Day 1 SuiteWorld Keynote
Ready for Next: 
The ERP You Stop Logging Into

Is NetSuite ready for the AI era? Yes, and SuiteWorld 2026 showed why.

 

Its 28 years of transaction history across 44,000+ customers are what make its AI reliable. Evan Goldberg's keynote backed that claim with products already shipping. NetSuite Next is live in the US and Canada. Ask Oracle, AI portlets, SuiteSkills and the AI Connector Service come at no extra charge. More than 13,000 companies connect NetSuite to Claude, ChatGPT and other assistants, a year after that service launched.

​

The advantage won't go to companies that switch on the most AI features. It will go to companies whose data, policies and permissions are clean enough for agents to act on them safely. Gartner predicts that CFOs who deploy AI strategically will gain an additional 10 points of margin growth by 2029. Gartner also warns finance leaders: "They must not mistake activity for impact." Here is what each executive can do this quarter.

What changes for buyers?

 

In ERP, the database is becoming more valuable than the screens on top of it. Goldberg's own line: "AI sees a number. NetSuite gives it the context behind that number." When an ERP can be queried from any chatbot, buyers stop judging it on its menus. They judge whether its data, controls and business rules can be trusted when an agent acts on them.

NetSuite enters 2027 as the mid-market ERP best placed to turn AI into retention and expansion revenue. But it still has three things to prove:​

 

1. Thousands of Next adopters have to become tens of thousands.

2. Agent-driven work has to hold up under audit.

3. SuiteExpress has to make 60-day implementations normal, not a showcase anecdote.

What NetSuite announced, and what each piece is for

Goldberg organized the keynote around three claims: AI as the UI, ERP for you, and the autonomous suite. Taken together, they amount to one bet. NetSuite wants to stop being an app people log into and become the trusted record that every AI assistant, agent and dashboard draws from.

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

​

 

 

 

 

 

 

Quotes worth remembering

• "Reasoning without context is just guessing with confidence." Evan Goldberg, founder and EVP, Oracle NetSuite. This sums up the keynote's argument in one line.

• "AI sees a number. NetSuite gives it the context behind that number." Goldberg.

• "The expertise comes from NetSuite, not from the model." Goldberg, after the same receivables question returned $127.6K past due across 10 customers in three different AI tools.

• "Building forecasts like that used to take a data scientist. Now it takes a question." Goldberg.

• "Notice what the agent didn't do. It didn't change anything on its own." Goldberg, on the inventory exceptions agent.

• "When you buy a car, you don't pay extra for the engine, and AI is the engine of NetSuite Next." Goldberg, on pricing.

• "This is the end of ERP as we know it, and we're the ones ending it." Goldberg.

• "I didn't expect us to get there so quickly." Steve Hehl, COO, Worksmith. Worksmith budgeted six months for its implementation and went live in 60 days.

• "Give people an approachable path where you're not burning the lifeboats on the way in." Connor Wray, co-founder and CFO, JonnyPops. JonnyPops has cut its close nearly in half this year.

• "We need you to be sitting in the control tower, helping to direct traffic." Wray, on the new job of ERP administrators.

• "We leave all of that up to humans." Lee Rice Epstein, head of enterprise applications, Therabody, on who controls what the AI close checks act on.

Strategic implications for NetSuite and comparable ERP platforms

NetSuite's AI strategy accepts that it no longer owns the front door. The returns depend on whether it can own the facts behind every door. That choice affects every mid-market ERP vendor in five ways.

1. Does the ERP interface still matter?

Less than it used to. The trusted record behind the interface matters more. By shipping the AI Connector Service before its own assistant had fully matured, NetSuite lets employees work in Claude or ChatGPT and keeps the data, permissions and audit trail in NetSuite. Microsoft takes the opposite approach with Dynamics 365, routing users through Copilot and its own productivity suite. NetSuite is betting that buyers would rather pick their own assistant than have one chosen for them. Rivals face a stark question here. If any assistant can query your ERP, a polished UI no longer wins deals. What wins is a data model clean enough that an agent can act on it without making things up.

