Manufacturing
Connect demand, materials,
capacity, production and cost.
ComplexityControl
Manufacturing performance depends on decisions that cross planning, procurement, inventory, warehouse, production, quality, maintenance, finance and customer demand. We help manufacturers connect those decisions through Microsoft business applications, data and operational intelligence.
Manufacturing complexity should be understood before it is configured.
In short
What a manufacturing platform is actually bought for.
Boards do not fund modules. They fund delivery promises that hold, working capital that behaves, margin that can be explained and fewer surprises late in the month.
Promises the business can keep
Delivery dates that reflect real material, capacity and supplier constraints rather than optimism.
Working capital under control
Inventory held because the plan needs it, not because nobody trusts the plan.
Margin that can be explained
Cost and variance traced back to the product, process and decision that created them.
Fewer surprises late
Shortages, capacity constraints and quality issues visible while there is still time to act.
Less manual effort
Planning, reporting and administration that no longer depend on a spreadsheet and one person.
A platform that can change
A supported operating core that can take on new sites, products and requirements without a rebuild.
Value chain
Every outcome depends on an operational condition.
An outcome without the condition behind it is a hope. This is the chain we work backwards along before agreeing what a programme should deliver.
On-time delivery
Depends on: Realistic planning parameters, reliable product structures, honest capacity and visible shortages.
Measured by: Schedule adherence and OTIF against your own baseline, not a benchmark.
Inventory and cash
Depends on: Owned planning policy, accurate stock, working replenishment and disciplined data maintenance.
Measured by: Inventory value, coverage and slow-moving exposure over time.
Margin control
Depends on: Costing model agreed, routings that reflect real work, scrap and variance recorded consistently.
Measured by: Production variance and product-level margin, reconciled to finance.
Operational visibility
Depends on: Production recorded at source, one reporting context, integrations that hold.
Measured by: Time to answer an operational question without rebuilding the data.
The manufacturing reality
The production order is only one part of the problem.
A single manufacturing decision can depend on demand, forecast, materials, supplier lead times, inventory, BOMs, routings, capacity, labour, machine availability, subcontracting, quality, warehouse, cost and delivery commitments. When those signals are disconnected, the plan becomes less reliable.
Each dependency is manageable on its own. The difficulty is that they move at the same time.
A late supplier affects production
One missed delivery moves work, capacity and promised dates that were already committed.
A production delay affects fulfilment
The customer experiences a planning problem as a service problem.
A planning assumption affects inventory
Lead times, safety stock and coverage decide how much cash sits on the floor.
Inventory affects cash
More stock is not automatically more resilience. It is often deferred visibility.
Warehouse congestion affects service
If material cannot be staged or finished goods cannot move, output does not reach the customer.
Quality affects rework and margin
A non-conformance consumes capacity that was already planned for something else.
Asset reliability affects the plan
An unplanned breakdown removes capacity that the schedule had already sold.
Manufacturing does not become complex because there are many screens. It becomes complex because every decision depends on another decision.
This is why manufacturing programmes more often struggle on data, planning assumptions and process ownership than on software capability.
One operating model
Demand becomes output through one connected chain.
This is the core flow. Quality, product data, capacity, warehouse, traceability and cost are not stages in it: they apply across all of it, which is why treating any of them as a single step tends to fail.
Demand
Understand what is real, what is forecast and what has changed.
Plan
Turn demand into supply, production and purchasing proposals.
Source
Place and manage supply against lead times and commitments.
Receive
Bring material into stock with the right tracking and inspection steps.
Produce
Issue material, record output and manage exceptions as they happen.
Fulfil
Move finished goods to the customer against the promised date.
Measure
Explain cost, variance and performance in the same context.
Applies across every stage
Quality
Applies at supplier, receipt, in process, at completion and after despatch.
Product data
Items, BOMs, routings and versions shape every stage of the chain.
Capacity
Constrains planning, execution and the promise made to the customer.
Warehouse
Decides whether material reaches production and output reaches the customer.
Traceability
Required across receipt, production, stock and despatch, not at one point.
Cost
Accumulates through every movement, not only at period end.
Where complexity appears
What ERP should actually control in a manufacturing business.
