// week 02 · day 08 · wed

Interpret Metrics in Context.

$/data-analysis · ~25 min
// 01 · what it does

Numbers with context.

data-analysis classifies the analysis type (metric interpretation, funnel analysis, cohort analysis, or anomaly investigation), produces verified insights with cited calculations, and generates 1–3 programmatic visualizations. Every numeric claim references a calculation ID that traces back to a reproducibility bundle — a replayable Python script, raw inputs, derived outputs, and a verification log. The math is checked before the output is delivered, not assumed.

The problem with product data isn't access — most teams have dashboards. The problem is interpretation: a number without context is noise, and context without a number is speculation. data-analysis connects them. It also surfaces what the data can't tell you — limitations, confidence intervals, and the hypotheses the data supports without proving. That distinction matters more than the findings themselves.

Day 8 adds the quantitative layer to the evidence week. Day 7 told you what users said; Day 8 tells you what the numbers show. Together they make a defensible claim. The analysis you produce today will become the evidence base for the business case on Day 10.

Setup note: This skill requires matplotlib. Run python3 -m pip install matplotlib before this day if you haven't already. See the setup page.
// 02 · sample prompts

Two ways in.

prompt.basic.txt 
/data-analysis

Here's a simple data table. Tell me what's happening in the booking funnel:

listing_type,search_sessions,detail_views,booking_starts,completed_bookings,cancellations
request_to_book,1200,408,53,46,7
instant_book,300,111,19,17,2

What is the Instant Book conversion lift compared to request-to-book? Show both search-to-booking and detail-to-booking rates, and label the denominator for each.
prompt.advanced.txt 
/data-analysis

I have two tables from Fernando Lopez (Data Lead). He exported these from Amplitude + Looker last Tuesday. Please analyze both.

Questions I need answered:
1. What is the Instant Book conversion lift relative to request-to-book on both search-to-booking and detail-to-booking? Label the denominator for each, and note whether the lift is consistent across the 16-week period or changing over time.
2. There appears to be a cancellation anomaly somewhere in the weekly data. Identify which weeks it occurs, estimate the magnitude, and surface any hypotheses about cause.
3. Which experience category has the highest cancellation rate — and does Instant Book reduce cancellations consistently across all categories or only some?
4. Are there any data quality issues I should flag before taking this to a leadership deck?

Please generate visualizations where they add clarity. Cite all calculations.

---

TABLE 1: Weekly Booking Funnel by Listing Type (16 weeks)

week,listing_type,search_sessions,detail_views,booking_starts,completed_bookings,cancellations
1,request_to_book,5200,1768,229,198,27
1,guide_pro_request_to_book,1600,624,90,79,11
2,request_to_book,5350,1819,236,204,28
2,guide_pro_request_to_book,1650,644,93,82,11
3,request_to_book,5500,1870,243,210,29
3,guide_pro_request_to_book,1680,655,95,84,11
4,request_to_book,5600,1904,247,214,30
4,guide_pro_request_to_book,1720,671,97,85,12
5,request_to_book,5750,1955,254,220,30
5,guide_pro_request_to_book,1750,683,99,87,12
6,request_to_book,5900,2006,261,226,31
6,guide_pro_request_to_book,1800,702,102,90,12
7,request_to_book,6050,2057,267,231,32
7,guide_pro_request_to_book,1850,722,105,92,13
8,request_to_book,6200,2108,274,238,33
8,guide_pro_request_to_book,1900,741,107,94,13
9,request_to_book,6000,2040,265,230,32
9,guide_pro_request_to_book,1900,741,107,94,13
9,instant_book,750,278,47,42,4
9,guide_pro_instant_book,270,111,20,18,2
10,request_to_book,5900,2006,261,226,31
10,guide_pro_request_to_book,1880,733,106,93,13
10,instant_book,1100,407,69,62,6
10,guide_pro_instant_book,390,160,29,26,3
11,request_to_book,5800,1972,256,211,37
11,guide_pro_request_to_book,1870,729,106,87,14
11,instant_book,1350,500,85,70,14
11,guide_pro_instant_book,480,197,35,29,6
12,request_to_book,5700,1938,252,208,37
12,guide_pro_request_to_book,1860,725,105,86,15
12,instant_book,1500,555,94,78,14
12,guide_pro_instant_book,530,217,39,32,6
13,request_to_book,5600,1904,247,215,29
13,guide_pro_request_to_book,1850,722,104,92,12
13,instant_book,1800,666,113,102,10
13,guide_pro_instant_book,620,254,46,41,4
14,request_to_book,5400,1836,238,207,28
14,guide_pro_request_to_book,1830,714,103,91,12
14,instant_book,2100,777,132,119,12
14,guide_pro_instant_book,730,299,54,49,5
15,request_to_book,5200,1768,230,200,27
15,guide_pro_request_to_book,1800,702,102,90,12
15,instant_book,2500,925,157,141,15
15,guide_pro_instant_book,880,361,65,58,6
16,request_to_book,5000,1700,221,192,26
16,guide_pro_request_to_book,1780,694,100,88,12
16,instant_book,2900,1073,182,163,18
16,guide_pro_instant_book,1040,426,77,69,7

---

TABLE 2: Booking Funnel by Experience Category (weeks 9–16, beta period only)

category,listing_type,detail_views,booking_starts,completed_bookings,cancellations,avg_booking_value
surf_lessons,request_to_book,4800,682,559,123,95
surf_lessons,instant_book,1800,306,266,35,95
alpine_climbing,request_to_book,2400,312,287,26,420
alpine_climbing,instant_book,900,135,127,8,420
kayaking_tours,request_to_book,3800,494,432,60,140
kayaking_tours,instant_book,1400,238,214,22,140
foraging_walks,request_to_book,1600,192,175,17,75
foraging_walks,instant_book,600,90,84,6,75
fly_fishing,request_to_book,2200,286,243,43,220
fly_fishing,instant_book,750,120,107,13,220
// 03 · reflection

Three questions.

  1. 01Which finding did the skill surface that you would have missed scrolling through a dashboard — and how would that finding change a decision you're currently facing?
  2. 02Where did the skill note a data quality limitation — and would that limitation matter enough to delay a leadership presentation?
  3. 03How would you use this analysis as evidence in the business case you'll build on Day 10 — what claim does it support, and what would make that claim stronger?
← Day 7 Day 9 →