2. Why do SuiteSkills matter more than the chatbot?

SuiteSkills give AI answers that finance can sign off on. Goldberg's demo asked one receivables question in three different AI tools and got the same answer each time. That consistency is the real product. CFOs won't let an agent near the general ledger if it gives a different answer depending on which model is asked.

AI-native challengers like Rillet, which raised a $100M Series C at a $1B valuation in August 2026, start with a modern architecture. They start without decades of accumulated process knowledge. NetSuite has turned that history into a reusable asset. Every other incumbent now needs its own version, or it is asking customers to write the prompts that encode their business logic themselves.

3. What happens to implementation economics?

They change, and partners feel it first. Oracle CEO Mike Sicilia told investors in September that go-live times are "coming down from double digit months down to single digit weeks". Worksmith's 60-day go-live, against a six-month plan, is the proof point. SuiteExpress records discovery meetings, turns them into to-do lists and handles configuration, data migration and testing with agents.
For the partner channel, billable configuration hours shrink. Advisory work, agent design and data cleanup grow. For rivals whose value depends on long system-integrator projects, a 60-day comparison is hard to answer in a competitive bake-off.

4. What does "AI included" do to pricing?

It moves the fight from AI fees to adoption. When AI is bundled, NetSuite earns its return through retention and expansion: more modules, subsidiaries and users. Wray described the result plainly. More of his staff touch the ERP than ever, "and they have absolutely no idea that they are engaging with ERP." That creates an open question for investors. If more work happens in ChatGPT, does it still need a NetSuite seat? NetSuite has not said how licensing will evolve as agents and outside assistants do more of the work.

5. Where is NetSuite expanding its reach?

Both up and down. QuickStart is free and targets startups before they settle on an AI-native ledger. Industry agents push into specialist territory. Construction AIA pay applications, retainage tracking and continuous work-in-progress (WIP) reporting put NetSuite into deals that once went to construction-specific accounting packages.

How NetSuite stacks up

NetSuite's AI is ahead of its revenue. Its product story is the most complete in the mid-market. But Oracle's own CEO acknowledged in September that NetSuite is growing more slowly than the rest of Oracle's software business. Rivals that publish ERP-level figures are growing in the double digits.

 

 

 

 

 

 

 

 

 

 

 

Growth rates are each vendor's own reported measure, so they are directional rather than strictly comparable.

What can we conclude about NetSuite's position?

• Adoption is NetSuite's strongest early indicator. On September 9, Sicilia said more than 10,000 customers used the AI Connector Service. Goldberg cited more than 13,000 on October 26. That is roughly 3,000 new companies in seven weeks, about 30% of the installed base overall. Sicilia called it "one of the fastest adopted capabilities in the whole history of NetSuite."

• The product leads; revenue has to catch up. Sicilia blamed "slower decision cycles" in fiscal 2026. NetSuite Next became generally available in July. Its effect on new bookings will show up in Oracle's next two or three quarters, not this one.

• NetSuite is selling model neutrality harder than any major rival. Microsoft steers users to Copilot and SAP to Joule. NetSuite connects to Claude, ChatGPT and Grok today, and Gemini and Copilot are "coming soon." For buyers who have standardized on a non-Microsoft AI tool, that neutrality is a deciding factor.

• Sage Intacct is the rival to watch in the core mid-market. Its 26% US growth comes from the same CFO-led buyers NetSuite targets. Sage has published no Copilot usage figures, so this contest pits NetSuite's product depth against Sage's sales momentum.

• AI-native ledgers threaten new deals, not existing customers. Rillet's $1B valuation shows investors believe a fresh architecture can win new logos among startups. QuickStart, which is free, exists to keep those startups in NetSuite's pipeline.

What is NetSuite's growth trajectory?

Our estimate is that NetSuite is growing in the low double digits today, behind SAP and Sage Intacct. It has a credible path back to the mid-teens in 2027. Three signals support that path:

1. Its base of 44,000 customers is adopting AI faster than any feature in the product's history.
2. SuiteExpress shortens sales cycles by taking implementation risk out of the decision.
3. Bundled AI makes switching away cost more.

