Program report · filed November 16, 2026
CRM & Sales TechMeasured report
Legacy CRM’s 10% usage estimate sharpens the replacement debate
CRM environments often grow for 20 years or more and can constrain AI-led service, while Treasure AI executive Rafael Flores estimates buyers use only 10% of what they pay for.
By Daniel Okafor4 min read739 words
Program notes
- Organizations typically build CRM environments for 20 years or more, according to ServiceNow’s Daniel Wilks.
- Enterprises typically use about 10% of traditional CRM functionality but pay for 100%, Rafael Flores estimates.
- Legacy CRM may record customer activity but struggle to coordinate request resolution across the wider business.
- No migration-cost, payback or implementation-time benchmark is provided in the source.

By Daniel Okafor
About 10% of traditional CRM functionality is typically used by an enterprise, while the buyer pays for 100%, according to Treasure AI executive Rafael Flores. The estimate, offered in a CX Today discussion without a disclosed sample or method, is not an independent benchmark; buyers can test it against feature use, integration work and total cost of ownership.
ServiceNow Vice President of CRM EMEA Daniel Wilks told CX Today: “Organizations have typically spent 20 years or more building CRM systems that were really designed to build a single view of customer and became these vast applications with huge amounts of customer data.”
Organizations layered custom code, integrations and workarounds onto systems designed mainly to record activity and provide a single customer view. The result was software that could be expensive to maintain, difficult to change and poorly suited to modern customer service.
What does this change for CX buyers?
Buying teams now face two tests: whether installed CRM can evolve enough for AI-driven service, and whether its architecture will constrain organizational outcomes.
For an enterprise buying team at the evaluation stage, replacement risk begins with discovery. Customer records, business processes, integrations and custom applications can become so intertwined that even determining what must move becomes a major project. Buying teams need a documented baseline before comparing platforms or judging whether the installed CRM can evolve.
- Pre-contract evaluation, enterprise buying team: Map customer records, business processes, integrations and custom applications; identify necessary processes, redundant integrations and customizations tied to important requirements.
- Technical validation, enterprise buying team: Test whether the platform can interpret conversations, understand customer signals and trigger actions across business processes, rather than only retain structured records.
- Operational validation, customer-facing team: Test whether it can coordinate request resolution across the wider business, not merely record activity and show what happened during an interaction.
- Commercial diligence, enterprise buying team: Compare maintenance, customization and costs for functionality rarely used with the scale of the migration assessment.
Wilks’s broader diagnosis is that CRM intended to serve customers had ended with organizations serving the system. When customer-facing teams spend significant time maintaining records, navigating applications and working around disconnected processes, the technology can obstruct the experience it was introduced to support.
Traditional CRM architecture creates a more specific problem for AI adoption. Many platforms were built around human data entry and structured databases. AI-driven customer operations require systems that can interpret conversations, understand customer signals and trigger actions across processes.
Aurasell Chief Executive Officer Jason Eubanks previously told CX Today that buyers cannot expect the same level of execution and outcomes from software built around a different architectural paradigm. That is an architectural claim, not a comparative performance result; the source provides no test sample or benchmark.
For customer-facing teams handling service resolution, the operational test goes beyond capturing an interaction. Traditional CRM can record customer activity and show what happened during a conversation, but it may struggle to coordinate the work required to resolve a request across the wider business.
Adding AI to an existing installation does not automatically repair those constraints. Wilks cautions against building AI capabilities on top of disconnected systems.
Keeping legacy CRM also carries costs. Over time, businesses connect CRM platforms to billing, order management, service operations, marketing platforms and other enterprise applications. They also accumulate poorly documented custom objects, workflows and business rules. Wilks compares the combination to a monster assembled from different applications, held together by extensive customization and a high total cost of ownership. Each added layer complicates migration.
Flores said: “You only use about 10 percent of it typically, but you pay for 100 percent of it.” He also argues that the end of support for a legacy product creates an opportunity to reassess the wider customer technology stack, including whether CRM should remain its central component.
The practical change for enterprise buyers is a two-part test. First, verify the 10% estimate against their own feature, integration and support requirements. Second, treat end of support as a trigger for a stack-level review, not an automatic instruction to replace CRM.
The source gives no migration-cost, payback or implementation-time benchmark; those figures must come from each buyer’s records and assessment. The next CRM buying cycle will turn on whether buyers can connect AI to coordinated customer action rather than add another layer to systems that already consume maintenance time and budget.
via cxtoday.com (Source)
More from Daniel Okafor
Show full bio
Correspondent covering regulation, policy and their impact on Sales technique and methodology, sales enablement at Quota Brief.
127 articles