How to Choose an ERP System: The Complete Selection Framework [2026]

Why ERP Selection Matters More Than Implementation

Most companies spend 70-80% of ERP project budgets and energy on implementation. That's a mistake. The real risk starts before implementation — during selection.

A bad ERP selection decision doesn't just delay your go-live; it locks you in for 8-12 years. Once you've spent 6-12 months implementing, migrated all your data, trained your workforce, and embedded processes into the system, you're psychologically and financially committed to that platform. Switching costs $500K-$2M+ and another 6-12 months of disruption.

The companies that avoid painful ERP implementations don't necessarily have perfect ERP projects — they choose the right ERP in the first place.

The selection phase is where you prevent 80% of ERP project failures. A poor selection choice leads to poor fit, which leads to poor adoption, which leads to poor results. And that's unavoidable through implementation alone.

This guide walks you through a proven 7-step selection framework used by companies that get ERP implementations right from the start.

Step 1: Define Your Business Requirements

Requirements gathering is the most critical phase of ERP selection, yet it's also the most rushed. Many companies skip straight to vendor shortlists and demos, treating requirements as something you figure out later. That's backwards.

Requirements gathering must consume 30-40% of your total selection timeline. If you're planning a 3-month selection, spend 4-6 weeks on requirements. If you're planning 6 months, spend 8-10 weeks.

How to run effective requirements gathering:

• Assemble a cross-functional team. You need Finance, Operations, Supply Chain, Manufacturing (if applicable), HR/Payroll, IT, and at least one C-level executive. Each function sees different needs; one group's requirement is another group's "nice-to-have."
• Distinguish must-haves from nice-to-haves. Create two lists. Must-haves are deal-breakers; if an ERP doesn't have it, you eliminate them from consideration. Nice-to-haves are differentiators; they matter for shortlist ranking but don't eliminate vendors. Most companies find 5-8 true must-haves; everything else is nice-to-have.
• Involve end users, not just department heads. Process owners and end users understand what really happens on the ground. Department heads sometimes optimize for control or reporting; end users optimize for workflow and efficiency. You need both perspectives.
• Document current state vs. future state. How are processes different 3 years from now? Are you expanding into new geographies? Doubling headcount? Moving from custom software to off-the-shelf? Future-state requirements are as important as current-state requirements.
• Build a requirements matrix by function. Structure requirements by business area (Finance, Operations, Supply Chain, Manufacturing, CRM, Reporting, Compliance). This makes it easier to evaluate vendors systematically against each category.

Requirements Matrix Template

Business Function	Sample Requirements	Must-Have?	Priority (1-5)
Finance & Accounting	Multi-currency support, consolidated GL, automated accruals, cash forecasting, budget variance analysis, SOX compliance reporting	Finance core features: Yes; Advanced: Varies	1-3
Operations	Purchase-to-pay, procure-to-pay, expense management, vendor management, contract management, procurement analytics	P2P core: Yes; Advanced: Varies	1-3
Supply Chain	Demand planning, inventory optimization, supplier quality management, logistics tracking, multi-warehouse support	Depends on complexity	1-5
Manufacturing	Bill of materials, work order management, production scheduling, quality management, lot/serial tracking, costing methods	If manufacturing: Yes	1-2
CRM & Sales	Sales pipeline, forecasting, quote-to-cash, opportunity management, customer hierarchy support	Varies by company	2-4
Reporting & Analytics	Real-time dashboards, self-service BI, financial reporting, operational KPIs, ad-hoc query capability	Basic: Yes; Advanced: Varies	1-3
Compliance & Security	Data encryption, audit trails, role-based access, SOX/HIPAA/GDPR compliance, regulatory reporting	Core security: Yes; Advanced: Varies	1-2

Step 2: Set Your Budget and Timeline

ERP projects fail primarily for one reason: they exceed budget or timeline. Why? Because companies didn't plan either realistically.

Budget and timeline aren't just constraints; they're selection criteria. An ERP that requires 12 months and $2M might be perfect functionally, but if you need 6 months and $1M, it's wrong for you. The wrong timeline creates pressure that cascades into bad decisions downstream.

How to calculate total cost of ownership (TCO):

Most companies underestimate ERP costs because they only count software licenses. Real TCO includes:

• Software licenses: Per-user costs, implementation seats, add-ons (analytics, security modules, industry-specific packs)
• Implementation services: Partner labor, internal staff costs, change management, training development
• Data migration: Data cleansing, mapping, ETL tools, testing
• Training: Train-the-trainer development, end-user training sessions, documentation
• Integration & customization: Custom code, third-party integrations, middleware
• Ongoing support: Annual maintenance, premium support, cloud infrastructure costs
• Infrastructure (if on-premises): Hardware, networking, backup systems, disaster recovery
• Contingency: 10-15% buffer for overruns

Typical Budget Allocation (Percentage of Total Project Cost)

Cost Category	Percentage of Budget	Notes
Software Licenses (Year 1)	20-25%	Varies by user count and system choice
Implementation Services	35-45%	Includes partner consulting and internal staff
Data Migration	10-15%	Higher if complex legacy systems
Training & Change Management	8-12%	Critical; often underfunded
Infrastructure & Integration	5-10%	Lower for cloud; higher for on-premises
Contingency & Other	10-15%	Essential buffer for overruns

Timeline expectations by company size:

• SMB (50-250 users, $10M-$100M revenue): 4-8 months typical; 3-month minimum (risky), 12-month maximum (organizational stress)
• Mid-Market (250-1000 users, $100M-$500M revenue): 6-12 months typical; 6-month minimum (very risky), 18-month maximum
• Enterprise (1000+ users, $500M+ revenue): 12-24 months typical; 12-month minimum (high risk), 36-month maximum

Step 3: Research and Create a Shortlist

The global ERP market includes 50+ systems. Your job is to narrow that down to 3-5 finalists in 2-3 weeks. The goal of shortlisting is efficiency, not perfection. You're eliminating systems that clearly don't fit, not identifying the perfect choice.

