The BI Backlog: The Silent Bottleneck Holding Back Decisions at Brazilian Companies

Last week I sat in on a leadership meeting with the CEO of a retail chain: 200 employees, an ERP that's been running for 6 years, Power BI licenses, a dedicated BI analyst.

Halfway through the meeting, he asked: "Which sales channel brought in the most margin this quarter?"

Silence.

The sales director pulled up WhatsApp and messaged the analyst. The analyst answered 4 hours later with a spreadsheet split across 3 tabs, and none of them answered the exact question. By then the meeting was over, and the decision had already been made on gut instinct.

I've seen this scene play out hundreds of times over the past 20 years, at companies with structured data, paid tools, and a dedicated technical team. The data is right there in the database, clean and up to date. But between the manager's question and the database's answer sits a person. And that person has a backlog.

None of this is that person's fault.

The Analyst Became a Human Translator

The BI analyst is good at the job. But over the years the role got bent out of shape, and the analyst turned into a human translator. On one side is the CEO who wants to know "how the month went." On the other is a database that only speaks SQL. In between, one person bridges the gap by hand, on demand, with 15 tickets in the queue.

That translation costs more than it seems. The 3-to-5-day wait for an ad hoc question is only part of it. The bigger cost is what happens while you wait: decisions made without data, opportunities that slip away, a sales team that course-corrects too late because pipeline visibility showed up on Friday instead of Monday, a CFO who finds a mismatch between the ERP and the controller's numbers on audit day instead of before it.

And the analyst? They spend 80% of their time churning out repetitive reports: pulling, formatting, exporting, emailing. That leaves 20% for the work that actually matters, like strategic analysis, spotting patterns, and making recommendations. The company hired an analyst and got a spreadsheet generator.

Two Decades of Infrastructure, and Nothing for the Last Mile

Brazil spent two decades building data infrastructure: ERPs, data warehouses, dashboards, governance, LGPD (Brazil's data protection law). Layer after layer. But we never solved the last mile, meaning how the person who needs the data can get to it without depending on someone else.

Power BI didn't solve it. It democratized visualization, but not the question. Tableau didn't either, since it was built for people who already know what they want to see. Metabase tried to make things simpler, but it still takes someone technical to build the query. None of them was designed for the CEO who needs a number at 10 a.m. that isn't on any existing dashboard.

The Real BI Bottleneck in 2026

That's the real BI bottleneck in 2026. Companies have plenty of data and plenty of tools. The trouble is that a human queue still stands between the question and the answer, and that queue doesn't scale.

As the business grows, so does the volume of questions. But the analyst is still one person. Maybe two. The queue only gets longer, and the gap between the data and the decision-maker keeps widening, right when it should be shrinking.

At your company, when someone on the leadership team urgently needs a number, does the answer come from the data or from a person?