9 Ways Ops Teams Find Bad Data (And Why All of Them Are Too Late)

Emily Goh · 2026-04-04

The same story, nine different ways.

Ask any ops team how they find bad data. The answers are almost always the same nine things. And every single one of them is reactive. The bad data gets discovered after it's already cost something.

Here's the full pattern.

1. It breaks a report you're presenting.

You're building a board deck. You pull the pipeline forecast. The number looks high. You do a spot check and find three deals that shouldn't be there, stale, duplicated, or already closed.

Now you're rebuilding the forecast on the 27th because you're presenting on the 28th.

Cost: Hours of work. Stress. Risk that the corrected number is still wrong.

The data was broken the whole time. It only surfaced because it was about to go in front of important people.

2. A customer mentions something that contradicts your record.

A customer calls the CSM. "Hey, sending you that quote we discussed last week." The CSM pulls up the account. Last activity was three months ago. No quote. No record of that conversation.

The AE was emailing them directly. Never logged it. The system shows zero activity. The CSM looks like they've been ignoring the account.

Cost: Customer experience hit. Team miscommunication. Time spent figuring out what actually happened.

The data was incomplete for three months. It surfaced when the customer called.

3. A deal doesn't close and you have to figure out why.

You were forecasting a $150K deal to close this quarter. Two weeks past the expected close date, it's still not in closed-won. You ask the rep. "Oh, they pushed back six weeks."

You check the record. Close date: March 15. It's April 28. Nobody updated it.

Cost: Forecast is wrong. Pipeline is wrong. You only found it because the deal was late.

4. Month-end reconciliation surfaces inconsistencies.

Finance is closing the books. Revenue in Salesforce doesn't match the GL. Someone opens a spreadsheet. Four hours later: deals recognized twice, invoices without a matching record, open deals that should have closed months ago.

Cost: Hours of finance time. Delayed close. Risk that the reconciliation misses something.

The data was inconsistent the whole time. It surfaced because of month-end.

5. An integration breaks and nobody notices for days.

The Zapier sync from the lead form to Salesforce breaks. New leads stop flowing. Nobody notices until someone asks why pipeline hasn't moved in five days. The integration broke on Day 2. Seventy-two hours of leads are gone.

Cost: Lost leads. Wrong pipeline projections. Time troubleshooting.

The data was incomplete the whole time. Found it by accident.

6. Someone audits a slice and finds problems everywhere.

The compliance person runs an audit. They spot-check 100 customer records and find that 12% have stale data, wrong phone numbers, old titles, missing required fields.

If 12% of 100 are bad, the full database could be 12% bad. But no one knows for sure.

Cost: Audit risk. Possible compliance issue. You're guessing at the scale.

7. A campaign fails and you have to dig into why.

You send a campaign to the "active customers" segment. Bounce rate is 22%. Way above normal. You dig in. Half the segment is actually inactive. The filter logic didn't work the way anyone thought.

Cost: Wasted spend. ISP reputation hit from the bounce rate. Hours debugging segmentation.

The data was wrong. Found it because a campaign failed.

8. A new hire asks a question nobody can answer.

A new CSM asks: "If a customer has multiple records, which one is the real one?" And the answer is... unclear.

Someone digs in. Forty-seven companies have duplicate records. No one knows which is the source of truth. The new person doesn't have a reliable way to do their job.

Cost: Onboarding delays. Confusion. Loss of confidence in the data.

The duplicates were always there. They surfaced because a new hire asked a basic question.

9. An executive makes a big decision on data that turns out to be wrong.

The VP of Customer Success looks at churn data: "3.2% and trending down." She staffs based on that number. Six weeks later, accounting says actual churn is 6.7%. The data was three months old and included accounts that had already churned.

She overstaffed. Costs get cut. Morale takes a hit.

Cost: Budgeting error. Team morale. Hiring costs that get reversed.

The data was bad the whole time. An executive made a major decision on it.

The pattern.

None of these are planned. Bad data doesn't get found proactively, it gets found when something else goes wrong and someone has to investigate.

All of them are expensive. Time, credibility, customer experience, or actual money.

All of them could have been caught earlier. If the data issue was visible before it affected a decision, the damage was either avoidable or manageable.

The difference between teams that handle this well and teams that don't isn't better auditors. It's continuous visibility. They know what's in their data before they use it for something that matters.

What actually needs to happen.

The goal isn't a better spreadsheet audit or a monthly review. It's a standing process that continuously asks:

Are there duplicates? Stale records? Missing required fields? Inconsistencies between systems? Orphaned records?

Not quarterly. Ongoing. And visible before a decision gets made on it.

Start with the three most critical data sources: the CRM, the customer database, the activity log. For each one, ask: "What would I need to know about this data before I'd be comfortable using it for a decision?"

That's the audit. That's the starting point.

Read next:

The Myth of the Single Source of Truth, why the "one system" fix doesn't solve the underlying data problem.

Do You Actually Trust Your Salesforce Pipeline?, the honest test for whether your pipeline forecast is decision-ready.

Finding bad data reactively is one problem. The other is what it costs you while you're looking. Read How Much Time Are You Spending Reverse-Engineering Your Own Data? -- that's the time tax you're absorbing every week.

And if the bad data is in your finance chain specifically: From Contract to Cash: The Finance Data Chain Nobody Audits.

Gohvio runs continuous data checks so your team isn't finding problems in board meetings. Start your free trial to see what's in yours.