Questions Your Company Can't Answer in Under a Week

Here is a test you can run on your own business this afternoon. Pick a question that sounds like it should take thirty seconds to answer. Not a hard question — a simple one, the kind a curious board member might ask offhandedly. Then try to actually answer it, all the way, with confidence. Watch what happens.

Something like this. "Which of our customer relationships are actually profitable, once you count the work we do for free?" Or: "If the person who runs our most important process left tomorrow, what exactly would stop working — and what only lives in their head?" Or, simplest of all: "Across every part of our business, how many of the thing we sell, manage, or own do we actually have right now?"

These questions sound answerable. That is what makes them dangerous. You will not get an answer in thirty seconds. If you are honest about what "answer" means — verified, current, agreed-upon — you may not get one in a week. You will get a person. That person will open several systems, export a few things, notice the numbers don't line up, ask someone else, wait for a reply, reconcile by hand, and eventually produce something they'll caveat with "roughly." And everyone will accept the caveat, because everyone already knew the real answer was unavailable.

We have quietly agreed to stop asking.

The strange part is that none of these questions are unanswerable in principle. The information exists. Every fact you'd need is sitting in your business right now. The profitability question is just revenue minus the true cost to serve — both of which are recorded somewhere. The key-person question is just a description of a process that someone performs every day. The how-many question is the most concrete fact imaginable; the things are real, they physically or legally exist, someone is responsible for each one.

So why can't you get to it? Because the answer isn't missing. It's scattered. It lives in pieces across systems that were never designed to talk to each other, plus a meaningful and unwritten remainder that lives in specific people's memories. No single place holds the whole picture. To assemble it, a human has to become the integration layer — pulling a number from here, a list from there, a "well, actually it works like this" from a colleague, and stitching them together by hand each time the question is asked. And because it's stitched by hand, it's stale the moment it's finished, and it has to be rebuilt from scratch the next time anyone asks.

This is why the answer is always "give me a few days." The few days aren't spent finding information. They're spent reconciling information that disagrees with itself.

The disagreement isn't an accident or a sign that someone bought the wrong software. It's structural, and it's almost universal. Each tool your business runs on was built to do one job well, and to do that job it had to make its own decisions about what things are. Your accounting system has a definition of "customer." Your operational system has a definition of "customer." The person who closes deals has a third definition in their head.

These definitions overlap enough that nobody notices the gaps — until a question requires all three to line up, and they don't. One system counts a relationship that another system split into two. One counts something the moment it's promised; another, only when it's paid. Neither is wrong. They were never trying to agree.

So your business doesn't have one model of itself. It has as many partial, conflicting models as it has tools and key people — and the work of holding them together has been silently delegated to whoever is willing to do it by hand. Usually that's a small number of long-tenured people who have, without anyone deciding it, become the only place the whole business actually exists.

It's tempting to file all this under "annoying but survivable." Every business has some friction; you route around it; the reports get made eventually. But look at what the workaround actually is. Every question you can't answer quickly is a decision you're making slower, or making blind. Every fact that lives only in someone's head is a dependency on that person remaining available, willing, and accurate. Every reconciliation done by hand is a place where an error can enter and no one will catch it, because there's nothing to check it against.

The business runs, but it runs on a small number of people's memory and patience, and on the hope that no one asks too many hard questions at once. That is not friction. That is structural fragility wearing the costume of normal operations. It stays invisible precisely because it works — right up until the key person leaves, or the question is asked by someone you can't fob off with "roughly," or two of your systems disagree about something that turns out to matter a great deal.

Here is the part worth sitting with. The businesses that can answer these questions in seconds are not smarter than yours, and they don't have better people. They've done one thing differently: they've made the structure of their business explicit. Somewhere there is a single, shared, current model of what exists and how it relates — one place the systems and the people both point to, instead of each holding a private version. When a simple question gets asked, there's somewhere to ask it.

That's the whole difference. Not more tools. Not more data. One agreed-upon picture of the business, kept alive, that doesn't have to be reassembled by hand every time someone is curious.

So run the test. Pick your question. Time the answer. Because the real measure of how well you understand your business isn't your revenue, or your headcount, or how busy everyone is. It's how fast you can answer a simple question about it — and what it costs you that you can't.