Upside comes from SuiteExpress expanding beyond software-industry customers and NetSuite Next going live in the UK and Australia. The downside is continued deal slippage while buyers wait to see whether AI-native ledgers mature.

Prognosis: what could go wrong, and what has to go right

NetSuite's design choices are sound. Its main risk is execution: getting announced features into customers' hands in time and making them hold up under scrutiny. The industry's track record is a warning. Gartner predicts that more than 40% of agentic AI projects will be canceled by the end of 2027 because of rising costs, unclear value or weak risk controls. In a March 2026 survey, 63% of finance organizations said AI implementation had been slower than expected.

Where NetSuite is leading

• Controls are part of the product. Every agent action in the demos waited for a human. The autonomous close logs who approved what and which agent did the work. When the close agent found that auto-approving low-risk journal entries could save 5%, it proposed a monitored pilot instead of changing policy. That design directly answers Gartner's "inadequate risk controls" concern.

• Adopting it is low-risk. Customers can switch NetSuite Next on role by role, user by user, with no reimplementation. JonnyPops ran Next in production while users could still switch back to the classic interface.

• Bundled AI removes the cost objection. CFOs don't need to justify a separate AI budget.

Where NetSuite is contending

• Forecasting is a trust test. In Gartner's same survey, finance leaders rated forecasting and insight generation among the weakest AI use cases. Ask Oracle predictions will be judged on accuracy over time. Showing a range of outcomes and the reasoning behind it is the right start.

• AI exposes messy data. Wray said JonnyPops is restructuring its general ledger (GL) segments so AI answers come out "more seamless and more accurate." Many customers will find their chart of accounts isn't ready for agents.

• Too much is still "coming soon." The Workspace Builder, Ask Oracle predictions, outbound connectors, and Copilot and Gemini support are all still on the roadmap. Goldberg's own line, "give us another 90 days," shows how fast the team is moving. It also means buyers are being asked to take parts of the roadmap on faith.

• Next is limited by geography. It is generally available only in the US and Canada. Multinationals with UK, EMEA or APAC subsidiaries will wait.

• Licensing is unclear. NetSuite hasn't said how seat counts will evolve as more work moves into outside AI tools.

Signals to watch

​

​

​

​

​

​

​

​

​

​

 

What's your edge?

For CEOs: 

1. Consolidate before you automate. The Data Intelligence demo succeeded because finance, campaign and policy data could be read together. AI agents built on fragmented systems inherit the fragmentation. Fund data consolidation ahead of new agent pilots.

2. Measure AI by the capacity it frees up. Every Man Jack's figures are a useful benchmark. Sicilia cited nearly 5,000 hours and $350K a year saved from the AI Connector Service alone. Set an hours-returned target and assign the freed capacity to a named growth initiative.

3. Build early wins into the rollout. Follow JonnyPops: switch Next on for one team, keep the classic interface as a fallback, and engineer an early "aha" moment. Adoption follows trust, and trust follows a visible result.

For CFOs:

1. Pilot the autonomous close on one subsidiary. Record your current days-to-close and exception counts first. JonnyPops cut its close nearly in half this year. Therabody uses AI close checks with Claude, sorted by subsidiary, and is aiming for its shortest close ever.

2. Clean up the chart of accounts before you scale. AI answers are only as clean as your GL segmentation. Treat restructuring it as AI infrastructure, not housekeeping.

3. Move policies out of PDFs and into workflows. The keynote showed an agent finding three orders that exceeded a 12% discount limit without approval. Describe your approval and discount rules once in SuiteFlow, and they're enforced on every transaction.

For CIOs: 

1. Choose the assistant, then lock down the roles. The AI Connector Service respects NetSuite permissions. That makes role design your real AI security control. Audit roles before connecting Claude, ChatGPT or Copilot.

2. Standardize on SuiteSkills for financial questions. Ad hoc prompts produce answers that vary from one model to the next. SuiteSkills produce consistent answers finance can rely on. Make them the default for anything that touches revenue, cash or close.

3. Run a registry of every agent. Use the SuiteAgent Management Workspace to record what each agent can see and do, its approval thresholds, and who owns it. That registry protects you from becoming one of the 40% of projects Gartner expects to be canceled.