ERP Implementation: A 7-Step Guide to Success for SMEs

An ERP implementation transforms processes, data flows, and workflows throughout the entire company. Errors in planning, data migration, or training therefore often lead to delays, additional costs, or acceptance problems within the organization.

Crucially, it's not just the selection of ERP software that matters, but a structured approach from requirements analysis to go-live—the productive launch of the new ERP system. This is precisely where many projects in medium-sized businesses fail.

This guide shows how an ERP implementation works in practice, which typical mistakes companies should avoid, and which seven steps are crucial for a stable ERP project.

Table of contents

• Why ERP projects succeed or fail
  
• Step 1: Clearly define project start and scope
  
• Step 2: Identify processes and requirements
  
• Step 3: Properly separate ERP selection and implementation
• Step 4: Design the solution and plan interfaces
• Step 5: Prepare data migration thoroughly
• Step 6: Organize change management, training, and testing
  
• Step 7: Plan go-live, rollout, and stabilization
• Costs, budget, and business case
• Conclusion
• FAQ on ERP implementation
  
  

Why ERP projects fail or succeed

An ERP implementation succeeds when goals, processes, data, and responsibilities are clearly defined before go-live and the project is consistently managed. Typical success factors include a robust scope, clean master data, realistic resource planning, integrated testing, and active change management (supporting employees in adopting new processes and ways of working). This reduces risks to operations, costs, and deadlines.

In SMEs, ERP projects rarely fail due to a single error, but usually due to unclear goals, poor data quality, a lack of prioritization, or departments being involved too late. Successful projects, on the other hand, rely on clear responsibilities, realistic planning, and standardized processes.

The benefits arise primarily from reliable data, fewer media breaks, and more transparent processes. Crucially, the implementation must deliver measurable improvements—such as shorter lead times, fewer manual corrections, or better inventory quality.

Studies on ERP projects repeatedly show that clear project management, management support, data quality, and sufficient resources are crucial for project success.

In practice, the following success factors and risks are particularly evident:

• Success drivers: clear scope, prioritized requirements, defined process responsibility, clean data, testable cutover plan
• Risk drivers: too many custom solutions, lack of key users, unclear interface responsibilities, insufficient testing, incomplete training
• Measurable objectives: order processing time, on-time delivery, inventory discrepancies, closing time, error rate in documents
  
  

What goals are realistic when implementing an ERP system?

Realistic goals are specific, measurable, and achievable within a project cycle. Target metrics along end-to-end processes, such as order-to-cash or procure-to-pay, are often suitable. These include clear responsibilities, defined master data rules, and a consistent document logic across all locations.

Unrealistic goals are those that mask structural issues, such as "all problems will disappear after go-live" or "every department will get its preferred logic." An ERP system (central software for inventory management, accounting, purchasing, sales, and processes) stabilizes processes but does not replace a lack of process discipline. Therefore, focus on a few, prioritized improvements that noticeably reduce operational workload.

Typical realistic and unrealistic expectations usually differ significantly:

• Realistic: Standardization of core processes, improved data quality, central inventory and order overview, faster reporting.
  
• Unrealistic: Complete customization without additional costs, simultaneous reorganization of all departments, go-live without temporary additional workload.
  
  

What are the most common errors in ERP projects?

Many risks arise early on: a constantly expanding scope, the late definition of master data rules, or a lack of available experts. Later, unclear integration boundaries and insufficient test coverage have an impact.

A common mistake is the transfer of poor master data. Duplicates, inconsistent article structures, or incorrect units later cause significant support and correction efforts during operation.

The following problems are particularly likely to lead to delays, additional costs, or unstable processes:

• Scope creep without a formal change management process
• Customizing as a substitute for process decisions
• Data migration without data owners and quality rules
• Tests without realistic use cases and without reconciliation reports
• Training that is too late or too generic, without role and process relevance
  
  

Typical stumbling blocks and countermeasures in SMEs

Step 1: Clearly define project start and scope

A clean project start reduces friction later on. Define which processes and locations will be included in the initial go-live, which systems will be replaced, and which integrations are mandatory. The scope is not just a list of modules, but a business-specific framework: which document types, which warehouses, which pricing logics, and which production methods.

