Agentforce Slow Adoption Exposes Enterprise Data Readiness Gap

What happened

Salesforce is facing slower-than-expected adoption of Agentforce, its flagship agentic AI platform, with enterprise customers struggling to deploy the technology at meaningful scale. The difficulties, reported by MarTech, are not primarily a product problem — they reflect deep-seated issues with data quality and operational readiness inside the organisations attempting to use it.

Agentforce is designed to allow AI agents to autonomously execute marketing, sales and service tasks on behalf of businesses. But the gap between that promise and live deployment has proved wider than Salesforce anticipated, with customers finding that their underlying data infrastructure is too fragmented, incomplete or poorly governed to give AI agents reliable inputs to act on.

The situation has drawn attention to a broader pattern in enterprise AI adoption: vendors can ship capable models and orchestration layers, but readiness on the customer side — clean data, clear workflows, defined governance — remains the binding constraint on real-world performance.

Why it matters

For customer experience and service-design practitioners, the Salesforce episode is a sharp reminder that agentic AI is not a plug-in upgrade — it is an organisational transformation that surfaces every pre-existing data and process failure. An AI agent tasked with resolving a customer query, personalising an offer or escalating a complaint can only perform as well as the data it is handed. Poor data quality does not just limit AI accuracy; it actively manufactures bad customer experiences at machine speed and scale.

From a behavioural economics perspective, there is also a significant expectation-management problem. Enterprises that have invested in Agentforce on the basis of vendor-led optimism may now face internal credibility damage — making future AI initiatives harder to fund and staff. The lesson for CX leaders is that the sequencing of AI adoption matters enormously: data readiness is not a precondition to tick off eventually, it is the first deliverable.

The Renascence take

Most of the commentary around slow Agentforce adoption will focus on Salesforce's sales pipeline or its competitive position against Microsoft and ServiceNow. That misses the more consequential story sitting directly underneath it.

The real revelation here is not that agentic AI is hard to sell — it is that most enterprises do not yet know what their customers' data actually looks like at the point of action. Organisations have spent years building CRM systems and data lakes while quietly tolerating the inconsistencies inside them, because human agents could compensate through judgement and improvisation. AI agents cannot. They will act on what they are given, faithfully and at volume. The operators who will win with agentic AI are not those who move fastest to deploy it, but those who have already done the unglamorous work of auditing, unifying and governing their customer data — and who treat that infrastructure as a CX asset, not an IT cost centre.