Enterprise AI Displacement — Digest for Current Cycle
TL;DR
The ERP market has entered structural disruption. Three venture-backed AI-native startups—Campfire, Rillet, and DualEntry—have collectively raised over $300 million and are winning direct customer migrations from NetSuite and SAP by compressing implementation timelines from 12–18 months to 4 weeks or less. Legacy incumbents have responded with massive defensive AI platforms (SAP's Autonomous Enterprise, Oracle NetSuite's NetSuite Next), but the core vulnerability remains unchanged: the 48-hour migration and 5x faster close times are real, and they undermine the lock-in that has protected ERP vendors for two decades.
The Implementation Timeline Barrier Has Collapsed
The single largest moat protecting legacy ERP vendors—the dreaded 12-to-18-month, half-million-dollar implementation—is being dismantled by migration automation, and that changes everything about switching economics.
"DualEntry built a highly optimized migration engine that can migrate companies from legacy systems (like QuickBooks and NetSuite) in 24 to 48 hours, eliminating the typical 18-month, high-cost implementation risk." — The Rise of AI-Native ERPs
+3
Rillet is executing 4-week implementations where legacy systems required a year. Campfire reduces financial close cycles from 15 days to 3 days—a 5x improvement—by automating reconciliations and variance analysis through its proprietary Large Accounting Model trained exclusively on financial data. These aren't marginal gains. They're fundamental shifts in the unit economics of ERP deployment.
The strategic implication is acute: if switching costs collapse, then switching happens. Named customers like PostHog, Replit, and Postscript are already making the move. Windsurf and Postscript—both operating at $100M+ ARR scale—are running their entire finance operations with only 2 people and closing their books in 3 days, a productivity delta that legacy systems cannot match.
What to watch: Whether legacy ERP vendors can match the 4-week implementation timeline through their own AI-driven migration tooling, or whether the switching window remains open through 2026.
AI-Native Startups Are Winning Through Speed and Purpose-Built Data Models
The displacement isn't happening because AI-native ERPs are cheaper—it's happening because they were architected from the ground up to treat AI as a first-class citizen in the data model, not a bolt-on feature.
"Campfire introduced LAM (Large Accounting Model), a proprietary AI model trained exclusively on financial data achieving 95%+ accuracy on reconciliations and variance analysis. Automatically itemizes and reconciles complex variable costs (such as AWS cloud bills). Rillet integrates native data ingestion directly into a 'smart general ledger,' eliminating the need for separate spreadsheet workarounds or bolt-on analytics tools." — The Rise of AI-Native ERPs
The three dominant startups have collectively raised over $300 million: Campfire ($100M total), Rillet ($100M+), and DualEntry ($90M Series A alone). But the capital is flowing because the product differentiation is real. Campfire's LAM achieves 95%+ accuracy on reconciliations by being trained exclusively on financial data, not on general-purpose text. Rillet's smart general ledger eliminates the spreadsheet workarounds that have plagued finance teams for decades. DualEntry's Unified Ledger handles 40+ billion records with real-time reporting and an auditable control layer.
These aren't incremental improvements to existing systems. They're category rewrites. And the market is responding: Campfire reported 10x year-to-date revenue growth as of late 2025 and serves over 100 customers including high-profile tech companies. Rillet doubled its ARR in a 12-week period and signed over 200 customers. DualEntry achieved an outstanding ~50% customer referral rate—a signal that product-market fit is genuine.
What to watch: Whether these three startups can scale to enterprise customer bases without hitting the operational complexity wall that has historically limited venture-backed ERP vendors, or whether they become acquisition targets for larger platforms.
Legacy Incumbents Are Racing to Embed Agents, But the Damage Is Already Structural
SAP and Oracle NetSuite have launched comprehensive AI platform upgrades designed to automate away the manual work that made their systems vulnerable. But the defensive posture itself signals that the category is shifting from "how do we manage this business" to "how do we automate this business."
