National mortgage brokerage
Replacing an enterprise CRM saved roughly $500K a year
- saved per year in licensing and consulting costs
- ~$500K
- users doubled. Software costs didn't.
- 40 → 80
- the quoted AI add-on price, designed out entirely
- $600/user
saved per year in licensing and consulting costs
users doubled. Software costs didn't.
the quoted AI add-on price, designed out entirely
The situation
The brokerage ran its deal flow on Salesforce enterprise licenses. The seats were expensive, but the bigger cost was friction: workflows designed for someone else's business, changes that took months and outside consultants, and a hard ceiling on what AI could do inside it. Every new feature came with its own price tag.
When leadership priced what it would take to properly use AI and agents on the platform, with Agentforce and the AI add-ons, the quote came to roughly $600 per user. That tracks with public list pricing: Salesforce's Agentforce 1 editions start at $550 per user, per month. The bill would grow every time the team did.
What we did
We embedded with the team and mapped how deals actually move, from first contact through funding, including every workaround staff had built to survive the old system.
Then we built a replacement designed around those real workflows, with AI woven through the whole operation: calls transcribed and analyzed, documents processed, deals analyzed, leads qualified and distributed, and accounting handled all the way down to commission payouts. The brokerage owns the system, so there is no per-seat meter, and improvements ship in days, shaped directly by the people who use them.
What we enabled
- Call transcription & analysis
- Document processing
- Deal analysis
- Lead qualification
- Lead distribution
- Accounting & commission payouts
- AI summaries & follow-ups
- Data hygiene & record keeping
The result
The brokerage saves roughly $500K a year in licensing and consulting costs. The cost curve changed too: when the team grew from about 40 users to 80, the software bill barely moved. Owning the system means growth no longer buys the same software twice.
And because we designed the system for change, new technology slots in easily. When a better model or capability ships, the brokerage has it in days, not contract cycles, and the team keeps saving time as the tools improve.
Per-seat licensing + AI add-ons
Cost doubles when the team doubles.
The system they own now
40 → 80 users. The bill barely moved.
Facing something similar? We’ll tell you straight whether AI can help.