Plan governance and decision-making processes so that business decisions can be made within a few days. In medium-sized businesses, speed is an advantage when roles are clearly defined: Steering Committee for priorities, Project Manager for schedule and budget, Process Owners for end-to-end decisions, and Key Users for day-to-day validation.

The following organizational foundations should be defined right from the start:

• Project Charter: Objectives, benefits, scope, assumptions, risks, budget
  
• Roles: Sponsor, Project Manager, Process Owners, IT, Key Users, Vendors
  
• Decisions: Change management procedures, escalation path, acceptance criteria for each milestone
  
  

Wie definiert man Scope und Rollen richtig?

Formulate 5 to 8 project goals as measurable results, not as activities. Complete a scope statement with clear in/out of scope points. In this way you avoid discussions about whether a topic should be added “quickly”. Also determine who is allowed to make process decisions, such as pricing logic, inventory management or production feedback.

Set up a change board that evaluates changes based on benefits, risks and effort. Every change has an impact on scheduling, costs and tests. This means that decisions remain comprehensible and the team can work in a focused manner.

• Goals as KPI: e.g. B. Completion time, inventory accuracy, order lead time
• Scope as a process map: Prioritize core processes and locations
• Governance: fixed decision-making meetings, documented decisions, acceptance logic
  
  

What content should be included in an ERP project plan?

A robust project plan defines not only deadlines but also deliverables, responsibilities, and acceptance criteria. Plan data migration—the transfer of existing customer, product, inventory, and transaction data—as a separate work package with test runs. Establish test phases with clear test case coverage and reconciliation reports. Include a cutover plan that outlines the transition steps and responsibilities until the first successful transaction.

Resource planning is critical: Key users require guaranteed time allocations. Without adequate support, delays will occur that can only be compensated for later through overtime. Also, plan for hypercare capacity for the first few weeks after go-live.

• Deliverables: Process map, role model, backlog/requirements specification, integration concept
  
• Tests: Test concept, test cases, UAT protocols, acceptance checklists
  
• Go-live: Cutover plan, support concept, hypercare plan, KPI tracking

Especially before a project starts, an external perspective helps to realistically assess the scope, roles, milestones, and risks. FIGULI CONSULTING supports SMEs in IT, ERP, and digitalization projects and helps them to prepare ERP projects in a structured way – from process mapping and project planning to seamless coordination with software and implementation partners.

Preparing an ERP project in a structured way



Step 2: Identify processes & requirements

Requirements engineering is the core of the process because it guides selection, design, data, and testing. The goal is not to document every exception, but to understand the value chain: where do waiting times, media breaks, rework, and a lack of transparency arise? Utilize workshops along end-to-end processes and supplement them with key performance indicators (KPIs) to objectively justify priorities.

A combination of process mapping (as-is), target state (to-be), and a prioritized backlog has proven effective. This keeps requirements testable and close to implementation. Ensure that business units don't just want "functions," but also define process rules, such as approvals, booking logic, and responsibilities.

How do you analyze processes before implementing an ERP system?

Start with a process map and identify the three to five most important core processes. For each process, document the individual steps, roles involved, systems used, relevant data, and typical exceptions.

It's important not only to capture the workflow but also to identify weaknesses—such as manual transfers, duplicate data entries, unclear approvals, or missing information.

In addition, define a few measurable key performance indicators (KPIs) for each process so that improvements remain traceable later. Typical KPIs include throughput times, error rates, inventory variances, or the proportion of manual entries.

