The interesting thing about broken business reporting is how long it goes unnoticed. The dashboard still loads. Reports still land in inboxes. Monthly meetings still happen. The failure is quieter than that — and more expensive.
A reporting system can be technically functional and operationally useless at the same time. The numbers get produced. Nobody acts on them. The business keeps making decisions on instinct, habit, and incomplete information — while generating the appearance of data-driven management.
None of the signs below require a technical audit to identify. All of them are observable this week, in your next team meeting, or in the last three months of management accounts sitting in your inbox.
Sign 1: The Last Three Reports Didn't Drive a Decision
Pull up the last three monthly reports. For each one, ask: what decision did someone make differently because of this? Not "we discussed it." Not "it was useful context." What actually changed?
If the answer is "none" across three consecutive reporting cycles, you don't have a BI problem — you have a reporting-theatre problem. Reports are being produced, meetings are happening, but the outputs aren't changing behaviour.
This is the most reliable indicator of broken reporting. The purpose of management information is to drive decisions. If it isn't, it's an expensive administrative exercise.
The fix isn't always more data or better visualisations. It's often simpler: identify one decision that the data should be driving, and restructure that section of the reporting to answer that question explicitly.
Sign 2: The Same Question Gets Asked Every Monthly Meeting
Every business has a version of this. The leadership team sits down with the monthly report in front of them, and within five minutes someone asks the same question they asked last month: "But what's our margin by product line, actually?" Or: "Where is the cash going — the P&L looks fine but the balance is lower." Or: "Which customer is driving the revenue growth?"
These questions recur because the reporting doesn't answer them. The monthly pack covers the standard set — P&L, pipeline, headline KPIs — but the questions the business actually needs to answer aren't in it.
Recurring questions are a direct map of the gaps in your reporting. Write them down. If the same question has been asked in three or more consecutive meetings without being resolved by the reporting, it belongs in the dashboard. Not in the meeting.
Sign 3: Someone Maintains a Parallel Spreadsheet
This is the most visible sign of lost trust in a reporting system. Somewhere in your business — usually Finance, Sales, or Operations — someone is maintaining their own version of a key metric. They pull the raw data themselves, build their own model, and bring that number to the meeting. Not the official one.
They do this because they've been burned before. The official number was wrong. Or defined differently than they expected. Or updated at the wrong time. So they built something they trust.
The problem isn't the spreadsheet. It's what the spreadsheet represents: a parallel source of truth. Now the business has two revenue figures, two customer counts, two views of margin — and a meeting that starts with ten minutes of reconciliation before any decisions can be made.
Every shadow spreadsheet is a vote of no-confidence in the official reporting. Find the shadows. They tell you where the trust broke down.
The root cause is almost always a data definition problem — the official metric was never agreed. Fixing it requires one conversation: what exactly does this number mean, and which system is the authoritative source?
Sign 4: Decisions Keep Getting Deferred "Pending More Data"
BI that works produces confidence. BI that's broken produces the opposite: a permanent sense that you almost have enough information to decide, but not quite.
The deferral pattern looks like this: a decision lands on the table. Someone asks for more data before committing. The data arrives. It raises new questions. Those questions get sent back. The decision gets pushed to next month. By the time a decision is made, the context has shifted.
In some cases, more data genuinely is the answer. But if this pattern is recurring across different decisions and different parts of the business, the issue isn't a missing dataset — it's a reporting system that doesn't produce decision-grade confidence.
Decision-grade BI has three properties: the data is trusted, it's presented at the right level of aggregation, and it's explicitly connected to a decision. If any of those three are missing, the request for "more data" is a proxy for "we don't trust what we have."
Sign 5: Nobody Can Name Who Owns Each Metric
Pick any three metrics in your current dashboard. For each one, ask: who is accountable for this number? Who explains it when it moves? Who is responsible for the decisions it's meant to drive?
If the answer is "everyone" or "the finance team" or "it's shared" — you have a metric without an owner. And a metric without an owner is a metric without accountability. It gets reported, discussed, and filed. Nothing changes.
Ownership isn't about blame. It's about closing the loop between data and action. The revenue retention number needs someone who reviews it every month, explains the trend, and is accountable for what happens next. Without that person, the metric is decoration.
The ownership question is the governance layer of BI. It's the part most businesses skip because it requires a conversation about accountability that feels uncomfortable. But it's the part that determines whether any of the reporting actually changes anything.
What All Five Signs Have in Common
None of these failures are technical. The dashboard isn't broken. The data isn't necessarily wrong. The failure is in the connection between data and action — the governance and decision-design layer that most BI implementations skip entirely.
Building that layer doesn't require new software. It requires answering four questions for every metric in your current stack: What decision does this drive? Who owns the answer? When was the last time it changed behaviour? What happens if it disappears?
Any metric that can't answer those four questions is a candidate for retirement. The ones that can are the foundations of a BI system that actually works.