't wordplay, it's a revolution in communication efficiency." —— CEO, a certain tech company

---

Summary

Business Language Generation Skill's core value isn't in "sounding good", but in:

1. •Conclusions first: Let readers understand core conclusions in 30 seconds
2. •Analogy explanation: Make complex metrics understandable to everyone
3. •Comparison enhancement: Turn isolated numbers into impactful information
4. •Audience adaptation: Speak differently to different people to ensure effective communication
5. •Action orientation: Every data story ends with suggestions to drive decisions

The value of data analysis isn't in the analysis itself, but in making decision-makers understand and take action. Business language generation is the bridge from analysis to action.

---

Extended Reading

• •Metric Interpretation Skill: Translating Data into Business Language - Single metric interpretation
• •Report Orchestration Skill: Auto-assembling Professional Analysis Results - From narrative to complete report
• •Executive Data Assistant Agent: Concise and Efficient Operations Assistant - Typical user of business language generation
• •AskTable Built-in Skills and Agents