Shortlisting screening criteria:

• Must-haves match: Does the ERP have all 5-8 of your functional must-haves? If not, eliminate.
• Industry fit: Is it proven in your industry? Manufacturing ERPs are different from distribution ERPs. Vertically-focused systems often outperform horizontal systems in specialized industries.
• Company size fit: Is it sized for your company? A system designed for enterprises may be overkill for mid-market. A system designed for SMBs may lack features you need at scale.
• Deployment model: Cloud vs. on-premises. This is often a must-have decision before shortlisting. Cloud is faster to implement; on-premises offers more control but higher infrastructure costs.
• Implementation time & cost: Does the typical implementation timeline match your constraints? Some systems are famous for long, expensive implementations.
• Vendor stability: Is the vendor going to exist in 5 years? Is it gaining or losing market share? Private equity-backed vendors are riskier than public companies or established private companies.
• Partner ecosystem: Can you find skilled implementation partners in your region? Some systems have strong partner networks; others don't.

ERP Systems by Company Size (Typical Market Positioning)

Company Size	Typical ERP Options	Implementation Timeline	Starting License Cost (Year 1)
SMB: $10M-$50M, 50-200 users	Business Central, Odoo, Xero, Wave, NetSuite (high end)	3-8 months	$40K-$150K
Mid-Market: $50M-$250M, 200-750 users	Business Central, NetSuite, Infor CloudSuite, Epicor, Aptean	6-12 months	$150K-$500K
Large Mid-Market: $250M-$1B, 750-2000 users	NetSuite, SAP S/4HANA, Infor, Aptean, Lawson (now Infor)	9-18 months	$500K-$1.5M
Enterprise: $1B+, 2000+ users	SAP, Oracle Fusion, NetSuite (scale), Infor, JD Edwards	12-36 months	$1.5M+

Step 4: Evaluate Vendors Through Structured Demos

This is where many companies go wrong. They sit through canned vendor demos, see what vendors want to show them, and make selection decisions based on marketing.

How to run effective, structured demos:

Before the demo:

• Create a scenario script. Don't let vendors demo their "highlights reel." Instead, give them a scripted business scenario from your company. Example: "A customer places an order for 100 units of product X with custom colors. Walk us through the entire quote-to-cash process: sales order, manufacturing planning, quality checks, inventory allocation, shipping, invoicing, AR collection." This forces vendors to demonstrate real-world fit, not marketing features.
• Define scoring criteria. Create a weighted scoring matrix before demos. Criteria might include: ease of use (20%), functional completeness (25%), performance (10%), vendor stability (15%), support quality (15%), TCO (15%). Evaluate each vendor against the same criteria so you can compare apples to apples.
• Assign demo observers. Have representatives from Finance, Operations, Supply Chain, Manufacturing, IT, and at least one end user in each demo. Different stakeholders will have different reactions; multiple perspectives catch things one person misses.

During the demo (allocate 4-5 hours per vendor across 2-3 sessions):

• First hour: Overview & navigation. How intuitive is the UI? Can someone find a basic screen without training? Try to break the system with unexpected steps.
• Next 2-3 hours: Scripted scenarios. Run your prepared business scenarios. Watch how many clicks, how many steps, whether the system forces your workflow or you're forcing it into the system.
• Last hour: Integration, reporting, admin. How does this ERP talk to other systems? How easy is reporting? Can non-IT users create reports? How does the vendor manage security and user access?
• Real-time scoring. Have observers score the demo in real-time using your matrix. Don't wait until later; scores fade quickly. Capture notes and impressions immediately.

Red flags to watch during demos:

• Vendor struggles with your scripted scenario but pivots to a different use case ("Well, we handle it a different way...")
• Heavy customization needed for basic processes (should be configuration)
• Reporting requires IT support for ad-hoc queries (should be self-service)
• Multi-step processes that seem unnecessarily complex
• Slow system performance or delays in screen loading
• Vendor demo person doesn't know how to answer basic configuration questions
• Mobile and offline access not functional (if you require it)
• Integration with critical third-party systems requires custom code

Step 5: Check References and Validate Claims

Vendor references are theater. They're handpicked to be happy and prepped by the vendor to say nice things. But reference calls are still worth doing; they're your chance to ask real questions of real customers who've implemented the system recently.

How to get real insights from reference calls:

• Request multiple references in similar industries and sizes. A reference from a $1B manufacturing company isn't relevant if you're $50M distribution. Request references from companies similar to yours (industry, size, complexity, timeline).
• Ask for recent implementations. A reference from a 2019 implementation is outdated. Software changes, partnership approaches change. Request references from the past 18 months.
• Ask these specific questions:
  • "What was the actual implementation timeline vs. planned? What caused delays, if any?"
  • "Did implementation costs stay within budget? What cost overruns occurred and why?"
  • "What surprised you negatively during implementation?"
  • "How long before your organization saw ROI? What contributed to delays or acceleration?"
  • "What would you do differently if you could start over?"
  • "How responsive was the implementation partner? The vendor support team?"
  • "Do your end users like the system? What's the adoption rate?"
  • "For our specific business scenario (describe), how does this ERP handle it?"
• Listen for hesitation and canned answers. References who say "It was perfect and we have no complaints" are lying or filtered by the vendor. Real references mention trade-offs, challenges overcome, and lessons learned. