2026 won’t be “business as usual.” AI, new operating models, and customer pushback are reshaping the enterprise stack faster than any time in the last 25 years. Here are the 10 shifts that matter, and practical moves to make now.
1) ERP’s monopoly is over
Why it’s happening:
- Modern businesses change faster than multi-year ERP roadmaps.
- Specialized apps (MES, WMS, CPQ, TMS, PLM) matured and now out-innovate “suite modules.”
- AI and integration tech reduce the penalty of running multiple systems.
Implications: Treat ERP as a transaction backbone, not the one platform to rule them all. Value comes from how quickly you can add/replace components, not how perfectly “single suite” your slide looks.
Watchouts: Governance and master data discipline must rise as you unbundle.
2) AI becomes the primary “system of interaction”
Why it’s happening:
- Users want answers, not menus. Conversational/agentic AI short-circuits report hunting.
- Retrieval, orchestration, and tool-use let AI traverse many apps and produce a single outcome (e.g., “Generate the variance analysis and open a corrective action”).
Implications: The UI your people touch daily will increasingly be an AI layer, while ERP/CRM/HCM recede into the background.
Watchouts: You need policy, auditability, and routing rules (who/what can act) or you’ll create shadow processes.
3) Composable ERP goes mainstream
Why it’s happening:
- Switching costs and business risk pushed companies away from “big bang.”
- Component choices (finance, SCM, manufacturing, service) can now be glued together reliably.
Implications: Architect around capabilities and events (order captured, work released, asset fails) and connect services to those events. Pick best-fit components per capability.
Watchouts: Without a canonical data model and API standards, composability turns into chaos.
4) Independent data & AI layers decouple from suites
Why it’s happening:
- Suites lag in cross-platform governance, lineage, and multi-source analytics.
- Neutral data layers (catalogs, quality, semantic models) and AI work better when not bound to one vendor’s stack.
Implications: Build a vendor-agnostic data foundation so AI can safely query, reason, and act across systems.
Watchouts: Decide data ownership early (who stewards product, customer, asset?). No ownership = no trust = no AI value.
5) Vendor lock-in backlash
Why it’s happening:
- Customers are wary of forced upgrades and shrinking optionality.
- Alternatives (point solutions + AI + stronger integration) restore negotiating leverage.
Implications: Demand portability: open APIs, data export guarantees, clear de-provisioning paths, and modular contracts.
Watchouts: Lock-in can hide in usage-based AI pricing, proprietary workflow builders, or private extensions.
6) Cloud gets pragmatic: hybrid wins
Why it’s happening:
- Not every workload benefits equally from SaaS.
- Data sovereignty, cost control, latency, and specialized hardware keep some estates on-prem/edge.
Implications: Classify workloads by sensitivity and differentiation. Put commodity processes in SaaS, keep crown-jewel logic/data where control matters.
Watchouts: Hybrid adds ops complexity. Invest in observability, identity, and network design up front.
7) Change management becomes a board metric
Why it’s happening:
- Failures are rarely technical; they’re adoption, roles, and incentives.
- AI amplifies impact on jobs and accountability, so leaders elevate OCM from “training” to operating-model change.
Implications: Budget and measure change like product work: readiness baselines, role maps, new skills, and incentive design.
Watchouts: “Train-the-trainer only” inflates go-live risk. Fund recurring enablement and performance support.
8) Legal, risk, and compliance expand (fast)
Why it’s happening:
- Breaches, privacy regimes, and AI IP/usage questions are mounting.
- Implementation disputes are up as projects grow more complex.
Implications: Involve legal/risk at design time: model provenance, data residency, acceptable use, incident playbooks, and shared liability with vendors/implementers.
Watchouts: If you can’t explain why a model acted (or who approved it), you can’t defend it.
9) Industry AI beats industry ERP (in many cases)
Why it’s happening:
- Targeted AI (forecasting, planning, maintenance, pricing, service) delivers impact faster than back-office rip-and-replace.
- Existing ERP can remain the system of record while AI becomes the system of decision.
Implications: Pilot AI where value is measurable (e.g., OTIF, MTTR, DSO, scrap rate). Keep scope tight, integrate outcomes back to systems of record.
Watchouts: Isolated AI wins that don’t write back into core workflows won’t scale.
10) Independent consulting & talent surge
Why it’s happening:
- Single-vendor implementers optimize for their product, not your operating model.
- Composable stacks, AI, and data require cross-disciplinary skills.
Implications: Build a blended bench (internal SMEs + independent process/data/AI/change specialists). Incent outcomes, not billable hours.
Watchouts: You still need a strong PMO and architecture authority or the mixed team fragments.
A pragmatic action plan (next 90–120 days)
- Phase-Zero refresh
Align business strategy ↔ tech strategy; define capability map, canonical data, and a 12–18-month roadmap. - Stand up an AI access layer pilot
One or two agentic use cases tied to P&L/working-capital KPIs; include approval/audit rails. - Harden data foundations
Assign data owners, define quality SLAs, implement lineage, and a shared semantic layer. - Contract for flexibility
Negotiate open APIs, export rights, kill-switches, modular scopes, and outcome-based milestones. - Fund adoption first
Role impact maps, new skills (prompt-to-procedure, exception handling), performance support, and revised incentives.
Bottom line
The center of gravity is moving from monolithic suites to capabilities orchestrated by data and AI. Winning teams will design for flexibility, protect leverage with vendors, and invest more in people than in logos on a slide.
Eric is recognized globally as a leading voice in digital transformation and ERP strategy. Over the past two decades, he has helped hundreds of organizations – including Nucor Steel, Fisher & Paykel Healthcare, Kodak, Coors, Boeing, and Duke Energy – define their technology roadmaps, modernize complex operations, and deliver real business value from large-scale transformation initiatives.
As Founder and CEO of Third Stage Consulting, Eric leads an independent, technology-agnostic advisory firm focused on helping clients navigate the shift from traditional ERP to more flexible, AI-enabled Digital Enterprise Operations (DEO) models. His work spans ERP selection, implementation quality assurance, organizational change, and operating model design across a wide range of industries and geographies.
Eric is also a prolific thought leader, known for his pragmatic takes on AI, cloud, and enterprise software trends, as well as his firm’s benchmark research and frameworks for de-risking transformation. He is dedicated to helping executive teams cut through vendor hype, make confident investment decisions, and successfully reach the “third stage” of their digital evolution.