For a thorough process analysis, the following points should be documented in particular:

• Current state: Processes, systems, data, roles, and exceptions
• Pain points: Waiting times, media breaks, queries, and sources of error
• Target state: Standard processes, responsibilities, approvals, and KPI targets

What requirements should be included in the specifications?

Formulate requirements in a way that makes them verifiable: inputs, rules, outputs, interfaces (interfaces to other systems, such as webshop, accounting, shipping, or time tracking), and permissions.

Distinguish between must-haves (go-live ready) and nice-to-haves (optimization after stabilization). Typical must-haves include correct posting logic, tax-compliant documents, master data fields, core reports, and the integrations that ensure smooth operation.

Use acceptance criteria to make tests unambiguous.

Example: "An order with a partial delivery automatically generates correct delivery notes and invoices, including backorder logic, and the postings match the reconciliation report."

The focus varies depending on the industry: In retail, variants, prices, and inventory processes are key; in manufacturing, bills of materials, material flow, and feedback are important; and for service providers, project, time, and billing logic are paramount.

Step 3: Properly separate ERP selection and implementation

ERP selection and implementation pursue different goals. While selection focuses on process fit, risks, operating concept, and budget, implementation centers on data migration, interfaces, testing, and go-live.

Mixing these two phases often results in projects that appear successful in demos but later struggle to replicate critical processes with significant effort.

• Selection: Process fit, vendor selection, and budget
  
• Implementation: Processes, data migration, interfaces, and go-live
  
• Goal: A solid basis for implementation decisions



What criteria really matter?

First, assess whether the ERP system accurately reflects your most important core processes. Then, examine scalability, interfaces, reporting, authorizations, and the appropriate operating model—such as cloud, on-premise, or hybrid.



If ERP systems are operated on-premise or in a hybrid environment, the underlying server infrastructure is also relevant. Learn more in the article "Building a Proxmox Cluster: High Availability for Your Server Infrastructure."

Industry-specific features are beneficial if they offer genuine process advantages, such as variant logic in retail or production control in manufacturing. Rare, special cases, however, should not be the primary criterion if they can be resolved more easily from an organizational perspective.

For many SMEs, 

 can also be a pragmatic option if inventory management, warehouse processes, and business operations need to be implemented quickly and with clear training. Especially for small and medium-sized enterprises, the advantage lies in the fact that setup, implementation, and training are often significantly streamlined compared to highly complex ERP systems. This often reduces effort, project duration, and initial costs – provided the functionality and processes are a good fit for the company.

FIGULI CONSULTING helps to assess this fit and implement blue office in a way that ensures processes, training, and ongoing operations work seamlessly together.

How do you compare providers in a structured way?

Don't use demos merely as product presentations, but rather as tests of defined use cases. Work with test data, roles, and expected results. An evaluation matrix helps to transparently compare process fit, effort, risks, and the operational concept.

Step 4: Design the solution and plan the interfaces

In many SMEs, the ERP system is not just a single piece of software, but the central component of the entire system landscape. Typical integrations include CRM, e-commerce, warehouse management systems, accounting, business intelligence, and industry-specific applications. Mobile data capture can also play a crucial role in warehouse processes. Learn more in the article "Mobile Data Capture with blue office: Optimizing Warehouse Processes."

Good solution design focuses on the target process, not existing departmental boundaries. This results in consistent data flows, fewer media breaks, and more stable daily operations.

To ensure integrations function reliably in the long term, the following points should be defined early on:

• Prioritize critical interfaces first
• Define data flows and responsibilities
• Plan monitoring and error handling early

What interfaces are typical for ERP systems?

Typical integrations include CRM systems for customer and sales data, e-commerce platforms for orders and prices, warehouse management systems for inventory movements, and financial systems for bookings, payments, and dunning. In production, MES, MES, or quality control modules are also added.

It is crucial to define clear leading data sources for each integration to prevent conflicting master data.

Furthermore, define early on how errors will be detected, reviewed, and corrected. Without clear processes, interfaces quickly become a persistent source of disruption.

