ERP System Comparisons for Discrete Manufacturing: Build a Defensible Shortlist
Set the comparison boundary around assembled products
Discrete manufacturing assembles identifiable parts and individual units into finished products such as automotive components, aerospace equipment, industrial machinery, furniture, and electronics. Your ERP comparison therefore needs to follow the way your products are defined, planned, built, costed, inspected, and traced.
Process manufacturing uses a different production logic. Formulas, recipes, percentages, yields, and batches typically drive the work there, while discrete manufacturing centers on distinct components and units. That difference affects the records your ERP must handle and the questions you should ask during evaluation.
The manufacturing model affects requirements for:
- Bills of material that describe components and assembly relationships
- Routings that describe operations and their sequence
- Scheduling, capacity, and production planning
- Costing and inventory movement
- Quality records and traceability
- Engineering changes and revised product definitions
These requirements connect. An engineering change may alter a bill of material, a routing, inventory requirements, production cost, quality checks, or traceability records. A comparison that reviews each function in isolation can miss the handoff between them. Follow a representative product from its definition through planning, production, inspection, and completion.
A system can contain manufacturing functions without fitting your particular operating model. Feature count does not settle that question. A system that handles formulas well, for example, may still require a different test than one intended to manage engineered assemblies, configurable products, or repeated production orders.
Start by writing a short boundary statement for your project: your products are assembled units, your production model is discrete, and your comparison will focus on the records and decisions that model requires. Exclude food, beverage, pharmaceutical, and other process-industry requirements unless your own operation actually depends on them. Use that boundary to remove irrelevant demonstrations before you score any vendor.
For a manufacturer in aerospace, automotive, electronics, or medical devices, the boundary should also name the product and production characteristics that drive the evaluation. Include revision frequency, product variety, order pattern, traceability expectations, and the way engineering information reaches production. Those details give every later question a defined operating context.
Compare the product and edition, not just the vendor name
The ERP product and edition under review—not the vendor name alone—should appear at the top of every comparison record. Record the exact product, manufacturing edition or module scope, deployment model, version or release being proposed, and the organizational entities included in the offer.
An ERP comparison becomes unreliable when one vendor is represented by a broad platform description and another by a specific manufacturing edition. Your team may then compare a product family with a configured package, or a general supply-chain product with a manufacturing scope that includes different functions. Put the boundaries on the same page before discussing capability.
Use this product identity record for each candidate:
- Product and edition name
- Manufacturing functions included in the proposed scope
- Deployment model and hosting responsibility
- Version or release used in the demonstration and proposal
- Legal entities, facilities, warehouses, and users included
- Interfaces, modules, or extensions assumed by the proposal
- Functions that remain outside the stated scope
- Product or release dependencies attached to each material requirement
Keep the product identity record separate from the capability score. The identity record tells you what is being assessed; the score tells you how that defined scope performed against your requirements. If the product changes between the demonstration and the proposal, preserve both records and explain the difference rather than carrying the earlier result forward.
ERP deployment options vary. Some software is cloud or SaaS only, while other vendors offer cloud, on-premises, and hybrid deployments. Treat that distinction as part of the product definition rather than as a substitute for functional comparison. A cloud label does not tell you how your product records, shop-floor transactions, or existing systems will work together.
Ask each vendor to identify the exact product and edition behind every answer. If the answer applies only to another edition, release, module, or deployment model, move it into a separate note instead of scoring it as available in your project. The same rule applies when a capability depends on an add-on or a separate execution product: record the dependency beside the requirement, not in a footnote that disappears during ranking.
Use one operational scorecard for every candidate
The operational scorecard should translate your discrete-manufacturing model into the same questions for every candidate. The manufacturing model affects bills of material, routings, scheduling, costing, inventory, quality, engineering changes, and traceability, so those areas belong in the shared comparison rather than in separate vendor-specific notes.
