logo

Are you need IT Support Engineer? Free Consultant

LO-VC or AVC? How to Think About Variant Configuration in SAP S/4HANA

  • By Sanjay
  • 05/11/2026
  • 32 Views



The Meeting Where Everyone Pretends to Agree

Picture a room where the project lead is gesturing at a migration timeline, the configurator in the corner is quietly calculating how long it would take to rewrite three thousand dependencies, and the IT manager is nodding at a slide that reads “AVC is SAP's strategic direction” while privately thinking: strategic for whom, exactly, and by when? If you have sat in that room — or run that meeting — this post is for you.

The question of LO-VC versus AVC is one of the most practically important questions in SAP variant configuration right now, and it is also one of the most tangled. Let us untangle it, without the bullet points.

Two Engines, One Job (Sort Of)

Think of LO-VC and AVC the way you might think of two professional chefs. The first trained in classical French technique, has been in the kitchen for thirty years, and can cook anything — but insists on doing things the old way and is not picking up any new recipes. The second trained on modern cuisine, works with precision instruments, and has an open calendar for whatever comes next. Both can feed you dinner. Only one of them is still learning new dishes.

LO-VC — Logistics Variant Configuration — is the legacy engine that has powered product configuration in SAP ERP since the 1990s. It sits deep in the ABAP stack, is tightly woven into business processes, and has been battle-tested across practically every manufacturing industry imaginable — from automotive and industrial machinery to high-tech electronics and engineered-to-order equipment. It is mature, stable, and still fully supported.

AVC — Advanced Variant Configuration — is the next-generation successor. It uses a constraint-based solver called Gecode — developed under the lead of Prof. Christian Schulte at KTH Royal Institute of Technology in Stockholm and extended by SAP in cooperation with the Fraunhofer Institute for Industrial Mathematics (ITWM) in Kaiserslautern, Germany. Gecode is an award-winning solver, and the performance difference is not marginal: AVC resolves complex configuration models significantly faster than the classic engine, with the advantage growing as model complexity increases.

It ships with modern Fiori applications, a dedicated simulation environment where you can test any configuration scenario before it touches a live order, embedded analytics at the characteristic level, and deep integration with SAP Commerce Cloud and SAP CPQ.

_Lovc_Avc_Img1_Lovc_Ui_Comparison.png

The Roadmap Message SAP Actually Means (Yes, Really)

There is one sentence from Dr. Petra Meyer, AVC Product Manager at SAP, that keeps appearing in every internal and customer-facing conversation about this topic: “No new functionality in LO-VC. The roadmap is empty. AVC is SAP's strategic solution.” That is not corporate hedging — it is a statement of engineering intent. LO-VC will receive maintenance and security corrections. The feature set, however, is frozen permanently, by design.

This matters in ways that go beyond a slide. LO-VC is not available in S/4HANA Public Cloud Edition, full stop. If your organisation is on a path toward public cloud, LO-VC is simply not part of that future. AVC is available in both private and public cloud editions and receives new capabilities with every release — Engineer-to-Order enhancements, Lot Size of 1 manufacturing scenarios, OData V4 APIs for sales order processing, support for complex multi-level solution quotations, and connectivity to SAP Commerce Cloud and SAP CPQ are all on active delivery. In Q1 2025, SAP also simplified the AVC licensing model to a single SKU covering all users — no separate tiers, no additional licences needed for engineers building models. The SAP Road Map Explorer entry for AVC is expanding; the one for LO-VC is, as promised, empty. SAP Note 3073312 contains the authoritative feature-by-feature comparison if you want the forensic detail.

So When Does It Make Sense to Stay on LO-VC (For Now)?

Honesty requires acknowledging that migrating from LO-VC to AVC is not a weekend project. The two engines use genuinely different modelling paradigms. Where LO-VC relies on procedural preconditions and procedures — step-by-step instructions the system follows — AVC uses declarative constraints, which describe what must be true without prescribing how to get there. For a model with thousands of dependencies written in LO-VC syntax, that is not a copy-and-paste migration. It is a re-design exercise.

If your LO-VC landscape is stable, your team is not cloud-bound yet, and you have no immediate pressure from a public cloud adoption programme, there is a reasonable argument for continuing to run LO-VC while planning the migration thoughtfully. “Thoughtfully” is the operative word. The window for that thoughtful planning is narrowing, not widening.

Why the Gap Is Bigger Than It Looks

The gaps show up in three places.

For end users, AVC's constraint engine — powered by Gecode — aggressively eliminates infeasible options as you configure. In LO-VC, a misconfigured dependency can silently allow combinations that only fail downstream — at production order creation, at procurement, or at delivery. AVC eliminates that class of error by design: invalid combinations are excluded as early as possible, so they rarely appear as choices at all. When a sales rep needs to change a value that has already been set, AVC lets them select an alternative without deleting the existing valuation. Filtering, undo, redo, and colour-coded tab guidance showing exactly where input is still needed round out a user experience that LO-VC's SAP GUI simply cannot match.

Before any of this reaches a live order, AVC also provides a dedicated Simulate Configuration Models app: a Fiori environment where you can test any product configuration scenario in a completely risk-free way. Run a full configuration, activate the dependency trace to see exactly which rules fired and why, inspect the resulting BOM, and understand why certain options appeared or disappeared — all without touching live data. This is as useful for sales training and customer demonstrations as it is for model validation and development.

