Executive Leadership Journey: AI-Driven Strategic Decisions

David: What data sources do we have for customer analysis?

Agent: I found 5 active data sources relevant to customer analysis in your
       organization's data catalog:

  ┌─────────────────────────────────────────────────────────────────────┐
  │  Source                Status    Last Sync       Tables  Quality   │
  ├─────────────────────────────────────────────────────────────────────┤
  │  NovaMart PostgreSQL   Active    12 min ago      14      87%       │
  │  Shopify Orders API    Active    8 min ago       6       92%       │
  │  Snowflake DWH         Active    47 min ago      23      89%       │
  │  Google Analytics      Active    6 hours ago     4       95%       │
  │  Customer Surveys      Active    2 days ago      1       94%       │
  └─────────────────────────────────────────────────────────────────────┘

  Key customer-related tables:
  - customers (2.1M rows) -- profiles, segments, lifetime value
  - orders (12.4M rows) -- full transaction history
  - clickstream (340M rows) -- web and app behavior
  - customer_nps (48K rows) -- satisfaction surveys
  - customer_segments (2.1M rows) -- ML-derived segments

  All sources are syncing on schedule. The customer surveys data was last
  updated 2 days ago (manual CSV import -- not auto-synced).

David: Show me all tables related to customer lifetime value.

Agent: I found 6 tables with columns or metadata related to customer
       lifetime value (CLTV):

  1. customers (novamart.public)
     - Column: lifetime_value (DECIMAL)
     - 2.1M rows, updated 12 min ago
     - Quality score: 87% (12.1% null emails)

  2. customer_segments (analytics.derived)
     - Columns: predicted_cltv, cltv_percentile
     - 2.1M rows, updated 1 day ago
     - Quality score: 91%
     - Owner: priya@novamart.com

  3. churn_features (analytics.derived)
     - Column: is_churned (binary label)
     - 1.98M rows, updated 8 hours ago
     - Quality score: 97%
     - Owner: priya@novamart.com

  4. cohort_revenue (analytics.facts)
     - Columns: cohort_month, revenue_month_1 through revenue_month_12
     - 24 cohorts, updated daily
     - Quality score: 99%
     - Owner: sofia@novamart.com

  5. revenue_hourly (analytics.facts)
     - Used to compute CLTV actuals
     - 2.8M rows, updated 15 min ago
     - Quality score: 98%

  6. customer_nps (novamart.surveys)
     - NPS correlates with CLTV retention
     - 48K rows, updated 2 days ago
     - Quality score: 94%

  The primary CLTV metric is defined in the semantic layer by Sofia's team.
  Current average CLTV: $412 (target: $500). Priya's churn model feeds into
  the CLTV prediction.

David: What was our revenue by channel last quarter compared to the
       same quarter last year?

Agent: Here are the results, using the Net Revenue metric from the
       semantic layer:

  Revenue by Channel: Q4 2025 vs Q4 2024
  ────────────────────────────────────────────────────────────────
  Channel     Q4 2024        Q4 2025        Growth      Share
  ────────────────────────────────────────────────────────────────
  Web         $25,812,340    $31,489,217    +22.0%      58.3%
  Mobile      $12,147,891    $16,234,782    +33.6%      30.1%
  Store       $ 6,234,112    $ 6,278,943    + 0.7%      11.6%
  ────────────────────────────────────────────────────────────────
  Total       $44,194,343    $54,002,942    +22.2%     100.0%

  Key observations:
  - Mobile grew 33.6% YoY -- now 30% of revenue (was 27.5%)
  - Store revenue is essentially flat (+0.7%) -- growing below inflation
  - Web remains dominant but mobile is closing the gap

  Generated SQL:
    SELECT
        channel,
        SUM(CASE WHEN order_date BETWEEN DATE '2024-10-01' AND DATE '2024-12-31'
                 THEN net_revenue END) AS q4_2024,
        SUM(CASE WHEN order_date BETWEEN DATE '2025-10-01' AND DATE '2025-12-31'
                 THEN net_revenue END) AS q4_2025,
        ROUND(100.0 * (SUM(CASE WHEN order_date BETWEEN DATE '2025-10-01' AND DATE '2025-12-31'
                                THEN net_revenue END)
              / NULLIF(SUM(CASE WHEN order_date BETWEEN DATE '2024-10-01' AND DATE '2024-12-31'
                                    THEN net_revenue END), 0) - 1), 1) AS yoy_growth_pct
    FROM semantic.revenue_by_channel
    GROUP BY channel
    ORDER BY q4_2025 DESC;

David: Why is mobile growing so much faster than web? Break down the
       mobile growth by customer segment.