An ERP requirements checklist should cover business, functional, technical, data, integration, reporting, security, and implementation needs. Use those categories to connect a production requirement with the surrounding business consequence. For example, a routing requirement may also affect capacity planning, operator reporting, costing, data ownership, and integration with a shop-floor system.
Build the scorecard in layers:
- Product definition: Test bills of material, revisions, configurations, engineering changes, and the relationship between design records and production records.
- Production planning: Test production orders, routings, operations, scheduling, capacity, substitutions, and changes to planned work.
- Materials and costing: Test material consumption, inventory transactions, backflushing where relevant, purchasing dependencies, and production cost records.
- Quality and traceability: Test the quality records you need, the links between material and finished units, nonconformance handling, and the level of traceability your customers require.
- Shop-floor interaction: Test how operators receive work, record progress, report quantities, and communicate production status.
- Business and technical fit: Test data migration, integrations, reporting, security, implementation responsibilities, and the internal ownership required after launch.
Give each requirement a priority, a test scenario, an expected record or output, and a status field. The status should distinguish demonstrated, configured, dependent on another product, proposed for customization, and unresolved. Use the same terms for every candidate. A shared scorecard defines your priorities before demonstrations and gives you a consistent way to compare vendors.
Add a consequence field beside each requirement. It should explain what decision the requirement supports: releasing a production order, calculating a cost, confirming material availability, proving a quality condition, or reporting progress. This keeps the scorecard tied to work your team performs instead of turning it into a list of software terms.
Do not force every requirement into a single numeric total. A mandatory traceability or integration condition may deserve a pass-or-fail boundary, while a preference about a report layout may be better recorded as a trade-off. The important point is that every candidate faces the same definition, priority, test, and status rule.

Compare the production record across each system
The production-record comparison should show how each product represents the information that moves from engineering into planning and then through production. The following side-by-side record uses only specific published descriptions; it is an illustration of comparison discipline, not a ranking of the two products.
| Shared criterion | Acumatica manufacturing page | Dynamics 365 Supply Chain Management description |
|---|---|---|
| Product definition and production connection | The page states that the system can create and maintain bills of material, routings, recipes, and formulas while connecting product design with production. | The description supports production orders, routings, and operations, but this passage does not establish the same product-design connection. |
| Production orders and material activity | The Acumatica description does not specify the full production-order and material-consumption scope. | The description says the system handles production orders, material consumption, and backflushing. |
| Scheduling and costing | The Acumatica description does not specify the full scheduling and production-costing scope. | The description says the system handles resource and capacity scheduling and production costing. |
| Shop-floor registration | The Acumatica description does not specify the operator-registration method. | The description says shop-floor registration can use job card device and terminal interfaces for reporting progress, time, and quantities. |
The table demonstrates why a product page and a demonstration answer different questions. A published capability statement can identify a function to investigate. It does not automatically show how the function behaves in your edition, configuration, workflow, or integration landscape.
Oracle’s manufacturing ERP guide identifies quality management, traceability, and CAD/CAM integration as relevant manufacturing capabilities. The guide also describes CAD/CAM integration as a way for engineering and production teams to work from the same design and measurements. Use those areas as test prompts, then ask each vendor to show the exact record, revision, handoff, and user action involved.
For every production-record test, follow the transaction across at least three points: the source record, the operational action, and the resulting record. If an engineering revision changes a bill of material, for example, ask how the revision is selected, which production orders use it, and where the change is visible to the people responsible for planning and production. If the vendor cannot show the complete path, record the missing handoff rather than awarding partial credit without a defined rule.
Do not award a point merely because a product page uses familiar terminology. Award the point only when the candidate’s defined scope and demonstration address the requirement you wrote. Keep a separate note for terminology that sounds similar but refers to a different record or process.
Make the ERP–shop-floor boundary explicit
The ERP–shop-floor boundary should be written into your comparison because production management and execution-layer requirements are not interchangeable. A description of Dynamics 365 Supply Chain Management and Business Central says neither provides real-time machine connectivity and OEE from live equipment signals, enforced digital work instructions with operator certification checks, or serialised as-built genealogy.
