🧠 Food for thought

Implications, context, and why it matters.

Salesforce pricing exposes AI adoption hurdles

• Salesforce says Agentforce 360 suits large companies. The model adds complexity that can slow rollouts. Buyers pick either conversation pricing at $2 per conversation 1 or Flex Credits (prepaid usage units) at $0.10 per action, which equals 20 credits 1. They cannot mix models in one org (a single-customer Salesforce tenant) 2. That forces a guess on usage before real data exists, though you can switch later by swapping SKUs (stock keeping units that represent price-plan choices) 2.
• An ROI estimator 2 and Digital Wallet tracking (an in-product ledger for usage and spend) 2 can forecast costs. Teams still lack baseline data for agent action counts across use cases. One case management setup that consumes 360,000 Flex Credits a month ($1,800) 2 looks manageable. Wider rollout without a rollover option 2 turns miscalculations into budget hits, which slows purchasing despite the flexible pitch.

Third-party firms can help with the semantic layer gap

• Tableau Semantics now ties into Data 360 3, which creates demand for data prep services that build a semantic layer (a business-friendly data vocabulary) across information scattered across many systems. Salesforce works with Databricks, dbt Labs, and Snowflake to align semantics 3. Many firms still hold unstructured content and clashing definitions, which Intelligent Context targets 4.
• Providers that focus on semantic mapping, document processing acceleration (promised in “hours” via low-code 4), and permission-aware unstructured data extraction (respecting existing access controls) can win work around Agentforce 360. Demand for a semantic layer goes beyond Salesforce customers. Any organization that needs AI-ready information with business context fits this need 5. Gartner (a technology research and advisory firm) says 60% of AI projects will fail by 2026 without solid data foundations 5.