Open only the areas that matter to your operation. Each one explains the business decision it supports rather than listing functionality.
- Sales orders, forecasts, contracts and customer schedules
- Seasonality, promotions, projects and historical patterns
- What demand is real, and what demand is forecast?
- What has changed since the plan was agreed?
- Forecast accuracy is a business discipline. No system removes uncertainty from demand.
How you manufacture
The planning model changes with the way you manufacture.
Not every model is equally supported by every platform or configuration. Platform fit should be assessed against the models you genuinely run, including the ones that only apply to part of the business.
Not one operating model
Manufacturing is not one thing.
Discrete and process manufacturing ask different questions of planning, quality, traceability and scheduling. Treating them as the same problem is how platform decisions go wrong.
Discrete
- Bills of material and routings
- Component-level traceability
- Assembly, machining and finishing operations
- Scheduling around work centres and machine capacity
Process and batch
- Formulas, batch sizing and yield
- Co-products and by-products where applicable
- Deeper quality, traceability and regulatory requirements
- Platform fit must be assessed carefully. We will not overclaim Business Central capability here.
Quality
Quality is a discipline across the chain, not a module at the end.
If quality only exists as a final inspection step, the cost has already been incurred. It should appear wherever a decision can prevent rework.
Supplier
Approved suppliers, incoming specification and performance history.
Receipt
Inspection, quarantine, certificate capture and release to stock.
In process
Checks at operation level, hold, rework and scrap recording.
Completion
Final inspection, release, batch or serial record and certification.
After despatch
Customer complaint, return, investigation and corrective action.
Improvement
Trend analysis that connects the issue to the process that created it.
Where regulatory or customer requirements go beyond standard capability, specialist quality management may be needed. That is a design decision to take early, because it affects platform fit, integration and testing effort.
Maintenance and reliability
Capacity you cannot rely on is not capacity.
Maintenance is usually discussed after the plan is built. It belongs in the operating model, because downtime, spares and planned maintenance all consume the capacity the schedule has already committed.
Asset and criticality register
Which assets carry the plan, and what happens to output when one of them stops.
Planned maintenance
Maintenance windows treated as capacity, so the schedule reflects them rather than ignores them.
Breakdown and downtime
Reason recorded consistently, so downtime can be analysed rather than argued about.
Spares and inventory
Spare parts planned, held and consumed with the same discipline as production material.
Reliability improvement
Downtime, scrap and quality data used together to decide where reliability investment pays.
System boundary
ERP maintenance capability, a specialist CMMS or both. The boundary is a design decision with cost and integration consequences.
Application architecture
Decide what each system owns before choosing any of them.
Manufacturers rarely run one system. The expensive mistakes come from two systems owning the same data, or from an integration nobody scoped.
ERP
Plan, commit, transact and account
- Demand, planning, purchasing and production orders
- Inventory, costing, finance and customer commitments
- The single financial and commercial record of what happened
MES
Execute and control at the machine and operation
- Real-time execution, machine data, operator interaction and detailed shop-floor control
- Only required where the operation genuinely needs control below ERP granularity
- Where it is required, the ERP boundary and data flow must be designed, not improvised
PLM or CAD
Define and change the product
- Engineering design, drawings, revisions and change control
- The source of product structure where engineering owns it
- Release of approved structures into ERP is an integration with governance, not a copy
WMS
Move and control material
- Directed put-away, replenishment, picking and despatch
- Standard or advanced warehouse capability within ERP is sufficient for many manufacturers
- A separate WMS should be justified by warehouse complexity, not by habit
Boundary rules we apply
- Decide the boundary before selecting anything. Two systems doing the same job is the most expensive outcome.
- One master for each data object. Product structure, stock and cost each need a named owner.
- Every integration is scoped, designed, owned, tested and supported. None of them are free.
- Prefer fewer moving parts where the operation allows it. Complexity has a permanent running cost.
- Where a specialist system is genuinely needed, we will say so rather than stretch ERP into that role.
Platform fit
Do not choose ERP by company size alone.
Both platforms can run manufacturing. The decision should be made against site structure, planning depth, warehouse sophistication, quality and traceability requirements, governance and integration reality.