How can media breaks be avoided in the ERP process?

Define clear status transitions, responsibilities, and booking logic for each end-to-end process. Without these rules, parallel processes often arise in Excel, email, or separate tools.

Clear approvals, traceable status changes, and consistent booking times are particularly important. This significantly reduces manual rework and data errors.

When is customization useful – and when is it not?

Customization is beneficial when it delivers clear business value and remains maintainable in the long term. It becomes problematic when custom logic merely reflects historical habits or complicates future updates.

Therefore, evaluate every deviation from the standard based on its benefits, risks, implementation effort, and long-term operating costs. Standardization should always take precedence over unnecessary custom logic.

• Customization should only be implemented with a clear rationale and defined objectives.
  
• Consider the subsequent costs for updates, testing, and support.
  
• Use standardization as the foundation for stable processes.

Step 5: Prepare data migration properly

Data migration is one of the most critical phases of an ERP implementation. Errors in master data, open items, or inventory are often only discovered shortly before go-live, causing delays, rework, or inaccurate reports.

Therefore, plan the migration not as a one-off export-import, but as a structured process with data cleansing, test runs, reconciliation reports, and clearly defined responsibilities. It is crucial to determine early on which data is truly needed and who will verify its accuracy.

In practice, the following points have proven particularly effective:

• Plan the migration with multiple test runs
• Thoroughly reconcile data inventories and open items
• Prepare the cutover with a timeline and clearly defined responsibilities

Which data should be migrated?griert werden?

Prioritize migrating data essential for day-to-day operations: customers, suppliers, items, prices, storage locations, accounts, cost centers, as well as open orders, open purchase orders, outstanding items, and relevant inventory.

Historical data should only be migrated if required for operational or legal purposes. Outdated, duplicate, or unclear data unnecessarily increases complexity and the risk of errors.

• Mandatory: Master data, open transactions, and inventory
  
• Optional: Limited history for analysis or documentation purposes
  
• Do not migrate: Duplicates, outdated items, or inactive records without value

How can data quality be improved before going live?

Start with simple rules: unique keys, mandatory fields, defined units, consistent naming conventions, and clear duplicate logic. Also, define responsible parties who are authorized to make business decisions, such as when merging customers or standardizing product attributes.

First, prioritize the data objects that are critical for core processes. For the initial go-live, a perfect data history isn't necessary, but rather a reliable minimum level of quality for customers, products, prices, inventory levels, and open transactions.

How do you test a data migration safely?

Document which data from the legacy system will be transferred to which fields in the new ERP system. Then, conduct several test migrations and systematically compare quantities, inventory levels, open items, and values.

A cutover plan is also crucial: It outlines the specific conversion steps, responsibilities, and checks required until the first successful posting in the new system.

• Document mapping and transformation rules.
• Conduct trial runs with fixed test data.
• Plan the cutover, including data freezes, import sequence, and validation.

For cost and efficiency analysis of migrations in medium-sized businesses, the University of Vienna study, "Cost-Benefit Analysis of an ERP Implementation" can provide additional guidance.

Step 6: Organize change, training, and testing

The technical implementation alone does not determine the success of an ERP rollout. Crucially, processes must be understood, accepted, and used effectively in daily operations. Therefore, change management, training, and testing should be planned early on.

Especially in SMEs, clear communication, short decision-making processes, and the active involvement of key users help reduce uncertainty and stabilize processes more quickly.

Testing forms the basis for a stable go-live. It helps identify errors in processes, permissions, and interfaces before the go-live – before they impact ongoing operations.

When implementing new ERP access points, roles, and end devices, workplace security should also be considered. More information can be found in the article "Endpoint Security."

The following points have proven particularly effective:

• Clear communication and clearly defined responsibilities
• Role-based training with practical examples
• End-to-end testing with documented results

How can change management be organized effectively?

