5 Signs Your Numbers Are Guessing (What an Analyst Sees First)

When I open a business's numbers for the first time, I'm not hunting for typos. I'm looking for one thing: whether these numbers can be trusted to make a decision. There's a real difference between the two, and after 25 years of cost analysis and audit work I can usually spot it in the first few minutes.

The decisions riding on those numbers are the ones that keep you up at night. Do I hire? Can I take on this bigger job? Which service actually pays the bills, and which just keeps me busy? You answer questions like those with your data — or, when the data can't be trusted, with your gut and a held breath.

Here's the uncomfortable part: numbers can look perfectly tidy and still be lying to you. Neat rows, every field filled in, and underneath them a total nobody checked, a category nobody set, two systems quietly disagreeing. Tidy is not the same as true. So these aren't signs your books are "messy" — they're the tells that your data isn't decision-grade yet, that when you ask it a real question, it guesses. Here are the five I notice first, and what each one is quietly costing you.

1. Nothing has been checked against reality

The first thing I look for is an anchor — some point where the numbers were tied back to what actually happened. When months go by without that, the report stops describing your business and starts describing whatever got typed in. It might be close. It might not. Nobody knows — and that's the problem. You're steering by a gauge no one ever calibrated, and every decision you stack on top inherits the wobble. You're not reading your business. You're reading a rough draft of it and hoping the draft is right.

2. A big chunk of it isn't labeled

Somewhere there's almost always a bucket — "Uncategorized," "Other," "Misc" — and it's bigger than you'd like. To me that pile isn't a chore to clear; it's a blind spot. Every unlabeled item is money moving through your business that you can't see, sort, or question. And it quietly breaks the questions that matter most: Which product line carries the business, and which one bleeds it? Is the new location pulling its weight? You can't answer any of those from an average, because averages hide the story — the money is made and lost in the segments. And you can't slice data you never tagged.

3. Two of your systems tell different stories

The bank says one thing. The sales system says another. The spreadsheet you actually look at says a third. When numbers never quite line up across your tools, the real issue isn't arithmetic — it's that there's no single source of truth. The instant two sources disagree, both become a maybe, and you're left picking whichever feels right, which is a polite word for guessing. Numbers that don't agree don't just cost you accuracy; they cost you nerve. You start hedging every call because the ground underneath keeps shifting. Confident decisions need one set of numbers everyone trusts — not three you have to referee.

4. You only find out long after it mattered

Ask a lot of owners how last month went and the honest answer is, "I'll know in a few weeks." By the time the numbers are assembled, the month they describe is gone — and so is your chance to do anything about it. Worse, getting a straight answer turns into a three-day archaeology dig through exports, so you stop digging. That's the real cost, and it's a quiet one: the questions you stop asking because the answer is too much trouble. Those unasked questions are usually the exact ones that run a business. (Clean, current data is also what makes the year-end hand-off to your CPA painless — that scramble is a symptom, not the disease.) You can't manage what you can't see, and you certainly can't manage it a season too late.

5. The numbers tell you what, never why

Even when an owner can find a number, it's usually a flat one — a total sitting by itself with no story behind it. Say last month was down. Fine: down because of a slow season, a price that slipped, one big job that ran long, a cost quietly creeping up? A single number is a snapshot, not a story. To decide anything you need the shape those numbers make over time — and that shape is the decision; everything before it is just raw material. This is the tell I care about most, because it's the whole point: the difference between data that reports and data that reveals. Without it, every real call falls back on gut, and gut doesn't scale.

What fixing this actually gets you

Notice what none of these are really about. Not one is about neat books for their own sake. Every one is about whether your numbers can answer a question you'd bet a decision on. That's what I actually build — cleaning the data is just where it starts.

From there it becomes a system. We get your data honest and checked against reality, give every piece a label, and pull your scattered tools into one source you can trust. Then we automate the flow so it stays current instead of going stale, and put it in front of you as a dashboard that shows what's really happening — sliced by job, product, location, or customer, updating on its own, ready the moment you have a question. That's the whole pipeline: send your data, clean and connect it, automate it, see the story. Raw data on one end, real answers on the other.

What comes out the far side isn't tidier books. It's the ability to ask "did I make money last month — and why," and get a straight answer, plus the visibility to spot the leak while you can still fix it. Send your data; it comes back as answers. Your numbers stop guessing and start telling you something you can act on.