What Does Garbage In, Garbage Out Mean?

Short Answer

Garbage in, garbage out, often shortened to GIGO, means a system’s output can only be as good as the information it’s given. Software, spreadsheets, and AI don’t judge whether data is right. They process it. Errors in customer lists, prices, inventory, or financial assumptions come back out as wrong messages, wrong reports, or wrong forecasts, often faster and with more confidence than a person would produce.

Quality of results = quality of inputs (the tool can’t add what the data lacks)

Doctor Owl, in Plain English

“It did exactly what it was told. That is the trouble.”

How This Guide Works

Want the idea fast? Read the definition and the checklist. Want the story? Follow Marco, a manager from Dr. Paul’s Fabled Forest Business School Series who watches an automatic reminder go out to a list nobody had cleaned in years. The table in the middle shows what one cleanup removed.

Are Your Numbers Built on Shaky Records?

Get your records and projections checked by Dr. Paul, or learn the basics first. Start with the list you rely on most.

Business Consulting

Bad data in your systems? Dr. Paul helps you find the errors that matter and fix them before you automate.

See consulting options →

Business Plan Templates

Industry templates with a Word plan and Excel model, built on sensible defaults you can verify.

Browse templates →

Small Business Education

Free guides and Fabled Forest books on data, technology, and numbers.

Start learning →

Meet Marco: A Manager Watching the Gate Fill Up

Marco has the pizza business to himself for a week in Dr. Paul’s Fabled Forest Business School Series. The trucks run perfectly, because the routines are written down. The trouble comes from the machine on the office wall.

It has sent a Monday reminder to every name on the customer list. By mid-morning, creatures are at the gate: a mouse from a bakery that closed in the autumn, a hedgehog with three identical messages, and more by ten o’clock. The list had been growing for years, and nobody had ever cleaned it.

Why Do Bad Records Cause Bigger Problems Once You Automate?

Because a person notices and a machine doesn’t. When staff use a messy list, they skip the closed business and the duplicate. When a tool uses it, every error goes out, every time, to everyone on it. The same is true of a spreadsheet forecast built on a guessed growth rate or an AI tool fed outdated prices.

That’s why data quality comes before workflow automation. It belongs to the Organize step of Dr. Paul’s Organize-Plan-Grow™ Strategy: clean the inputs, then let the tool run.

Signs your inputs are the problem:

  • Customers get duplicate or irrelevant messages.
  • Reports don’t match what you see in the shop.
  • Inventory counts in the system are often wrong.
  • Your forecast rests on numbers nobody can source.
  • Several versions of the same list exist.
  • AI tools give confident answers that turn out wrong.

In the Fabled Forest

Doctor Owl looked at the machine on the wall and then at the list it had used. The machine wasn’t broken, he said later. It had sent to every name, exactly as told. The names were the problem.

How Do You Check Data Before a Tool Uses It?

  • Remove duplicates. Merge people or items that appear more than once.
  • Remove what’s gone. Closed customers, discontinued products, departed staff.
  • Fix wrong details. Hold questionable records until they’re confirmed.
  • Honor every no. Screen against your do-not-contact list.
  • Source every assumption. In forecasts, note where each number came from.

In the Fabled Forest

Marco struck out the doubles and the creatures who were gone, then copied what was left onto a fresh scroll, one name to a line, with the same details in the same order on every line.

How Do You Stop Bad Data From Coming Back?

Give each important list one owner and one home, so there’s a single version everyone uses. A single source of truth stops copies from drifting apart. Then check it on a schedule, before any big send or report.

For financial projections, write down where each assumption came from: last year’s sales, a lease quote, a supplier price. If you can’t source a number, flag it as a guess, so nobody treats it as fact.

In the Fabled Forest

The customers with wrong numbers weren’t thrown away. They went on a card of their own until the drivers could find them, and four turned up by Saturday. Nothing was sent to the rest until they were.

What Does a Cleanup Remove From a Typical List?

List math ahead. Skip it if the checks are enough.

Here is Marco’s list, from the book. Run the same count on your own before automating anything.

Line itemsCount
Names on the list220
Duplicates−40
Closed, moved, or gone−18
Wrong contact details, held−11
Asked not to be contacted−6
Clean, ready to use145

Counts from the book.

Step 1: Count the bad lines. Seventy-five of two hundred and twenty, about a third.

Step 2: Picture every send. Without cleaning, a third of every message would be wasted or unwelcome.

Step 3: Clean before the next run. Then check monthly, so it never drifts that far again.

Book counts; your proportions will differ.

How Do You Avoid Garbage In, Garbage Out?

Four steps before any tool, report, or forecast runs on your data.

  1. Pick the list or numbers the tool will use.

    Customers, products, prices, or financial assumptions.

  2. Clean it.

    Remove doubles and departed records, fix errors, and screen out opt-outs.

  3. Name one owner and one home.

    A single version everyone uses.

  4. Recheck on a schedule.

    Monthly for lists, at every update for forecasts. The common mistake is automating old records as they are.

In the Fabled Forest

When Frankie came back and saw the clean list, he didn’t take it back. He handed it straight to Marco to keep, checked on the first Monday of every month.

What Did the Hedgehog Think He’d Ordered?

The book has the gate filling up by ten, the mouse from the closed bakery, the pencil down every line, the card for the missing numbers, and the hawk’s old list coming out of a drawer the same week.

Technology and AI for Small Businesses covers checking what machines are fed before letting them act.

How Does a Plan Writer Keep Bad Inputs Out of a Forecast?

Dr. Paul Borosky, DBA, MBA · Business Consultant

When Dr. Paul builds financial projections, every major assumption gets a source: last year’s sales, a signed lease, a supplier quote, or an industry benchmark. Anything he can’t source is labeled as an estimate, so a lender knows which numbers are firm. He sees the same principle in operations. Owners who automate marketing or ordering on an uncleaned list get fast, confident mistakes. So before any tool goes live, he has owners clean the data, name one owner for it, and set a recheck date.

What Else Do Owners Ask About Garbage In, Garbage Out?

What is an example of garbage in, garbage out?
An automated reminder sent to an old customer list reaches closed businesses, duplicates, and people who asked not to be contacted, because the list was never cleaned.
Does garbage in, garbage out apply to AI?
Yes. AI tools work from the information they’re given. Outdated or wrong inputs produce confident but wrong answers, so check what you feed them.
How does it apply to a business plan?
A forecast is only as reliable as its assumptions. Source every major number, label estimates clearly, and update them as real results come in.

Written by Dr. Paul Borosky, DBA, MBA, business consultant and author of books on Amazon. Want help setting this up? Check your data and numbers with Dr. Paul.

This page is for general education. It is not legal, security, or privacy-law advice, and no product is named or endorsed. Rules differ by place, so check yours. The example business is invented and the Fabled Forest is fictional.