Designate key users for each process or location to validate requirements, oversee testing, and serve as the first point of contact. Communicate regularly about goals, timelines, and the impact on daily work.

A structured approach to feedback is also crucial. Employees should be able to understand how feedback is evaluated and decisions are made.

FIGULI CONSULTING helps companies align change processes, project management, and communication in a practical way.

Which training formats work in SMEs?

Role- and process-based training with realistic case studies is particularly effective. A train-the-trainer approach, where key users later serve as internal contacts, often proves successful.

Short process instructions and exercises with real-world workflows significantly reduce follow-up questions after go-live.

Which tests are mandatory before going live?

Before going live, integration tests, user acceptance tests (UAT), and authorization tests should be conducted. These tests must examine not only individual functions but also complete process chains.

It is crucial that the results are documented and evaluated from a technical perspective. The key is not whether a process is "clickable," but whether bookings, inventory management, and reporting function correctly.

Step 7: Plan go-live, rollout and stabilization

Go-live is not just an IT event, but the transition to ongoing operations. Crucial factors include a tested cutover, clear responsibilities, and a support model that covers the initial weeks after launch.

For multiple locations or branches, a decision must also be made as to whether the ERP system will be implemented all at once or in stages. It's important to plan stabilization as a separate phase: Questions, minor errors, and adjustments will arise during the first few weeks.

After go-live, ERP quickly becomes an integral part of ongoing IT operations. The article "IT Maintenance for Companies: Why Regular Support Is Essential" explains why regular maintenance, monitoring, and clear support processes are critical.

For a stable launch, the following points should be prepared in advance:

• Go-live prerequisites: Acceptance testing, cutover testing, support readiness
• Rollout: Site sequence, training planning, authorizations
• Stabilization: Hypercare, error classification, KPI review

Big Bang or rollout in waves.

A big bang approach can be beneficial when processes are highly integrated and parallel systems pose significant risks, such as with centralized inventory management. However, it requires thorough testing, clean data, and a high level of training.

A phased rollout reduces risk because lessons learned from the first site can be incorporated into subsequent rollout phases. A hybrid approach is often also advisable: central functions and master data go live simultaneously, with sites following in phases.



How long does an ERP implementation take?

The duration depends primarily on process complexity, data quality, interfaces, and the availability of relevant departments. In many SMEs, a realistic timeframe for the initial go-live is approximately 6 to 12 months. With high integration density or production complexity, it can take longer.

The project can be accelerated through clear prioritization, early data trials, rapid decision-making, and implementation that adheres closely to standards.

• Drivers: Locations, interfaces, customization, data cleansing, testing scope
• Accelerators: Clear governance, early prototypes, processes that adhere to standards
• Dampers: Scope creep, lack of key user time, delayed interface clarification

What belongs in the hypercare phase?

Hypercare is the controlled stabilization process after go-live. Define support channels, responsibilities, ticket categories, and response times. Issues should be prioritized by severity: critical errors are resolved immediately, while optimizations are added to an improvement backlog.

Add KPI tracking to make stability measurable. Relevant metrics include throughput times, error rates, inventory discrepancies, completion times, and system availability.

Further technical best practices for high availability, cluster operation, and post-change stabilization are also described in the official Proxmox documentation.

• Support model: Key users, vendors, escalations
• Error management: Priorities, hotfixes, regression testing
• KPI review: Measurement points, target values, action plan

Costs, budget and business case of an ERP implementation

The costs of implementing an ERP system arise not only from the software itself, but primarily from implementation, integrations, data migration, training, and ongoing operations. For a realistic budget, all cost categories should be identified and prioritized early on.

A robust business case evaluates the benefits and costs based on concrete KPIs, such as time savings, reduced error rates, improved inventory management, or faster closing times. Scalability and reduced dependence on individual employees also play a role.

