Acme Financial Is in Your CRM 4 Times. That's Why You Hired 3 Reps You Didn't Need.

Emily Goh · 2026-05-05

The account exists. Four times.

Acme Financial. Acme Financial Bank. Acme Financial Holdings plc. Acme Financial -- North America.

Four records. Same company. One in your enterprise territory. Two in your mid-market bucket. One unassigned.

Your Sales Ops team runs the territory model in Q4. They pull all accounts from Salesforce and count them by region, segment, and industry. The model says the Northeast enterprise territory has 340 accounts. Good coverage. Enough to justify a new hire.

They hire the rep.

The rep spends 60 days working the territory. By month 3, they're reporting back: "A lot of these accounts are already customers or they're tiny." Coverage is thinner than the model suggested.

Because the territory had 340 records -- but only 180 real companies.

The other 160 were duplicates.

The math that breaks everything downstream

Duplicate accounts are not a housekeeping problem. They are a strategic planning problem.

Here is what they actually break, in order:

1. Territory size is wrong. If 30-40% of your accounts are duplicates -- and 45% of Salesforce records analysed across 12 billion records were found to be duplicates -- your coverage model is built on inflated data. You think reps have 300 accounts to work. They have 180.

2. Headcount model is wrong. Territory size drives headcount. If you are modelling one rep per 300 accounts, a territory that actually has 180 real companies should have 0.6 reps -- not 1. Stack this across 8 territories and you have over-hired by 3-4 heads.

3. Quota targets are wrong. Quotas are set against territory potential. If territory potential is based on 340 fictional accounts, the quota is fiction. The rep misses. You put them on a PIP. They were always set up to fail.

4. ABM tiers are wrong. Your top 50 enterprise accounts have 90 records because the same companies were entered by different reps over 3 years. Two reps emailing the same buyer. One at a discount you didn't approve.

5. Pipeline rollup is wrong. Opportunities attached to duplicate accounts appear twice. Your $12M pipeline might be $9M. You find out at quarter end.

What it looks like on a Tuesday

It is the Tuesday before your Q3 planning meeting.

Your Sales Ops lead is pulling territory counts from Salesforce. They send you a spreadsheet. 12 territories. Account counts. Segment breakdowns.

You look at the data. It looks clean.

It is not clean. It just looks like it is.

Acme Financial is in there four times. Globex Corp and Globex Inc are separate territories. Northwind Health is split between your pharma vertical and your healthcare vertical because two reps named it differently in 2023.

You make the headcount call based on this data. You hire. You set quotas. You miss.

And the post-mortem says "the reps didn't perform." But the reps were always working from a map that had the wrong number of cities on it.

The specific version of this that Sales Ops hates most

Here's the one that causes the actual fight.

A major account -- a real prospect at $200K ARR -- has three records in Salesforce. Three reps think they own it. One just sent a proposal at 8% discount. Another promised 12% to get the meeting. A third is mid-discovery call.

The prospect gets three different proposals in the same week. They go to your competitor. "You seem disorganized."

Sales Ops finds out during the win/loss review. By then, two reps are already on notice for missing quota and a third has quietly started looking.

What the manual fix looks like

Every quarter, a dedicated ops person runs a dedup review.

They export every account from Salesforce -- 12,000 records for a mid-market company. They try to normalize company names: strip "Inc," "LLC," "Ltd." They match on domain where the domain field is populated (it often isn't). They flag potential duplicates. They route them to account owners for confirmation. Owners either ignore the request or approve without checking.

This process costs 40-60 hours per quarter. At $80K fully-loaded for an ops analyst, that is roughly $15,000 a year spent finding and re-finding the same duplicates.

It also does not catch the ones being created right now, while the last review is still in progress.

What needs to happen instead

Continuous. Not quarterly.

Every time a new account enters Salesforce -- from a list import, from an SDR creating a record, from a HubSpot sync -- it should be checked against what already exists. Same domain? Flag it. Similar name? Flag it. Same phone or address? Flag it.

The human still makes the merge decision. But they are making it in real time, when the account is new and the cost of fixing it is one click -- not three weeks into a territory planning cycle when the cost is a bad headcount decision.

For how duplicate accounts affect your lead routing specifically, read Your Leads Are Being Routed to the Wrong Reps. For how pipeline rollup errors show up in board reporting, read Your Sales Board Deck Is Missing the Numbers That Matter. And for the full cost of decisions made on bad data, read The Real Cost of Dirty Data (That Nobody Tracks).

Gohvio is free to try -- gohvio.io