Concept
1. What makes people stop
Lead with a short, useful opening; the visible preview varies by device and layout. Write them last, after you know your best insight.
Good hooks:
- A surprising number: "Our 'attrition rate' looked like 26.8%. The real number was 5.2%."
- A before/after: "This report took 3 hours every month. Now it takes 2 minutes."
- A question: "Why do mobile users convert 35% less than desktop?"
Weak hooks: "Excited to share my new project!", "Day 23 of learning Excel."
2. The post structure
[Hook — 1–2 short lines]
[Context — the problem in 2 lines]
What I did:
→ Cleaned a 50,000-row synthetic sales export with Power Query
→ Built a data model with YoY and target measures
→ Designed a one-page dashboard with slicers
What the data showed:
1. Revenue +11.0% YoY, but only 96.6% of target
2. Growth came from more orders, not bigger baskets
3. West grew slowest (+6.3%) — the main gap
[Lesson / takeaway in 1 line]
[Question to readers]
#Excel #DataAnalytics #PowerQuery
Short paragraphs, white space, arrows or numbers — easy to read on a phone.
3. The visual
Choose an image, document or video that makes your project easy to inspect. Compare performance using your own analytics.
- Single image: the dashboard, cropped tight, readable on a phone (big fonts!).
- Carousel (upload a PDF): 5–7 slides — 1 hook, 2 problem/before, 3 approach, 4–5 insights (one per slide), 6 dashboard, 7 "what I learned + link in comments". Use 1080 × 1350 px (portrait) slides with large text.
- Short screen recording of slicers changing the dashboard.
4. Links
Put the GitHub link where readers can find it. Test post text and a comment if useful; neither placement guarantees greater reach.
5. After posting
- Reply to every comment in the first hour or two — answering questions helps readers understand your work.
- Ask a real question ("How would you measure attrition?") — it invites expert comments.
- Tag only people genuinely involved (mentor, course) — not 20 random influencers.
- 3–5 relevant hashtags are enough.
6. When and how often
Try a time when your intended readers are available, then compare your post analytics. One good project post every 1–2 weeks beats daily low-effort posts. Mix: project posts, short tips ("3 Power Query steps that saved me hours"), lessons from mistakes.
7. Example — HR attrition project
Our attrition looked like 26.8%.
The real number was 5.2%. Here's the difference 👇
26.8% = everyone who left since 2019 ÷ everyone ever hired.
That's cumulative — not an attrition rate.
Annual attrition = exits in the year ÷ average headcount
→ 25 ÷ 477.5 = 5.2% for FY2025-26.
Other findings from 650 employee records (Excel + Power Query):
1. Overtime employees left at 37.7% vs 22.2%
2. Customer Support & Sales had the highest exit share
3. "Better pay" was the top exit reason (64 of 174)
Lesson: define the metric before you calculate it.
How does your company measure attrition?
Dataset & workbook in the comments.
#Excel #HRAnalytics #DataAnalytics
8. Profile basics that convert views into calls
Headline with role + skills ("MIS Analyst | Excel · Power Query · Dashboards"), a Featured section with your best project post and GitHub, and "Open to work" settings for recruiters if you're job hunting.
Example scope: The posts use synthetic practice data. State this in your post and distinguish cumulative exit shares from annual attrition. LinkedIn post analytics.
Common mistakes
Generic hooks. Walls of text. Tiny unreadable screenshots. No numbers. Posting and disappearing. Hashtag spam.