That description also identifies electronic batch records as outside the stated capabilities. Keep that point in its proper scope: it describes the cited products and passage, not every ERP product, edition, or possible configuration. Use it as a prompt for direct product-specific testing.
Those distinctions matter for an assembled-product operation that needs to connect planned work with what happened on the floor. Your team should decide which records belong in ERP, which belong in an MES or another execution system, and which system owns the handoff between them.
Add a boundary test for each requirement:
- Machine data: Does the requirement involve a live equipment signal, a manual entry, or a transaction passed from another system?
- Performance measurement: Does the requirement need OEE or another calculation from equipment data, or only production quantity and time reporting?
- Work instructions: Does the requirement call for a document display, a controlled instruction, or operator certification checks?
- As-built history: Do you need a serial-level record linking components, operations, and finished units?
- Ownership: Which system creates, changes, stores, and audits each record?
Use the published limitation as a prompt for direct confirmation rather than treating it as a universal statement about every possible configuration. Ask the vendor to identify the product, edition, add-on, interface, and system owner for each execution-layer requirement. Keep an unresolved boundary visible in the shortlist record; do not convert it into a functional score.
A useful boundary record names the event that crosses systems and the information that must survive the handoff. For example, an operation completion may need to return quantity, time, status, material information, and an exception. If your team cannot state what the receiving system needs, the integration test is not ready for scoring.
Test integration as two-way data flow
The integration comparison should follow information in both directions between enterprise planning and shop-floor operations. API integration can connect manufacturing software with an existing MES without requiring MES replacement when both systems support it, and it can create a two-way flow between planning and execution.
That conditional wording should shape your test. Do not ask only whether an API exists. Ask what data travels, which system owns it, when it moves, how failures appear, and how a corrected transaction returns to the other system.
Use an integration record with these fields:
- Business event, such as a released order, material issue, operation completion, or quantity correction
- Sending system and receiving system
- Data object and required fields
- Direction of flow and timing
- Trigger, interface method, and error handling
- Duplicate, late, rejected, or corrected transaction behavior
- Ownership of the resulting record
- Reporting and audit requirements
Give every candidate the same transaction scenarios. For example, ask how a released production order reaches the execution system, how material consumption returns to ERP, and how a completed operation changes the production record. Ask whether the demonstrated flow is native, configured, dependent on middleware, or dependent on a separate product.
An integration answer is incomplete if it covers only the successful path. Require the vendor to show a rejected transaction and the process for correcting it. Record any assumption that both systems must support a particular API or data structure. Advance the candidate only when the two-way flow, ownership, and exception path match your requirements.
Include the people and decisions attached to each exception. A rejected quantity, late confirmation, or duplicate transaction may require an operator, supervisor, planner, or system administrator to act. Ask who receives the exception, what record they change, and how the correction is reflected in both systems. This exposes operational responsibility that a simple interface diagram does not show.
Keep integration scope separate from the existence of a connector. A connector may move data while leaving ownership, timing, correction, and reporting questions unresolved. Score the complete business event against the same criteria for each candidate.
Use deployment as a filter, not the selection answer
The deployment comparison should filter candidates against your operating requirements without deciding the shortlist by itself. ERP options include cloud or SaaS-only software as well as products offered in cloud, on-premises, and hybrid forms. Those models change the questions you need to ask about access, responsibility, data location, integrations, upgrades, and continuity.
Record the deployment conditions that matter to your operation:
- Required access for plants, warehouses, offices, and remote users
- Responsibility for infrastructure, maintenance, upgrades, backups, and recovery
- Connectivity requirements for production areas and connected systems
- Data location, retention, export, and administrative access requirements
- Integration ownership and the way interfaces are maintained
- Product version and release cadence included in the proposal
- Operational consequences of an outage or delayed connection
A deployment preference may remove a candidate that cannot meet a hard requirement. It should not compensate for a gap in bills of material, routings, scheduling, costing, quality, traceability, or integration. Deployment options are one shortlist criterion, not the sole determinant of ERP selection.
