When to Leave Salesforce: A 5-Year TCO Model
The honest math on per-seat licences, admin overhead, and what custom CRM actually costs at the 200-seat break-even line.
The honest math on per-seat licences, admin overhead, and what custom CRM actually costs at the 200-seat break-even line.
The visible bill
Most TCO conversations start at the line item: 200 seats × Sales Cloud Enterprise at ~$165/seat/month is $396k/year, before Service Cloud, before Marketing Cloud, before sandboxes. Five years compounds to roughly $2M. That number is bad enough on its own, but it is also the smallest number that will appear in this post — every line below adds to it, and none subtract.
What replacement actually costs
A custom CRM scoped to your actual workflows — not Salesforce parity — runs $180–$350k to ship and roughly 18–25% of that per year to maintain. The replacement is rarely "all of Salesforce" — it is the 22% of Salesforce your team actually uses, plus the integrations you already paid Salesforce to do badly. Most engagements ship in 4–7 months on a senior team.
The break-even line
The math gets compelling above ~150 seats once you account for hidden costs. Below that, Salesforce + a careful admin is hard to beat. Above 250 seats with a high-customisation setup, custom is almost always cheaper inside three years. The decision rarely turns on the build cost; it turns on whether you have engineering capacity to maintain the replacement.
When to leave (and when not to)
Leave when (a) Salesforce is the largest line in your software budget, (b) your admin team has become a bottleneck on every process change, and (c) at least three of your business-critical workflows are too custom for the platform. Stay when you have under 100 seats, when your sales motion is genuinely off-the-shelf, or when you do not have an engineering function. There is no honour in leaving for its own sake.