AI Budget Expansion Squeezes Traditional SaaS Market Share
Taylor Wilson
Companies are redirecting spend from Salesforce, ServiceNow, and other legacy SaaS toward Anthropic, OpenAI, and AI-native startups — some cutting software budgets by 50% — raising a stark question: will traditional vendors be reduced to data warehouses stripped of pricing power?
Where is the money moving — and to whom?
Multiple companies have begun cutting ServiceNow and Salesforce usage, shifting to self-built AI agents or third-party AI platforms, The Information reports.
This means → the total enterprise software wallet may not be shrinking, but the flow is structurally redirecting — away from subscription SaaS, toward AI-native tools.
French pharma giant Sanofi is a case in point: it used Anthropic's Claude Code and startup Elementum's software to build its own AI agents, reducing reliance on ServiceNow's IT management modules.
Why does the Sanofi playbook deserve a closer look?
Sanofi cut ServiceNow while deepening its partnership with SAP — connecting its own AI agents with SAP's to automate procurement-order audits.
In plain terms = not every legacy vendor loses; those that integrate deeply with AI (like SAP) actually gain budget share. Only those that can't integrate get replaced.
Sanofi also sharply reduced its use of Indian outsourcing firms — AI agents are displacing not just software, but human labor too.
How aggressively are smaller companies switching?
Mixology, a 500-person family-owned apparel firm, adopted Palantir last year to build custom AI apps for ad generation, demand forecasting, and shift scheduling — and plans to scale back Salesforce CRM and marketing modules.
CEO Jordan Edwards: "Before I touched Palantir, my entire business ran on Salesforce. I'm not sure it's necessary going forward."
Israeli startup Utila went further: it dropped 10 small-vendor SaaS tools at once — Clay, Vendelux, and others covering lead tracking, email marketing, event management — consolidating onto Swan AI's sales and marketing platform. Result: a 50% cut in software spend.
What role does legacy SaaS end up playing?
Utila killed 10 tools but kept HubSpot — not for its features, only as a data store for Swan AI's agents to pull from.
This means → legacy SaaS is sliding from "you use my features to get work done" to "you only use me to store data" — the role shifts from functional tool to data warehouse, and pricing power erodes accordingly.
Starbucks is also reported by Bloomberg to be exploring AI replacements for key Microsoft and IBM applications.
What is the central unresolved question?
EY global vice chair Julie Teigland says clients are using AI procurement as a chance to consolidate enterprise apps and clear historical tech debt, aiming to "make systems simpler and easier to migrate."
This reflects a deeper signal: companies aren't just swapping tools — they're using the AI transition to rethink their entire software architecture.
Whether traditional SaaS vendors can defend their premium through AI upgrades — or will be forced into a low-margin "data-layer" position — is the defining question of this market shift.
Content is for reference only, not financial advice.