Flat-rate cost estimates are of limited use for ERP projects because effort and budget depend heavily on process complexity, data quality, interfaces, number of users, operating model, and desired customization. A more reliable approach is to calculate costs based on work packages, risks, and clearly defined go-live goals.

For companies in Austria, subsidies may also be relevant, depending on the timing and program. The KMU.DIGITAL funding program provides an overview.

Which cost categories are frequently underestimated?

Interfaces, data cleansing, and training are often underestimated. Data migration and integrations, in particular, often result in additional coordination and testing efforts.

Internal costs should also be considered: key users' time, project management, testing, and temporary overtime in daily operations.

• One-time costs: Implementation, migration, integrations, testing, training
• Ongoing costs: Operation, support, updates, monitoring
• Internal costs: Project time, decision-making, hypercare effort

How do you calculate a realistic ERP budget?

Create a budget structure based on work packages and define responsibilities, assumptions, and risks. A risk-based buffer, for example, for data quality or interfaces, is also advisable.

For many SMEs, a "Minimum Viable ERP" (MVER) is recommended for the initial go-live: core processes are stable, data is clean, and critical interfaces are functional. Extensions can follow later.

• Structure the budget according to work packages.
• Include risk buffers for data and integrations.
• Clearly separate must-haves from future optimizations.

Overview of typical cost components of an ERP implementation

Conclusion

An ERP implementation is successful when it's managed not just as a software project, but as a process, data, and change management project. For SMEs, a clear scope, prioritized requirements, clean master data, realistic testing, and a stable go-live are paramount.

These seven steps help identify risks early and make structured decisions. Thorough planning of data migration, interfaces, training, and hypercare significantly reduces rework, costs, and operational disruptions.

FIGULI CONSULTING supports SMEs in methodically structuring ERP projects—from process mapping and project planning to data migration and interfaces, all the way to testing, training, and go-live preparation. This results not just in a software project, but in an actionable roadmap for stable operations.

Plan your ERP project with FIGULI

FAQ on ERP Implementation

What are the most important ERP implementation steps for medium-sized businesses?

The most important steps are: project start and scope, process and requirements work, selection and implementation planning, solution design with integrations, data migration with test runs, change/training and testing as well as go-live with Hypercare. Decisive are acceptances and measurable results for each phase.

How does an ERP implementation proceed when business operations must continue?

Plan a phased rollout with early prototypes, repeatable data runs, and clearly defined testing phases. Establish binding cutover steps, data freezes, and support readiness. This will reduce risks to day-to-day operations, delivery capability, and sales during the transition.

What typical mistakes lead to additional costs in ERP projects?

Common causes include scope expansions without a change management process, excessive customization instead of standardization, underestimated data cleansing, and inadequate testing. A lack of key user time also leads to delayed corrections. These factors significantly increase implementation, testing, and support efforts.

How can data be migrated to an ERP system without extensive rework?

Designate data owners, define minimum quality standards, and conduct several test runs with reconciliation reports. Document mapping and transformations, and test the cutover as a dry run. This will allow you to identify errors before go-live and avoid corrections during production.

Big bang or rollout in waves: When is which type of introduction sensible?

Big Bang is suitable when processes are highly integrated and parallel system environments would lead to inconsistencies. A phased rollout is appropriate when locations operate differently or capacities are limited. Crucial factors are data consistency, seasonality, test readiness, and support capability.

What are the minimum tests that should be carried out before going live?

At a minimum, integration tests across all interfaces, user acceptance testing (UAT) with key users along real end-to-end use cases, and authorization tests for roles and permissions are required. For large data volumes, performance tests are also necessary. Documented acceptance procedures and reconciliation reports ensure the technical correctness of the implementation.

How long does an ERP implementation take in medium-sized businesses?

ERP implementation in medium-sized businesses typically takes 6 to 12 months. Key factors include process complexity, data quality, the number of interfaces, the degree of customization, and the availability of relevant departments. A clear scope and early data trials significantly shorten the project duration.