Business Central may fit when
- Manufacturing processes are relatively standard.
- Entity and site complexity is manageable.
- Warehouse complexity is moderate.
- The integration landscape is controlled.
- The organisation wants a connected operational core without unnecessary enterprise complexity.
Finance & Supply Chain may fit when
- Multiple entities and multi-site manufacturing.
- Advanced warehousing and more complex planning.
- Deeper manufacturing and supply-chain control.
- Enterprise security, governance and intercompany requirements.
- Global operations and a large integration landscape.
The order the decisions should be taken in
Operating model
Agree how the business should manufacture in future.
Manufacturing fit
Test that model against real product, planning and warehouse examples.
Architecture
Decide what ERP owns and what a specialist system owns.
Platform fit
Choose the smallest platform that supports the model responsibly.
Delivery route
Choose the route the situation supports, not the fastest one available.
Choose the smallest platform capable of supporting the future manufacturing model responsibly.
Delivery route
The right route depends on how much is still undecided.
Transform, Implementation, RAPID, Recovery, Support and Optimise are different answers to different starting points. Optimise is optional and only worth doing once the operation is stable.
Transform
When the future operating model is not yet clear
Planning model, production model, warehouse, product structure, master data, site and entity model, platform, integration and standardisation all still need decisions. Do not configure the old manufacturing model into the new ERP by default.
Explore TransformImplementation
When the model is agreed and delivery needs control
Scope control, master data, proof with real scenarios, readiness and cutover carry the risk. The methodology is the same discipline whether or not the timeline is accelerated.
Explore ImplementationRAPID 30
When Business Central scope is genuinely contained
Where manufacturing requirements are contained, standard Business Central fit is strong and readiness supports a controlled first release, RAPID 30 may be assessed. Not every Business Central manufacturing programme qualifies.
Explore RAPID 30RAPID 90
When enterprise scope can be bounded
Where enterprise manufacturing scope is bounded, Dynamics 365 fit is strong, governance is ready and complexity can be controlled, RAPID 90 may be assessed. Acceleration is never automatic.
Explore RAPID 90Recovery
When the programme is already in difficulty
Manufacturing ERP problems rarely stay inside one module. Planning, inventory, product data, production reporting, warehouse, migration, integration and costing tend to move together.
Explore RecoverySupport
When the operation depends on the platform every day
Once manufacturing runs on ERP, support is operational continuity. Planning, production, warehouse, inventory, finance, integration and reporting all need to work on a normal Monday morning.
Explore SupportOptimise
When the platform is stable and value is still available
Optional, and only worth doing against a baseline. Planning effort, warehouse travel, inventory policy, reporting and deferred automation are the usual candidates.
Explore Optimise
Confidence gates
Four points where manufacturing confidence is re-confirmed.
A gate is a decision point rather than a milestone to pass. Each one can confirm the plan, pause it, or reduce scope, and each is evidenced rather than asserted.
Gate 1
Before design
Manufacturing model confirmed
The future operating model is described in business terms, including the models that only apply to part of the business.
Confirmed at this gate
- Production models in scope, by product family and site
- Planning approach, capacity treatment and warehouse ambition
- Traceability, quality and regulatory requirements defined by the business
- Architecture boundary between ERP and any specialist system
Decision
Proceed to design with an agreed model · Continue qualification where the model is still contested · Assess a specialist requirement before committing to platform
Gate 2
Before build completion
Product and master data credible
Planning output can only be as good as the product structures, lead times and capacity data behind it.
Confirmed at this gate
- Items, BOMs, routings, units and versions reviewed against real products
- Lead times, safety stock, calendars and planning parameters owned by a named person
- Migration scope, cleansing responsibility and reconciliation approach agreed
- Costing setup agreed with finance rather than assumed
Decision
Proceed with data ownership in place · Hold where data ownership has not been accepted · Reduce first-release scope where data readiness is uneven
Gate 3
Before cutover decision
Operating proof demonstrated
The proof is whether representative work runs end to end, not whether configuration is complete.