Agent: Mobile revenue growth decomposed by customer segment:

  Mobile Revenue Growth Decomposition: Q4 2025 vs Q4 2024
  ────────────────────────────────────────────────────────
  Segment     Q4 2024     Q4 2025     Growth   Contribution
  ────────────────────────────────────────────────────────
  New         $2,891,342  $5,124,891  +77.3%   54.7%
  Silver      $4,234,112  $5,012,347  +18.4%   19.1%
  Gold        $3,891,234  $4,534,891  +16.5%   15.8%
  Bronze      $1,131,203  $1,562,653  +38.2%   10.6%
  ────────────────────────────────────────────────────────

  Root cause: New customer acquisition on mobile drove 55% of the channel's
  growth. This correlates with the mobile app relaunch in August 2025 and
  the $2.1M increase in mobile-targeted ad spend.

  Supporting data:
  - Mobile app downloads: +142% QoQ (from app store data)
  - Mobile conversion rate: 2.8% -> 3.4% (from GA data)
  - Mobile-specific campaigns: 12 new campaigns in Q4

David: When was our customer data last updated? I have a board
       meeting tomorrow and need current numbers.

Agent: Here is the freshness status for all customer-related data:

  Data Freshness Report (as of Dec 15, 2025 2:34 PM)
  ──────────────────────────────────────────────────────
  Dataset                    Last Updated    Status
  ──────────────────────────────────────────────────────
  customers                  14 min ago      Current
  orders                     8 min ago       Current
  clickstream                52 min ago      Current
  revenue_hourly (fact)      18 min ago      Current
  revenue_daily (fact)       6 hours ago     Current (next: 3 PM)
  customer_segments          1 day ago       Current (weekly refresh)
  churn_features             8 hours ago     Current (daily refresh)
  customer_nps               2 days ago      Stale (manual import)
  ──────────────────────────────────────────────────────

  All automated pipelines are running on schedule. The customer_nps
  data is 2 days old because it requires a manual CSV upload.

  Recommendation: If you need updated NPS data for the board meeting,
  ask the CX team to export the latest survey results for import.
  All other metrics are current and ready for reporting.

David: I need a weekly summary of our top 10 KPIs sent to the
       leadership team every Monday morning. Can you set that up?

Agent: I can draft the report configuration for you. Here is what I
       recommend based on the KPIs defined in the semantic layer:

  Proposed Weekly KPI Report
  ──────────────────────────
  Metrics: GMV, Net Revenue, AOV, Conversion Rate, Churn Rate,
           CLTV, CAC, ROAS, Inventory Turnover, NPS
  Comparison: Week-over-week and year-over-year
  Format: PDF with embedded charts
  Schedule: Every Monday at 7:00 AM
  Recipients: david@, cfo@, ceo@, vp-marketing@, vp-ops@

  Shall I submit this as a report creation request to the BI team?
  Sofia (BI Lead) can set this up in the BI Workbench and have it
  running by next Monday.

David: Yes, submit it.

Agent: Done. I have created a request ticket and notified Sofia.
       Request ID: RPT-2025-0847. You will receive a confirmation
       once the report is configured and scheduled.

David: Which product categories are underperforming this quarter?

Agent: I analyzed revenue trends for all 20 product categories
       comparing Q4 2025 to Q3 2025 and Q4 2024. Three categories
       show declining revenue:

  Underperforming Categories: Q4 2025
  ─────────────────────────────────────────────────────────────
  Category          Q3 2025     Q4 2025     QoQ      YoY
  ─────────────────────────────────────────────────────────────
  Books & Media     $1.82M      $1.54M     -15.4%   -8.2%
  Kitchen Appl.     $2.41M      $2.18M      -9.5%   -3.1%
  Office Supplies   $1.13M      $0.97M     -14.2%  -11.7%
  ─────────────────────────────────────────────────────────────

  Root Cause Hypotheses:

  1. Books & Media (-15.4% QoQ)
     - Competitor analysis: Amazon launched same-day book delivery
       in October in 8 of our top markets
     - Price comparison: Our avg book price is $18.42 vs Amazon $14.99
     - Customer feedback: 3 NPS comments mention "cheaper on Amazon"

  2. Kitchen Appliances (-9.5% QoQ)
     - Seasonal effect: Q3 includes back-to-school / new apartment
       purchases. Q4 drop may be partially seasonal.
     - YoY is only -3.1%, suggesting this is not structural
     - Inventory data: 4 top SKUs were out of stock for 8+ days
       in November (supplier delays)

  3. Office Supplies (-14.2% QoQ)
     - Structural trend: 3 consecutive quarters of decline
     - Likely cause: Remote work normalization reducing office supply
       demand. This segment may need strategic review.