Use two separate fields in the scorecard: deployment fit and operational fit. A candidate can satisfy your hosting condition while failing a production-record test. Another candidate can fit the production model while requiring a deployment arrangement your organization cannot support. Keeping those results separate makes the trade-off visible and prevents a familiar deployment label from carrying more weight than the operational requirement it is supposed to serve.
Ask the same continuity questions regardless of the preferred model. Identify what happens to production reporting when connectivity is interrupted, which transactions can be recorded later, and how corrections are reconciled. These are evaluation questions, not assumptions about any particular deployment type. Record the vendor’s answer against the exact product and edition proposed.
Do not let deployment terminology hide a change in scope. A cloud proposal, hosted arrangement, and on-premises proposal may place different responsibilities on your team even when they appear to support the same functional requirements. Keep those responsibilities beside the product identity and commercial records.
Normalize commercial comparisons before ranking candidates
The commercial comparison should place every proposal on the same cost and scope assumptions. A low license figure is not comparable with a broader proposal that includes more users, facilities, modules, implementation work, migration, interfaces, training, or support.
Your commercial record should show:
- License or subscription model and the user or access assumptions
- Products, editions, modules, and extensions included
- Implementation activities, roles, and estimated effort
- Data migration scope, cleansing responsibility, and conversion assumptions
- Infrastructure, hosting, maintenance, support, and upgrade responsibilities
- Integration, reporting, customization, and testing allowances
- Training, documentation, and post-launch support assumptions
- Contract term, renewal conditions, exclusions, and change-order treatment
If a proposal uses named or concurrent licensing, record the access pattern assumed by the quote. Licensing guidance says concurrent licensing can reduce license-count requirements by 20–40% for distributed workforces when users access the system at different times across time zones. Treat that figure as conditional guidance, not as a saving that applies to every organization. Your own user pattern and the vendor’s licensing rules determine whether the model is relevant.
Ask vendors to price the same scope and to identify every assumption behind the figure. Keep software, services, infrastructure, migration, and training visible as separate lines before you calculate a total comparison. If one proposal leaves an item unpriced, mark the item unresolved rather than treating the blank as zero.
Commercial normalization does not require a predicted outcome. It gives you a consistent way to see what each candidate includes, excludes, and expects your organization to supply.
Create a comparison period before reviewing totals. Separate one-time implementation and migration charges from recurring license, subscription, support, infrastructure, or maintenance charges, then identify the assumptions attached to each period. This does not produce a forecast by itself; it prevents a partial quote from being compared with a broader scope.
Keep the commercial record connected to the operational scorecard. If a mandatory integration, traceability requirement, or production function appears only as an optional module or service, mark that dependency beside the requirement. A candidate should not receive the same operational status as a product that includes the function in the defined scope.

Separate documented capability from evaluation evidence
Documented capability and demonstrated capability should have different labels in your evaluation record. A vendor page can describe a function, while a demonstration can show how that function works in the product, edition, configuration, and workflow proposed for your operation.
Use a controlled status vocabulary:
- Documented: A published product description states that the function exists.
- Demonstrated: The vendor shows the function using your agreed scenario and the proposed scope.
- Configured: The function works only after a stated configuration decision, and the configuration is recorded.
- Dependent: The function requires another product, interface, module, add-on, or deployment condition.
- Proposed change: The vendor identifies customization or development work rather than an existing function.
- Unresolved: The vendor has not shown or explained the requirement sufficiently to score it.
Acumatica’s manufacturing page describes creating and maintaining bills of material, routings, recipes, and formulas while connecting product design with production. Oracle’s manufacturing ERP guide identifies quality management, traceability, and CAD/CAM integration as relevant capabilities.