"SAP unveiled its strategy to transition customers to an 'Autonomous Enterprise' by combining its core business applications with sophisticated AI agent networks. The SAP Autonomous Suite embeds more than 50 domain-specific Joule Assistants that orchestrate over 200 specialized agents to automate processes from start-to-finish. For instance, the Autonomous Close Assistant automates journal entries and reconciliations to compress the financial close from weeks to days." — The Rise of AI-Native ERPs
SAP committed €100 million to partner incentives and introduced agent-led migration tooling designed to reduce ERP migration efforts by 35% through automated code remediation. Oracle NetSuite launched the Model Context Protocol (MCP)–based AI Connector Service and SuiteAgent frameworks, allowing custom autonomous agents to run directly on the SuiteCloud Platform. Both moves are credible technical responses.
But they're also admissions. The fact that legacy vendors need to spend $100M+ on partner incentives to keep customers from switching suggests that the technical superiority of the new systems is no longer deniable. SAP and NetSuite are fighting to make their existing install bases more valuable, not to win new greenfield customers. The consulting firms that have built entire business units around 18-month implementations are being asked to retrain on AI architecture and managed services—a transition that will take years and will compress margins for the entire SI ecosystem.
What to watch: Whether SAP's agent-led migration tooling can compress implementation timelines enough to close the gap with AI-native startups, or whether it becomes a tool for managed exit rather than customer retention.
The Consulting Model Is Being Hollowed Out in Real Time
As AI coding assistants and automated migration engines compress the billable hours required to implement an ERP, the traditional services-heavy model that has sustained system integrators for two decades is collapsing.
"As AI coding assistants compress the hours required for custom scripting and migration tools automate up to 35% of ERP migration effort, traditional IT consulting and system integration (SI) firms must pivot from billing hourly development labor to offering high-margin AI architecture, data governance, and process re-engineering." — The Rise of AI-Native ERPs
This is not a gradual transition. NetSuite analysts have explicitly flagged a "12-month partner window" leading into late 2026—a compressed runway for services firms to transition from traditional scripting to AI-readiness assessments and managed FP&A services. Firms that don't make that pivot will find themselves competing on hourly labor against AI coding assistants that cost a fraction of a consultant's billing rate.
The winners will be firms that can position themselves as transformation architects—firms that understand how to redesign financial processes around autonomous agents, not firms that can code faster. But that repositioning requires both cultural change and margin pressure that many mid-market SI firms will struggle to absorb.
What to watch: Whether traditional SI firms can successfully rebrand as AI transformation partners before their labor arbitrage disappears, or whether they become subcontractors to AI-native consulting firms built from the ground up to operate in this new model.
What Surprised Us
-
The velocity of capital flowing into AI-native ERPs is genuinely unprecedented. Campfire and Rillet each closed their Series B rounds within 10–12 weeks of their Series A, and DualEntry raised $90M in a single Series A. This isn't typical venture velocity—it's a signal that LPs believe the category is genuinely being rewritten, not just improved. The fact that Sequoia, Accel, a16z, and Lightspeed are all betting heavily on different horses in this race suggests they expect multiple winners, not a single consolidator.
-
Rillet's ability to double ARR in 12 weeks while also signing over 200 customers suggests that the sales motion for AI-native ERPs is fundamentally different from legacy vendors. These companies aren't selling to procurement committees; they're selling to finance teams that are desperate to reduce close cycles. The product is doing the selling.
-
SAP and NetSuite's defensive moves are credible but come too late. Both vendors are now racing to match the speed and autonomy of AI-native systems, but they're doing so while carrying the technical debt of 1990s-era database architectures. The 35% migration improvement SAP is promising is still not competitive with 48-hour migrations. This is a case where the incumbents may win the battle (keeping some customers) but lose the war (the category is being redefined without them).
Open Threads Worth a Vote
- What are the pricing and monetization models of AI-native ERP startups? — Since legacy ERPs rely on per-seat licensing and heavy implementation services, understanding how Campfire, Rillet, and DualEntry are pricing their platforms will be critical to assessing whether they can build durable, high-margin businesses or whether they're competing on volume with unsustainable unit economics.