  Data quality note: These figures use the net_revenue metric from the
  semantic layer. All underlying data passed quality checks within the
  last 24 hours. Revenue figures are accurate to +/- $0.01.

  Shall I drill deeper into any of these categories?

Daily Revenue Snapshot -- Dec 15, 2025
──────────────────────────────────────
Yesterday's Performance:
  GMV:          $487,234  (+12% vs last Monday)
  Net Revenue:  $421,891  (+8% vs last Monday)
  Orders:       7,234     (+15% vs last Monday)
  AOV:          $67.40    (-3% vs last Monday)
  Churn Rate:   17.8%     (down from 18.3% last month)

Alerts: None triggered yesterday.
Full dashboard: [View in BI Workbench]

David: Prepare a one-page summary of our Q4 performance for the
       board presentation. Focus on revenue growth, customer health,
       and strategic risks.

Agent: Here is a board-ready summary:

  ═══════════════════════════════════════════════════════════════
  NOVAMART Q4 2025 PERFORMANCE SUMMARY
  Prepared: December 15, 2025
  ═══════════════════════════════════════════════════════════════

  REVENUE
  - Q4 GMV: $54.0M (+22.2% YoY)
  - Q4 Net Revenue: $46.8M (+19.4% YoY)
  - December run rate: on track for $22.9M monthly GMV (record)

  CUSTOMER HEALTH
  - Active customers: 2.1M (+14% YoY)
  - Churn rate: 17.8% (down from 18.3%, target: 15%)
  - Churn prediction model (deployed Nov) identifies at-risk
    customers 23% more accurately, saving est. $47.7K/month
  - NPS: 42 (stable, target: 50)

  GROWTH DRIVERS
  - Mobile channel: 30.1% of revenue (was 27.5%), +33.6% YoY
  - New customer acquisition: +77% on mobile (app relaunch effect)
  - Conversion rate: 3.1% (up from 2.8%)

  STRATEGIC RISKS
  - In-store revenue flat (+0.7% YoY) -- needs strategic review
  - 3 product categories declining (Books, Kitchen, Office)
  - Inventory stockouts cost est. $340K in lost November sales
  - CAC trending up: $34.20 (target: $30)

  RECOMMENDATIONS
  1. Increase mobile investment -- highest ROI channel
  2. Review in-store strategy (close, convert, or differentiate?)
  3. Address inventory stockouts -- demand forecasting system live
  4. Reduce CAC via organic/referral programs
  ═══════════════════════════════════════════════════════════════

  This summary was generated from live platform data. All metrics
  verified against the semantic layer definitions. Shall I export
  this as a PDF or PowerPoint slide?

Weekly KPI Digest -- Week of Dec 8-14, 2025
────────────────────────────────────────────────────────────
KPI                 This Week    Last Week    Target    Status
────────────────────────────────────────────────────────────
GMV                 $3.41M       $3.12M       $4.5M/wk  Below
Net Revenue         $2.95M       $2.69M       $3.9M/wk  Below
AOV                 $67.40       $69.10       $75.00     Below
Orders              50,578       45,234       60K/wk    Below
Conversion Rate     3.1%         3.0%         4.0%      Below
Churn Rate          17.8%        18.1%        <15%      Above
CLTV                $412         $409         $500      Below
CAC                 $34.20       $33.80       <$30      Above
ROAS                4.8x         5.1x         6.0x      Below
NPS                 42           42           50+       Below
────────────────────────────────────────────────────────────
Overall: 0/10 KPIs at target. 3 improving, 2 stable, 5 declining.

David: Churn rate dropped from 18.1% to 17.8% this week. Is that
       because of Priya's model, or something else?

Agent: Good question. I analyzed the churn rate change by decomposing
       it into model-influenced and organic components:

  Churn Rate Decomposition: Week of Dec 8-14
  ───────────────────────────────────────────
  Component                  Impact on Churn Rate
  ───────────────────────────────────────────
  Model-driven retention     -0.18%
    (customers scored high-risk and targeted
     with retention campaigns: 231 retained)

  Organic improvement        -0.12%
    (holiday shopping brought back 1,847
     customers who would have been classified
     as churned next week)

  Net change                 -0.30%
  ───────────────────────────────────────────

  Roughly 60% of the improvement is attributable to the churn model's
  retention campaigns. The remaining 40% is seasonal -- holiday shopping
  temporarily reactivates dormant customers. We may see churn increase
  again in January when the holiday effect fades.

