Olist

Muito orgulho de pertencer. #GoOlist 💙

Olha o bonde da oportunidade passando:

Nobody tells junior analysts what most of the job actually looks like. The job listings say: SQL, Python, Tableau, data modeling, statistical analysis. The reality for most junior analysts in their first 12 months is something else entirely. Roughly 80% of the work falls into two categories: cleaning data and managing expectations. Data cleaning means tracing why a revenue figure in the dashboard doesn't match the one in the spreadsheet the CFO sent. It means finding that a field called "created_date" actually stores the date a record was last modified. It means writing 40 lines of SQL to handle nulls, duplicates, and inconsistent formatting before a single analysis can begin. Managing expectations means learning to say "that depends on how we define it" before answering any question involving the word "total." It means understanding that stakeholders don't want data -- they want a decision confirmed or a problem explained. None of this is a complaint. It's the part of the job that builds actual instincts. An analyst who has cleaned enough messy data eventually develops a mental model of where data breaks. That's not something any course teaches. The analysts who struggle aren't usually the ones who can't code. They're the ones who expected the job to be more model, less detective work. What surprised you most about your first analytics role? #DataAnalyst #DataAnalytics #DataCareer #SQL #AnalyticsCareers #JuniorAnalyst #DataProfessional

The most common waste in analytics? Building something nobody asked for clearly. It happens constantly. A stakeholder requests a dashboard tracking 40+ KPIs across multiple departments. No clear owner, no decisions tied to any of it. Just "we want visibility." The analyst builds it. Weeks of work. It gets opened a few times, then forgotten. The problem isn't the analyst. It's that vague requests produce vague outputs. A simple 3-question gut check before starting any request changes this: 1. What decision does this enable? If the stakeholder can't name one, the request needs more scoping, not more data. 2. Who's accountable for acting on it? Analysis without an owner dies in a Slack thread. 3. What does "done" look like? Without a definition, there's no way to know when to stop. When a request fails all three, the right response isn't a hard no. It's: "Help me understand the use case so I can build something you'll actually use." That reframe shifts the conversation from friction to collaboration. Most stakeholders appreciate the pushback once they realize it protects their time too. Saying no to bad requests isn't being difficult. It's doing the job well. Have you ever built something that never got used? What was missing from the original ask? #DataAnalyst #DataAnalytics #OpenToWork #StakeholderManagement #AnalyticsLeadership #HiringDataAnalyst #DataCareer