Best Data Science Learning Path for MBA Graduates

Phase 1 — Daily Schedule (2-3 hours)

6:30 AM — Read one SQL or stats concept
7:30 AM — 3 SQL practice queries (30 min)
8:30 AM — Statistics reading / notes (30 min)
Evening  — One practical exercise (60 min)
             Week 1-2: SQL on a real schema
             Week 3: EDA on a business dataset
             Week 4: Hypothesis test writeup
             Week 5: A/B test design doc
             Week 6: Pandas report

Daily rules:
- Every query you write, write a business question first
- Every statistic you learn, find a business example
- One note per day: what confused you

What you skip in Phase 1:
✗ Deep learning
✗ Advanced CS theory (compilers, OS internals)
✗ Building a recommender system
✗ Kubernetes, MLOps
✗ Kaggle competitions
Those come later if you actually need them.
Most MBA-track roles never need them.

Week 1 Review:
□ Can I write a GROUP BY query with HAVING?
□ Do I know why COUNT(*) ≠ COUNT(column)?
□ Can I filter and sort confidently?

Week 2 Review:
□ Can I explain the difference between INNER and LEFT JOIN?
□ Can I use ROW_NUMBER() and RANK()?
□ Can I write a query that answers a business question?

Week 3 Review:
□ Do I know when median beats mean?
□ Can I spot an outlier and explain it?
□ Can I describe a distribution in plain English?

Week 4 Review:
□ Can I explain a p-value to a non-technical person?
□ Do I know what statistical significance does not mean?
□ Can I state a null hypothesis for a business question?

Week 5 Review:
□ Do I know how to choose a sample size?
□ Can I explain power and MDE?
□ Do I know the pitfalls of peeking?

Week 6 Review:
□ Can I read a CSV with Pandas?
□ Can I group and summarise without SQL?
□ Can I output a clean CSV report?

Phase 2 — Daily Schedule (3-4 hours)

7:00 AM  — Review yesterday's analysis
8:00 AM  — Deep work (90 min): build the week's project
10:00 AM — Business writeup (30 min): what would you do?
Lunch
2:00 PM  — Refine: charts, tables, README
4:00 PM  — Read one industry case study
Evening  — Polish and commit

Rules for every project:
✓ Start with a business question
✓ Show the "so what" in the first paragraph
✓ Use a real dataset from your industry
✓ Write the recommendation before the model
✓ End with what you'd do next

Project types that land interviews:
✓ A/B test with a real recommendation
✓ Cohort/churn analysis with actions
✓ Pricing / revenue analysis
✓ KPI dashboard for a real business
✗ Titanic / Iris / generic Kaggle (skip these)

Project template (use for all three):

1. Business question
   What decision does this support?
   Who would act on it?

2. Data
   Source and provenance
   Time period
   Known limitations

3. Method
   Why this method over alternatives
   Key assumptions

4. Finding
   The result in one sentence
   Two supporting charts

5. Recommendation
   What should change
   Expected impact (with a range)
   What could make it wrong

6. Next step
   What you'd measure in 30 days

Every project ends with a one-paragraph
"if I were the business owner" statement.
That's the paragraph that gets read.

Resume structure for MBA → data:

Header
  Business Analyst (Data) | SQL · Python · A/B Testing

Summary (3 lines)
  "MBA with 4 years in sales strategy. Built three
   analytics case studies — A/B testing, churn, and
   pricing — that produced specific revenue
   recommendations."

Experience
  Rewrite each bullet to show data work:
  ✗ "Managed regional sales"
  ✓ "Built SQL dashboards tracking regional pipeline;
     identified a churn driver in the SMB segment and
     proposed a retention offer tested with 12k users."

Projects (3 items, named)
  - "Pricing elasticity model — recommended a 6% price
     increase for a specific tier, projected ₹X impact"
  - "A/B test on onboarding copy — 8% lift in activation"
  - "Cohort analysis — identified a drop at week 3"

Skills (be honest)
  SQL (daily use)  ·  Python (Pandas)  ·  Tableau
  A/B testing  ·  Statistics  ·  Stakeholder analysis

For every project, one line of business impact.
That's what a hiring manager at any company reads.

Role-specific prep:

BUSINESS ANALYST
☐ Requirements-gathering stories
☐ Dashboard building
☐ SQL fluency
☐ Stakeholder management examples

PRODUCT ANALYST
☐ A/B test design and pitfalls
☐ Product metrics: DAU, retention, funnel
☐ SQL on user event data
☐ One experiment with real analysis

CONSULTING ANALYST
☐ Case study interviews
☐ Quick turnaround analysis (30 min)
☐ Clean visualisation
☐ Executive summary writing

MARKETING ANALYST
☐ Attribution models
☐ Cohort and funnel analysis
☐ CAC / LTV thinking
☐ Campaign performance SQL

DATA PRODUCT MANAGER
☐ Roadmapping for data products
☐ Metrics definition and governance
☐ Data engineering basics
☐ Stakeholder negotiation

12-week path tracker:

WEEKS 1-2 (SQL)
☐ 20 practice queries
☐ 10 business queries written
☐ Joined data from 2 tables

WEEKS 3-4 (Statistics)
☐ EDA on a real dataset
☐ 3 hypothesis tests interpreted
☐ One p-value explained in plain English

WEEK 5 (A/B Testing)
☐ One A/B test designed
☐ Sample size and power understood
☐ Common pitfalls documented

WEEK 6 (Python)
☐ Pandas report produced
☐ Automated CSV output

WEEKS 7-8 (Projects)
☐ A/B test case study written
☐ Churn analysis with recommendation

WEEKS 9-10 (Projects)
☐ Pricing analysis with recommendation
☐ Dashboard deployed

WEEKS 11-12 (Portfolio)
☐ ML decision frame written
☐ 3 case studies polished
☐ Resume updated with impact lines

Weekly log:

Week _____

Hours spent:
  SQL: ___
  Statistics: ___
  Python: ___
  Projects: ___

Questions I answered this week:
1. ______________________
2. ______________________
3. ______________________

What I could explain to a business owner:
_______________________________________________

What still confuses me:
_______________________________________________

Next week's single priority:
_______________________________________________

One rule: if I can't write the business implication
of what I learned, I haven't learned it.

Skill matrix (rate 1-5 each month):

TECHNICAL
SQL basics              [ ]
SQL joins               [ ]
SQL window functions    [ ]
Python / Pandas         [ ]
Data visualisation      [ ]
Dashboard building      [ ]

STATISTICS
Descriptive stats       [ ]
Hypothesis testing      [ ]
A/B testing             [ ]
Confidence intervals    [ ]

BUSINESS
Problem framing         [ ]
Stakeholder comms       [ ]
Recommendation writing  [ ]
Metric definition       [ ]

Target after 12 weeks: 3+ everywhere.
Anything below 2 is your next priority.
Anything at 4+ is a differentiator — mention it.