Salesforce is typically the third-largest software line item, but almost no one in finance owns it. Here are the five places the money leaks — and the five questions that surface waste in a single meeting.
Most software has a clear owner and a clear outcome: the data warehouse team owns Snowflake, and query performance is the metric. Salesforce is different. Sales owns it. Marketing uses it. Finance reports from it. Ops builds in it. Nobody is accountable for the return.
This diffusion of ownership means no one asks the uncomfortable question: are we getting back more than we're spending? That silence is expensive. The median mid-market company pays $85–150K per year in licenses alone, before admin headcount, integrations, and consulting.
Most orgs have 15–25% of their seats unused or underused — reps who left, roles that changed, users who do everything in a spreadsheet. Salesforce reports this clearly in the login history. A 20-seat org paying $150/user/month with 5 ghost licenses is burning $9,000/year on access nobody's using.
The fix: Pull a 90-day login report. Seats without a single login get downgraded or removed at next renewal. Do this before the renewal conversation, not after.
The biggest hidden cost in Salesforce isn't the license — it's the admin. A full-time Salesforce admin costs $85,000–$123,000 all-in, and most of their week goes to recurring work: pipeline hygiene, data deduplication, report requests, field updates. That's work AI agents now handle for $500–850/month.
If your admin is spending more than 30% of their time on recurring maintenance rather than process improvement, you're paying senior-engineer rates for clerical work.
Finance builds models on pipeline coverage. If the pipeline is padded with zombie deals — stale close dates, no recent activity, amounts that haven't changed since Q1 — then every forecast conversation is starting from a lie. The board deck looks fine until it doesn't.
A clean pipeline isn't just a sales hygiene issue. It's a financial reporting issue. The CFO relying on a 3x pipeline coverage model needs to know whether that 3x is real.
Salesforce is rarely the only system. Most orgs have it connected to HubSpot, NetSuite, Salesloft, Outreach, Gong, or some combination. Each integration is a data contract. When the contract breaks — a field mapping gets out of sync, an API key rotates, a record type changes — data starts flowing incorrectly or not at all.
The failure is usually silent. Reps don't notice because they're not looking at the integration layer. Finance notices months later when the numbers don't reconcile.
The hardest spend to quantify is the decision that was made on bad data. A territory plan built on inaccurate win rates. A headcount model anchored to a pipeline that was 40% zombies. A product roadmap prioritized by deal stage data that reps entered wrong.
None of these show up as a line item. They show up as missed targets and post-mortems that conclude the data was unreliable — which everyone knew but nobody fixed.
If any of these questions produce silence or 'I'm not sure,' you've found the leak.
Start with the free read-only Pipeline Hygiene Audit — findings within days, no install, no obligation.
Get the free audit