_Lovc_Avc_Img2_Avc_Inference_Differentiator.png

For the engineers who build and maintain these models, AVC removes entire categories of workaround. Tasks that required multi-step helper dependencies in LO-VC — type conversions, string handling, complex table lookups — are handled concisely in AVC constraint syntax. Variant tables in AVC are more expressive too: they support multiple values per cell, intervals, wildcards, and rules that define forbidden combinations — capabilities that in LO-VC require workaround structures that bloat models and slow them down. When something is not behaving as expected, AVC's tracing environment lets you step through individual rule executions, filter by characteristic, and see exactly which inputs produced which outputs — turning what was a half-day debugging exercise into a fifteen-minute one.

_Lovc_Avc_Img3_Avc_Trace_Differentiator.png

For the business, the advantages show up where they matter most: order accuracy and speed.

AVC is the only engine that supports switching from Configure-to-Order to Engineer-to-Order mid-process. If a customer's requirements go beyond what standard options can cover, the salesperson switches to Engineer-to-Order mode — and every value they have already entered is preserved. This matters most for manufacturers who build products to a single customer's specification with no stock held: the “lot size of 1” scenario that industrial machinery, specialised equipment, and custom engineering companies deal with every day. The AVC model starts with a super BOM containing every possible component. As the configuration is completed, the engine selects the right parts, eliminates the incompatible ones, and delivers a complete, accurate, ready-to-manufacture bill of materials for that one customer's product. No manual rework, no post-order corrections.

Beyond individual order accuracy, AVC's embedded analytics give you business intelligence that LO-VC simply cannot provide: which characteristic combinations appear most often in orders, which configurations drive the most revenue, where customers consistently hit dead ends in the configuration process. These insights live at the characteristic level, flow into SAP analytics without custom development, and give your product management and sales teams the data to make informed decisions about your variant portfolio.

What Migration Actually Involves (No Glossing Over)

_Lovc_Avc_Img4_Transition_Workbench_Migration.png

SAP provides dedicated tooling to make the transition from LO-VC to AVC manageable. The starting point is Transition Workspaces (transaction VCH_L2A_WORKSPACES), where you group your existing classic configuration models into packages. Each workspace is defined by a configurable product and one of its configuration profiles, and any number of models can be grouped together for collective processing — giving you an aggregated view of transition progress across related models.

The actual work happens in the Transition Workbench (transaction VCH_L2A_WORKBENCH). The Workbench assesses each model for AVC compatibility, converts the dependencies it can handle automatically, flags what still needs manual attention, and lets you simulate the transition steps before committing anything to the live system. Once models have been adapted, the Compare Engines tool (VCH_L2A_COMPARISON) runs a side-by-side comparison of LO-VC and AVC results on the same order or material variant — so you can confirm both engines produce equivalent output before you cut over.

LO-VC and AVC coexist in the same system during migration — you migrate model by model, validate each one, and carry the landscape forward incrementally. What migration cannot do is eliminate the re-design work for complex dependency logic. For large, mature landscapes — such as those in industrial machinery where manufacturers have been running LO-VC since ECC days — the translation effort is the honest centre of the migration cost estimate. SAP's tooling has matured significantly, and partial automation is now possible. But sizing the remaining manual work accurately, early, is what separates a migration that lands on time from one that keeps getting pushed.

The Integration Story and Why It Matters

Once you are on AVC, the integration picture changes fundamentally. AVC integrates with SAP Commerce Cloud and SAP CPQ — what SAP describes as the ability to “model once and configure anywhere.” The mechanism is a Knowledge Base: a packaged runtime copy of your product configuration rules that is replicated to external systems. A B2B or B2C customer configuring a product on your Commerce Cloud storefront works from exactly the same constraint rules as your internal sales team in S/4HANA. The same options are available, the same rules apply, and the resulting configuration flows back to create a correctly configured sales document. No duplicate model maintenance, no synchronisation overhead.

Sales Order processing is also exposed via OData V4 API, enabling external integrations, custom applications, and automation workflows to interact with AVC-configured orders in a standardised, modern way.

One practical note for organisations already on LO-VC and using SAP Commerce Cloud or SAP CPQ — or planning to: certain classic LO-VC variant table constructs are explicitly documented by SAP as incompatible with those external platforms. AVC variant tables do not have this limitation. Migration may already be relevant to integration challenges your team is dealing with today, not just a future-state consideration.

None of these integrations exist for LO-VC. If your business needs to configure products through a web channel or connect to an external quoting tool, AVC is not just the better option — it is the only one on SAP's roadmap.

A Practical Thing You Can Do This Week

If you are still on LO-VC and the migration conversation keeps getting deferred, there are two useful starting points — depending on whether you want to understand what you are migrating toward, or what it will cost.

To understand what AVC looks like in practice: open the Simulate Configuration Models app on any AVC-mode product — or a simple test material — and run some configurations. Watch infeasible options disappear in real time as you make selections. Activate the dependency trace and see exactly which rules are firing. Inspect the resulting BOM. It usually takes less than an hour to shift from “I have read about AVC” to “I understand why this matters.”

To understand what migration will actually cost: use the Transition Workbench (VCH_L2A_WORKBENCH). Create a workspace in VCH_L2A_WORKSPACES for one of your simpler configuration models and run the compatibility check. The Workbench produces a clear picture of what converts automatically and what needs manual re-writing — transforming migration from an abstract risk into a bounded workplan. That assessment usually takes less than a day.

If you want the full picture, the SAP Road Map Explorer at roadmaps.sap.com shows the AVC delivery pipeline through 2027 and beyond. If you are earlier in your thinking and want to understand what AVC looks like in practice, the first two posts in this series cover the fundamentals: what variant configuration is and why it matters, and what AVC's specific capabilities look like in SAP S/4HANA. Start there, come back here, and then pick up the Transition Workbench.





Source link

Leave a Reply

Your email address will not be published. Required fields are marked *

Chat with us on WhatsApp!