Those descriptions give you specific questions to test. They do not establish that every function is included in every edition, configured for your workflow, integrated with your systems, or suitable for your operating conditions. Keep the wording narrow. Record what is stated, what is shown, and what remains to be resolved in separate fields.
A comparison table should therefore include a claim, its product scope, the test scenario, the demonstration result, and the remaining condition. This structure lets you distinguish a documented statement from a result your team has actually observed.
Use the status label at the requirement level, not only at the vendor level. A product may be documented for production orders but unresolved for the way your team handles engineering revisions or shop-floor corrections. Requirement-level labels preserve those differences and make the final proposal review more precise.
When a vendor changes the answer during the evaluation, preserve the earlier statement and record the reason for the change. The change may reflect a different edition, an add-on, a configuration decision, or a correction to the original scope. Each possibility has a different consequence for the shortlist.
Keep trade-offs visible instead of naming one universal winner
The shortlist should show which conditions make each candidate worth further evaluation instead of declaring one universal winner. Manufacturing functionality alone does not establish fit for a specific manufacturing model, and deployment options should not be the sole determinant of selection.
A useful trade-off record connects each candidate to the requirements it appears to address and the conditions that still need proof. Use categories such as:
- Product-definition complexity and engineering-change needs
- Production planning, scheduling, and capacity requirements
- Material, inventory, costing, quality, and traceability needs
- ERP–MES or other integration boundaries
- Deployment and infrastructure constraints
- Commercial scope and ownership assumptions
- Unresolved requirements that could change the shortlist position
Avoid turning a feature count into a conclusion. A candidate with many listed functions may still require an extension, a separate execution product, or a particular edition for the requirement that matters most to your operation. A candidate with a narrower documented scope may remain relevant if it meets your highest-priority production scenarios with fewer unresolved dependencies.
Write the trade-off in conditional language tied to your requirements: a candidate remains in scope if it demonstrates the required routing and production-record behavior; a candidate needs further testing if the required shop-floor data depends on another system; a candidate leaves the shortlist if a mandatory deployment condition cannot be met. These are decision rules, not universal rankings.
Your shortlist is defensible when another project stakeholder can see why each candidate advanced, which requirement drove the decision, and what unanswered question could change it.
Make the decision rule visible beside the score. For each candidate, state the condition for advancement, the condition for hold, and the condition for removal. This is more useful than a single total when a mandatory requirement differs from a preference. It also gives operations, engineering, finance, and IT a common way to discuss the same result.
A trade-off is not a weakness that must be hidden. It is the relationship between a requirement and the conditions needed to meet it. Preserve that relationship until the product scope, demonstration result, integration responsibility, and commercial assumption are all clear.
Run equivalent demonstrations against the requirements
The vendor demonstration should use the same prioritized requirements and realistic discrete-manufacturing scenarios for every candidate. An ERP requirements checklist covers business, functional, technical, data, integration, reporting, security, and implementation needs, so the demonstration should include more than a tour of visible features.
Prepare a demonstration pack with:
- A representative assembled product with components, revisions, and an applicable bill of material
- A routing with operations, resources, timing, and a capacity constraint
- A production order that moves through material issue, progress reporting, completion, and costing
- An engineering change that affects a product definition or production record
- A quality or traceability event that follows the record to the required level
- An integration scenario involving a planning or shop-floor transaction
- A report or inquiry that an operations manager would use to make a decision
Give each vendor the same scenario, data assumptions, and time to prepare. Ask the vendor to show the user action, the resulting record, the control that prevents an incorrect transaction, and the report or downstream process that uses the result. If a vendor substitutes a slide, roadmap statement, or generic explanation for the requested scenario, record that as a different status from a demonstrated result.
We recommend evaluating critical requirements against a comparison table for electronics manufacturing operations. The same discipline applies across discrete manufacturing: connect every demonstration observation to a requirement and record whether the result was native, configured, dependent, customized, or unresolved.
