Presenting a project: README, screenshots, before/after

Concept

1. Recruiters skim

Most reviewers spend about a minute on a project. In that minute they should see: the question, one strong image, the result in numbers, and the tools used. Everything else is for the curious.

2. Project folder

hr-attrition-analysis/
    README.md                 ← the project page
    HR_Attrition_Analysis.xlsx
    HR_Attrition_Dashboard.pdf
    data/hr_attrition.csv
    images/dashboard.png
    images/before_after.png
    images/insights.png

File names without spaces, dated if versions matter. Remove personal/confidential data.

3. README template

# HR Attrition Analysis (Excel)

**Question:** Who is leaving the company, why, and what should HR change?

![Dashboard](images/dashboard.png)

## Key insights
- Employees with overtime leave at 37.7% vs 22.2% without overtime.
- Customer Support (34.5%) and Sales (32.2%) have the highest exit share.
- "Better pay" is the top exit reason (64 of 174 exits).
- FY2025-26 annual attrition is 5.2% (25 exits ÷ avg. headcount 477.5).

## Recommendations
1. Cap overtime in Support and Sales; review shift planning.
2. Benchmark salaries for roles with the most "Better pay" exits.

## What I did
- Cleaned 650 employee records in Power Query (dates, blanks, categories)
- Built age bands, tenure, and active/exited flags
- Pivots + dashboard with Department and City slicers

## Tools
Excel · Power Query · Pivot tables · Dynamic arrays · Charts

## Files
- `HR_Attrition_Analysis.xlsx` — workbook (README sheet inside)
- `HR_Attrition_Dashboard.pdf` — one-page view
- `data/hr_attrition.csv` — synthetic dataset

4. Screenshots that work

One hero image: the whole dashboard at 100% zoom, gridlines off, filters cleared, no ribbon (Win + Shift + S on Windows, Cmd + Shift + 4 on Mac).
Detail images: 1–2 close-ups (a KPI card row, a key chart) with a short caption.
Add arrows/boxes only to point at the insight; keep them in your accent colour.
A 10–15 second GIF of slicers in action (ScreenToGif on Windows) shows interactivity better than words.
Save as PNG; keep each under ~1 MB.

5. The before/after story

Put the raw mess next to the result:

Before	After
24 monthly CSVs, inconsistent dates, trailing spaces	one Power Query refresh, 50,000 clean rows
report built by hand in ~3 hours	refresh in 2 minutes
no target view	KPI cards with ▲▼ vs last year and target

Numbers make it believable: time saved, rows handled, errors removed.

6. Inside the workbook

A README sheet (Dashboards track, Module 8): purpose, data, steps, KPI definitions.
Clean layer structure: DASH, CALC, RAW.
DASH active on open, filters cleared.
A one-page PDF export for people who won't open Excel.

7. Be honest

Say the data is synthetic or public. Don't present course projects as client work. Credit sources. Interviewers will ask "why did you…?" — make sure every step is yours to explain.

Common mistakes

No README or a README that only lists tools. Blurry screenshots with gridlines and the ribbon. No numbers in insights. Confidential employer data in a public portfolio.

What you'll learn