Do You Actually Trust Your Salesforce Pipeline?

Emily Goh · 2026-04-09

The honesty test.

Pull up your Salesforce pipeline right now.

Now imagine the CEO walks into your office and says "I'm going to bet $50,000 on the accuracy of that forecast. Is it right?"

Would you take that bet?

If the answer is anything other than "absolutely," you have a pipeline trust problem. Roughly 80% of the mid-market teams I talk to would lose that bet. Not because they're bad at sales. Because the data underneath the forecast is messy.

What's actually broken.

Phantom opportunities. A deal gets created in 2023. Nobody closes it. Nobody marks it closed-lost. Two years later, you're still showing that deal as a potential close. The contact has moved jobs. The deal is dead. But it's still in your pipe.

Stale stages. A deal moves to "Negotiation" in January. The customer says they'll get back to you. It's now April. The deal is still in Negotiation. Your near-term forecast is counting something that stalled three months ago.

Wrong contact people. The contact person left the company six months ago. You didn't update the record. You're not following up with the right person. Your engagement metrics are off because you're tracking a ghost.

Duplicate deals. A deal gets created. Then someone re-creates it because they didn't see the first one. Now you have two deals for the same opportunity, in different stages, with different values. One company I worked with had a $600K deal listed four times.

Closed deals that aren't closed. A deal closes and hits the GL. But someone forgets to mark it "closed-won" in Salesforce. It's still sitting in Negotiation months later, inflating your current pipeline.

Why this matters.

None of these problems are visible when you run a standard pipeline report. You open Salesforce. The number feels real. It came from your CRM.

But that number includes phantom deals, stale stages, wrong contacts, and duplicates. When you make decisions on that number, you're making decisions on garbage.

You hire headcount based on pipeline that includes phantom opps. You commit to a board forecast inflated by 30%.

The audit test.

Pick 20 random open opportunities. For each one, ask:

• Is this contact person still at the company?
• Is this the right stage for the deal?
• Are there any duplicate records for this opportunity?
• Has there been communication with this contact in the last 30 days?
• Would I actually be surprised if this deal closed next month?

If more than 25% of your answers are "no" or "I don't know," your pipeline has a serious trust problem. Most teams score worse than that.

How to fix it.

You don't need a new CRM. You just need to know what's actually in your data.

A systematic check that answers:

• Which deals are duplicates?
• Which deals haven't had communication in 60+ days?
• Which contacts have left their companies?
• Which records are orphaned or broken?

Not manually. Not through a spot check. Systematically. Before you use that data for forecasting.

The real question.

Why would you bet $50K on a number you've never actually audited?

You wouldn't. So don't make bigger bets on that data without knowing what's in it first.

Pull your pipeline. Look at it. Ask that honesty question again.

If you still wouldn't bet on it, the problem isn't your sales team or your forecast methodology.

The problem is the data.

A 3-step pipeline trust audit.

Three questions a sales leader can ask in a forecast call to spot trust issues immediately:

1. Of the top ten deals in this forecast, when was the last real conversation with the buyer, and is that buyer still in role?
2. Which of these deals exist in only one place in our stack, and which appear in two systems with different values?
3. If we slipped every Stage 4 deal that has not moved in 30 days, what does the quarter look like?

If any answer is "I am not sure," that is where the trust gap lives.

This pipeline data problem shows up directly in board reporting. Read Your Sales Board Deck Is Missing the Numbers That Matter for how to present pipeline coverage, win rate, and forecast accuracy in a way the board can actually trust.