The demonstration score should not stand alone. Keep the scenario, product edition, data assumptions, participants, and unresolved questions with the result so the score remains understandable after the meeting.
Assign an owner to each observation before the session ends. The owner should identify the follow-up question, the evidence required to close it, and the date or decision point by which the question matters. This prevents unresolved answers from being forgotten when the team moves from demonstrations to proposals.
Ask the same process question in every session: what happens next? A screen that shows a production order is only one part of the test. Follow the order into material activity, progress reporting, completion, costing, quality, traceability, or integration wherever those records matter to your operation. Use the same stopping point for every candidate.

Require proof beyond the demonstration
The proof review should test whether the demonstrated fit resembles the operating conditions your team will actually manage. We recommend requesting examples from organizations with similar products, order patterns, engineering practices, and operational complexity during ERP evaluations.
Similarity should be specific rather than based on industry labels alone. Compare the reference organization against your operation on:
- Product variety, configuration, and revision frequency
- Engineer-to-order, make-to-stock, make-to-order, or mixed order patterns
- Production volume, routing complexity, and capacity constraints
- Number of facilities, warehouses, entities, and production teams
- Quality, traceability, customer, and reporting requirements
- Existing MES, CAD, PLM, warehouse, or customer-order integrations
- Deployment model, data migration conditions, and internal system ownership
Ask the vendor to connect each reference to the requirements you are testing, rather than offering a general success story.
Use a structured reference discussion. Ask what the organization implemented, which edition and modules were involved, what remained outside the ERP, how integrations were handled, and which assumptions changed during the project. Ask how users record production activity, how engineering changes reach production, and how the organization handles exceptions.
A reference cannot replace your own demonstration. It adds operating context to the result and helps you identify questions that a scripted presentation may not answer. Record the conditions behind the reference before using it to support a shortlist decision.
Treat reference evidence as conditional context. A similar product or market does not make another organization identical to yours, and a positive account does not settle an unresolved requirement in your own scope. Use the conversation to identify what to test, what to price, and what responsibility your team would carry.
Request examples that match the work your project team is comparing, not only the industry name attached to the customer. An electronics manufacturer, aerospace supplier, automotive component producer, or medical-device manufacturer may still have different product structures, order patterns, engineering practices, and operational complexity. Record the similarity and the difference side by side.
Use final commercial proposals to test shortlist discipline
The final proposal review should keep multiple credible candidates in the commercial process long enough to compare scope and assumptions consistently. Licensing guidance recommends obtaining competitive quotes from at least three vendors through the final commercial proposal stage before accepting pricing.
Apply that recommendation to a controlled proposal process:
- Send each finalist the same requirements, scope boundary, user assumptions, facilities, interfaces, and implementation expectations.
- Require each proposal to identify included products, editions, modules, services, migration work, training, support, and exclusions.
- Map every priced line to a requirement or project activity.
- Separate recurring charges from one-time charges and identify the period covered.
- Record assumptions about data quality, internal staffing, customization, integrations, and change requests.
- Compare unresolved requirements before comparing totals.
Do not treat three proposals as comparable merely because they contain three totals. A proposal with a broader scope may cost more because it includes work another vendor left unpriced. A lower figure may depend on assumptions that your team cannot meet. Return every proposal to the same scope record and ask the vendor to explain any difference.
The commercial stage is also a discipline check. If a candidate cannot describe the product edition, production scope, integration responsibility, or implementation assumptions in the proposal, move that uncertainty back into the evaluation record. Do not let a commercial deadline erase an unresolved operational requirement.
Before accepting a price, reconcile the proposal against the demonstration record. Confirm that the priced product and edition match the one shown, that dependent modules or interfaces are included where required, and that proposed changes are identified as such. If the commercial document changes the tested scope, reopen the affected requirement rather than treating the earlier demonstration as proof.