  Recommendation: Do not extrapolate this trend. January will be the
  true test of whether the model-driven improvement is sustainable.

David: What if we increase marketing spend by 20% next quarter?
       What is the projected revenue impact?

Agent: I modeled this scenario using historical marketing elasticity
       data from the last 8 quarters:

  Scenario: +20% Marketing Spend (Q1 2026)
  ─────────────────────────────────────────────────────────────
                     Current Plan    +20% Scenario    Delta
  ─────────────────────────────────────────────────────────────
  Marketing Budget   $4.2M           $5.04M           +$840K
  Projected GMV      $48.1M          $52.3M           +$4.2M
  Projected Net Rev  $41.4M          $44.8M           +$3.4M
  New Customers      24,500          31,200           +6,700
  CAC                $34.20          $32.30           -$1.90
  ROAS               4.8x            5.0x             +0.2x
  ─────────────────────────────────────────────────────────────

  Assumptions:
  - Marketing elasticity: 0.52 (1% spend increase -> 0.52% revenue
    increase, based on 8-quarter regression, R-squared = 0.78)
  - Diminishing returns kick in above +30% spend increase
  - Channel allocation remains constant (60% digital, 25% social,
    15% traditional)

  Sensitivity Analysis:
  ┌─────────────────────────────────────────────┐
  │ Spend Increase:  +10%    +20%    +30%       │
  │ Revenue Lift:    +$2.0M  +$4.2M  +$5.8M    │
  │ Marginal ROAS:   5.1x    5.0x    4.4x       │
  │ CAC:             $33.10  $32.30  $33.80     │
  └─────────────────────────────────────────────┘

  Note: Marginal ROAS declines past +20%, suggesting that is near
  the optimal spend increase for current channel efficiency. Beyond
  +30%, CAC starts rising again due to audience saturation.

  Confidence level: Medium. This model uses historical correlations
  which may not hold if competitive dynamics change.

David: Compare two options for me:
       Option A: Invest $1M in mobile app features
       Option B: Invest $1M in opening 3 new retail stores

Agent: Here is a side-by-side analysis using historical performance
       data and market benchmarks:

  Strategic Investment Comparison
  ════════════════════════════════════════════════════════════════
                          Option A              Option B
                          Mobile App             3 New Stores
  ────────────────────────────────────────────────────────────────
  Investment              $1.0M                  $1.0M
  Timeline                3 months dev           9-12 months build
  Expected Rev Lift       $4.8M/yr               $2.1M/yr
  Payback Period          2.5 months             5.7 months
  Incremental Customers   18,000/yr              4,200/yr
  Marginal CAC            $14.80                 $48.20
  Risk Level              Low                    High
  ────────────────────────────────────────────────────────────────

  Supporting Evidence:

  Option A (Mobile App):
  - Mobile revenue growing 33.6% YoY (fastest channel)
  - Mobile conversion rate improved 21% after last app update
  - 142% increase in app downloads after August relaunch
  - Benchmark: Similar e-commerce companies see 3-5x ROI
    on mobile UX investment

  Option B (New Stores):
  - Existing stores growing 0.7% YoY (slowest channel)
  - Average store generates $500K-$700K annual revenue
  - Build-out cost: $250K-$350K per location + lease
  - Industry trend: Store traffic declining 3% annually
  - Risk: Lease commitments of 3-5 years

  Recommendation: Option A (Mobile App) delivers 2.3x the revenue
  lift at lower risk with a faster payback period. Option B has
  strategic value only if stores serve as fulfillment hubs for
  same-day delivery -- which would require an additional $400K
  investment in logistics infrastructure.

  Caveat: This analysis assumes current trends continue. A major
  shift in consumer behavior (e.g., return to in-person shopping)
  would change the calculus. Recommend revisiting quarterly.
  ════════════════════════════════════════════════════════════════

  Technical Layer (invisible to David)      Executive Layer (David's experience)
  ─────────────────────────────────────     ─────────────────────────────────────
  Airbyte connectors, CDC, sync jobs   -->  "Our data was updated 12 minutes ago"
  Trino federation, SQL optimization   -->  "Revenue by channel last quarter..."
  Great Expectations quality checks    -->  "Data quality score: 98%"
  MLflow experiments, Ray Serve        -->  "Churn model saves $47.7K/month"
  Temporal pipelines, DAG scheduling   -->  "All pipelines running on schedule"
  Semantic layer metric definitions    -->  "Net Revenue = GMV - refunds - discounts"