Two Hires, One Bad Number: When Dirty Data Makes the Decision for You

Emily Goh · 2026-04-19

The pipeline said we were short.

A VP of Sales pulled the Q1 pipeline number from Salesforce last quarter. It looked light. Way under coverage.

So she did what any good sales leader would do. She went to the CEO and said -- we're short on coverage. We need headcount.

The CEO approved two new hires. Recruiter fees. Onboarding costs. Ramp time. Six figures committed before anyone asked one question:

Is that pipeline number actually right?

It wasn't.

Thirty percent of the "open" opportunities hadn't been touched in 90 days. Two large deals were duplicated across regions -- different reps had created separate records for the same account. One deal was already closed-lost, but nobody had updated the stage.

The real pipeline wasn't short. It was just dirty.

And now the company had two new hires it didn't need, solving a problem that didn't exist.

This is what bad data actually does.

Bad data doesn't just sit in your CRM looking messy. It makes decisions for you. Quietly. Confidently. With a spreadsheet attached.

You don't question the number because it came from the system. And the system is supposed to be your source of truth.

But the system only knows what someone told it. And if nobody's checking. Nobody's cleaning. Nobody's flagging the deals that went stale three months ago. Then the "truth" is just a guess with formatting.

That's the part that should scare you. Not that the data is wrong. But that you're making real decisions, headcount, budget, market bets, on data you haven't verified.

It's not just pipeline.

I've seen this pattern play out everywhere.

A company kills a product line because the revenue report says it's underperforming. Turns out half the revenue was attributed to the wrong product code. The product was actually doing fine.

A marketing team gets their budget slashed because "pipeline from campaigns" looked weak. Turns out the attribution model was broken. The campaigns were working -- the CRM just wasn't tracking the connection.

A finance team flags a customer segment as declining. Turns out the segment codes were inconsistent across regions. Pull the data differently and you see growth.

Every bad decision has a data trail. And somewhere in that trail, there's a number nobody verified.

The confidence trap.

Here's the thing about reports. They feel authoritative. A number in a dashboard feels like a fact. A pipeline total in Salesforce feels like the truth.

There's no red warning banner. No "this data might be wrong" disclaimer. The system just shows you the number, and you make a call.

That's the confidence trap. You trust the output because the system looks professional. But the inputs haven't been checked since the last time someone had a spare afternoon -- which was never.

What actually needs to change.

You don't need more dashboards. You don't need a better BI tool. You need to trust what's already in your systems.

And right now, you probably can't.

The fix isn't hiring a data analyst to run spot checks once a quarter. It's having a system that continuously watches your data and tells you when something looks wrong -- before that number ends up in a board deck or a hiring plan.

Not a person asking "does this look right?" A system doing it. Automatically. Every day.

That's the gap most mid-market teams face. Enterprise companies have data engineering teams that build validation pipelines. You have a spreadsheet and a hope that someone catches it before the meeting.

Start here.

Pull your most important number right now. Your pipeline total. Your revenue forecast. Your customer count.

Now ask yourself: would you bet $50K on that number being correct?

If the answer is no -- that's your real problem. Not the dashboard. Not the tool. The data underneath it.

The same bad pipeline data that drove that hire will show up in your next board presentation. See Your Sales Board Deck Is Missing the Numbers That Matter -- it shows exactly what the board actually needs to see, and what they're not seeing.

And for the wider time cost across your ops team: How Much Time Are You Spending Reverse-Engineering Your Own Data?.

Want to see what's hiding in yours? Gohvio is free to try, gohvio.io