Keep the proposal comparison readable to people outside the selection team. A finance reviewer should be able to see what is recurring and one-time; an operations reviewer should be able to see what production and integration scope is included; an IT reviewer should be able to see the interfaces and responsibilities. A common scope record supports all three reviews.
State what the comparison cannot establish
The comparison cannot establish that a system will fit your operation until the proposed product, edition, configuration, integrations, commercial scope, and operating assumptions have been tested together. A general product description cannot settle a requirement that depends on a particular release, module, add-on, interface, user role, facility, or workflow.
Keep these boundaries visible:
- A feature description does not establish how the function behaves in your configured process.
- A production demonstration does not establish the full implementation effort or migration condition.
- A cloud, SaaS, on-premises, or hybrid label does not establish operational fit.
- A successful interface scenario does not establish every exception, correction, or ownership rule.
- A case study or reference does not establish that another organization’s products, order patterns, or engineering practices match yours.
- A proposal total does not establish total commercial scope when assumptions or exclusions remain unresolved.
- A documented limitation for one product or description should not be generalized to every edition or deployment without direct confirmation.
Mark each unresolved item with an owner, a question, and a decision point. Ask the vendor for the exact product and edition involved, the condition under which the function works, and the responsibility assigned to your team. If the answer depends on another system or a future change, keep the dependency in the shortlist record.
An evidence gap is a reason to request a focused test, not a reason to award a positive or negative score. Remove a candidate only when a stated requirement fails, a required condition cannot be met, or the remaining uncertainty is unacceptable for your project.
This boundary also applies to outcomes. A demonstration, reference, or proposal can support a decision about the next evaluation step; it does not establish a particular implementation result for your operation. Keep the conclusion tied to the evidence actually recorded and state the remaining condition plainly.
When the team reaches a decision point, use the unresolved record to choose the next proof activity: a product-specific demonstration, an integration test, a commercial clarification, or a reference conversation. Do not use an unsupported assumption to fill the gap simply because the shortlist must be narrowed.
Compare the production record before advancing a candidate
Compare each candidate through the production record it would manage rather than treating the vendor name as the unit of comparison. Start with one representative assembled product—such as an aerospace component, automotive part, electronic product, or medical device—and identify the exact product, edition, deployment option, and proposed scope under review.
Trace the same scenario through:
- Product definition, bills of material, routings, and the connection between product design and production
- Resource and capacity scheduling, production costing, and shop-floor registration
- Quality management, traceability, and CAD/CAM integration where those requirements apply
- The boundary between ERP production management and execution-layer needs such as live machine connectivity, OEE, digital work instructions, serialised as-built genealogy, and electronic batch records
- API exchange with an existing MES, treating two-way information flow as conditional on support from both systems
Keep each result tied to the product and edition being considered. A vendor-published description is not the same as a demonstrated result. Record whether each requirement is documented, demonstrated, configured, dependent on another system, proposed for change, or unresolved, and ask the vendor to show the relevant behavior in the stated scope.
The distinction between ERP production features and execution-layer capabilities should remain visible. The same cited description says that Dynamics 365 Supply Chain Management and Business Central do not provide certain execution-layer capabilities, including real-time machine connectivity and OEE from live equipment signals, enforced digital work instructions with operator certification checks, serialised as-built genealogy, or electronic batch records.
Use the resulting production record to identify which candidates can account for the operational scenario and which leave dependencies or evidence gaps. Deployment and commercial terms can filter the comparison, but they do not resolve an unproven production or integration requirement.
Keep reading
Related articles
Free account
Create A Free Account
Get the clarity you need to make confident ERP decisions — without the sales pressure.
- Read every article (past & future)
- Compare ERP systems side by side
- Access exclusive guides, checklists, and cost breakdowns
- Stay current on industry updates and trends
Get expert guidance at no cost
Rather talk to a specialist?
Book a free call with one of our ERP specialists for a no-obligation needs assessment. We'll help you understand your options, cut through the noise, and figure out what actually makes sense for your business — before you ever talk to a vendor.
Book A Free Call