Data Analyst: Faire Sens Des Données (Qui Sont Souvent Buggy)

À Propos

On cherche un Data Analyst pour nous aider à comprendre notre business via les données.

Analytics est "super important" pour nous.

(Mais on n'a pas vraiment la stack en place.)

Responsabilités

• Build dashboards (pour que les gens voient les metrics)
• Investigate anomalies (pourquoi les numbers changent?)
• Provide insights (pourquoi la churn monte? pourquoi le revenue baisse?)
• Set up tracking (on a pas de proper event tracking yet)
• Ad-hoc analysis (CEO wants to understand something random)

Profil Recherché

• 2-3 ans analytics experience
• SQL expert (tu vas passer 50% de ton temps en SQL)
• BI tool experience (Tableau, Looker, etc.)
• Python/R is a plus
• Patience with data messiness

Ce Qu'On Veut

Nous: "On est data-driven!"

Reality: Nous n'avons pas d'infrastructure pour être data-driven.

Tu: Tu dois construire l'infrastructure et make insights out of nothing.

La Comp

Salaire: 45k€
Equity: 0.12%
Benefits: Standard startup stuff

Le Jour Typique

• 9h-10h: Fix yesterday's broken dashboard
• 10h-12h: Write SQL to analyze customer behavior
• 12h-13h: Lunch (while debugging a query)
• 13h-15h: CEO asks for a specific analysis ("How many customers from Paris?")
• You write the query
• Data comes back wrong
• You debug
• Find out: The data was never tracked properly
• You tell CEO: "We don't have clean data"
• CEO: "But I need the answer!"
• 15h-17h: Create a workaround to estimate the answer
• 17h-18h: Set up proper tracking for future analysis

The Data Quality Problem

Your company's data is probably:
- Incomplete (missing events)
- Inaccurate (wrong implementations)
- Inconsistent (different definitions across teams)

You'll spend 60% of your time fixing data quality.

30% doing analysis.

10% presenting findings that nobody uses.

The Interesting Contradiction

CEO says: "We're data-driven!"

Reality:
- Decisions are made on gut feeling
- Your analysis is requested but ignored
- The data you provide doesn't inform strategy
- People get defensive when data shows bad news

Example

You find: "Our conversion rate is actually 2%, not 5%"

CEO's response: "Are you SURE the tracking is correct?"

(It is. But he doesn't like the answer.)

After 18 Months

Best Case:
- You've built solid data infrastructure
- Dashboards are actually used
- Your insights informed strategy
- Company is more data-driven

Realistic Case:
- You have some dashboards
- They're partially used
- You've built some tracking
- But culture isn't changing
- You feel like your work doesn't matter

Worst Case:
- Your findings are ignored consistently
- You burn out
- You leave
- Data quality never improves

The Real Challenge

Analytics is only useful if:
1. Data is good (it usually isn't)
2. People act on insights (they usually don't)
3. Culture values data over gut (rare)

Without these, you're making pretty reports that nobody reads.

The Advice

If you like: Data, investigation, building infrastructure:
Maybe yes.

If you want: Your work to directly impact decisions:
No.

Analytics in early-stage startups is often ignored.

The founding team usually knows what they want (wrongly or rightly).

Your data is just... noise.

But someone's gotta build it anyway.