Confirmed at this gate
- Demand to plan to purchase to receipt to production to despatch, run with real scenarios
- Shortage, exception, subcontract, rework and scrap paths tested
- Quality, traceability and warehouse steps demonstrated by the people who will do them
- Costing and period close reconciled against expected results
Decision
Confirm the cutover window · Re-test the specific scenario that failed · Move the date rather than carry unproven risk into live operation
Gate 4
After go-live
Operational stability and value
Stability is judged on the operation running, not on the number of tickets closed.
Confirmed at this gate
- Planning output trusted and used, with exceptions being worked
- Production, warehouse and quality recorded at source by the people doing the work
- Integrations, month end and reporting holding under normal volume
- Baseline measures captured so improvement can be evidenced later
Decision
Move into support and a measured improvement cycle · Extend hypercare where a specific area is not yet stable · Open a bounded assessment where the cause is unclear
Already live
Manufacturing ERP problems rarely stay inside one module.
If several of these are true, the issue is usually programme and data control rather than a single configuration fault.
- Planning output is not trusted
- Inventory is incorrect
- BOM or routing data is unreliable
- Production reporting is inconsistent
- Warehouse processes are blocking production
- Data migration issues remain unresolved
- Integrations are unstable
- Costing is not trusted
- Users are working outside ERP
- Go-live confidence is low
Live and stable
A live manufacturing ERP can still contain significant friction.
Stability is not the same as performance. Most manufacturers carry improvement opportunities that the first release could never have addressed.
- Manual planning effort
- Excess warehouse travel
- Reporting that needs rebuilding every month
- Repetitive administration
- Inventory policy that has never been revisited
- Limited production visibility
- Fragile integrations
- Data quality gaps
- Automation that was deferred at go-live
When manufacturing depends on ERP, support becomes operational continuity. Planning, production, warehouse, inventory, finance, integration, data and reporting all need to keep working on a normal Monday morning.
The new ERP should not become a museum of legacy workarounds.
Customise where manufacturing genuinely differentiates you: a unique production model, a real customer or industry requirement, an integration need or an operational differentiator. Not to recreate familiar screens.
Every custom process should answer one question: what business value does this difference create? If the answer is "we have always done it this way", that is not enough by itself.
Foundations
Manufacturing planning exposes weak master data quickly.
Data quality, product structure, migration, integration, extensions, site design, security and licensing decide whether the operating model holds together once volume arrives.
- Lead times, safety stock and calendars
- BOMs, routings, units and capacity
- Inventory, supplier data and product status
- Poor planning output is not always a planning-engine problem. Sometimes the inputs are wrong.
Shared responsibility
Some of this only you can do.
Manufacturing programmes succeed or fail on decisions and data that sit inside the business. Saying so at the start is more useful than discovering it during testing.
InteliSense
- Qualification, solution design and platform fit advice
- Configuration, extension, integration design and delivery governance
- Migration tooling, testing approach, training materials and cutover planning
- Honest advice where a requirement is outside standard capability
Your organisation
- Decisions on the future operating model and standardisation
- Product data ownership: BOMs, routings, lead times, capacity and costing inputs
- Named business owners for planning, production, warehouse, quality and finance
- People released to test with real scenarios and to attend training
- Acceptance of the cutover decision and of what the first release will not include
Shared
- Scope control and the consequences of change
- Risk, issue and dependency management
- Data quality remediation before and after go-live
- Measurement of the outcomes the programme was funded to deliver
Cutover in a plant that has to keep running
- Cutover sequence agreed for stock counts, open orders, work in progress and open purchasing.
- Production stop and restart plan agreed with operations, including the quiet window.
- Opening balances, valuation and reconciliation approach signed off with finance.
- Fallback position understood before the decision is taken, not invented during it.
- Support model on day one: who answers a shop-floor question at 6am.
Adoption on the shop floor
- Training built around the job people do, not around the module they are in.
- Shop-floor recording tested by shop-floor users, in the environment where they work.
- Planners involved in setting the parameters they will be accountable for.
- Adoption measured by whether transactions happen at source, not by attendance.
From manufacturing data to decision
The operation is already generating signals about what may happen next.
Demand, orders, inventory, lead times, production, capacity, scrap, downtime, supplier performance, warehouse activity, planning and cost data all describe the near future. Most of it is never used that way.
Process
Agree how manufacturing should operate before automating it.
Data
Make BOMs, routings, lead times and capacity trustworthy.
Control
Record production, quality and movement consistently at source.
Insight
Explain production, inventory, supply and cost in one context.
Automation
Remove manual steps where the value is clear.
Prediction
Apply models only where the data foundation genuinely supports them.
Ownership
Name who acts on the output, otherwise the model changes nothing.
Predictive intelligence
What would a manufacturing manager want to know earlier?
Prediction is only useful when it changes a decision while there is still time to act.
- Will a material shortage affect production?
- Where will capacity become constrained?
- Which orders are most at risk?
- Where is warehouse backlog building?
- Which supplier delays may affect production?
- Which inventory is creating avoidable exposure?
- Which planning assumptions are becoming unrealistic?
Level 1 Demonstration
Warehouse slotting advisor
Optimisation and recommendation
Reduce travel and congestion where material staging and finished goods movement constrain output.
Level 1 Demonstration
Warehouse labour and backlog forecast
Forecasting
See where backlog is likely to form before it affects production supply or despatch.
Level 1 Demonstration
Pick wave SLA risk predictor
Risk prediction
Identify fulfilment work most likely to miss the promised date while there is still time to act.
Level 1 Demonstration means we can show the capability and assess it against your data. It does not mean a deployed manufacturing product with published customer results, and we will not describe it as one.
Potential use cases, not products
- Material constraint risk
- Production schedule risk
- Supplier lead-time risk
- Capacity forecast
- Scrap and quality risk
- Downtime and reliability risk
- Inventory exposure
- Demand risk
These would be assessed against your data before anything is promised. We do not describe them as available manufacturing products, because they are not.
Explore Predictive IntelligencePower BI
See production, inventory, supply and finance in the same decision context.
Reporting should explain the operation rather than reconstruct it. We do not publish benchmark figures or invented customer metrics.
Typical dashboard themes
- Production status
- Plan versus actual
- Material shortage
- Capacity and load
- Inventory
- Warehouse
- Supplier performance
- Quality
- Downtime
- Cost and variance
- Order fulfilment
KPI categories worth agreeing early
- Schedule adherence
- Throughput
- Production variance
- Scrap
- Work in progress
- Inventory
- Supplier performance
- Downtime
- Backlog
- Fulfilment and OTIF
- Working capital
- Cost variance
These are categories, not benchmarks. Targets should come from your operation and be measured against your own baseline.
Evidence
A manufacturing customer, in their own words.
This is the manufacturing video evidence we can publish. Hill & Smith describe transforming their business practices in five months. That is their experience, not a result we promise to repeat.
Customer voice
Manufacturing customer evidence.
Manufacturing
Hill & Smith
We transformed our business practices in just 5 months
We hold approved customer logos and long-term Microsoft delivery experience alongside this. Where a claim cannot be evidenced, we do not make it.
See all customer storiesHow we work
Understand the operation before configuring the system.
Understand before configuring
Manufacturing complexity should be understood before it is built into a system.
Platform fit first
Choose the smallest platform capable of supporting the future manufacturing model responsibly.
Standardise deliberately
Every difference should earn its place with business value, not history.
Validate with real work
Test planning, production and warehouse processes with representative data and real users.
Data as a discipline
BOMs, routings and lead times decide whether the plan is believable.
Stay after go-live
Support, optimisation and intelligence continue once the platform is live.
Where to start
Tell us where you are, and we will tell you what the situation supports.
Four questions. The answers are carried into the conversation, so nothing needs repeating. We do not select a platform or promise acceleration from a form.
Indicative signal
Answer the questions and we will show a likely starting point. This is orientation rather than a recommendation, and we do not decide platform fit or acceleration from a form.
Your answers are carried through, so you will not be asked to repeat them. Final qualification is always a conversation.
Common questions
Questions manufacturers ask us.
Start with the operation
Understand the complexity before configuring it.
If planning, materials, capacity, production, quality, maintenance, warehouse or cost are harder to control than they should be, we can help you understand where the constraint actually sits and which platform and delivery route your situation genuinely supports.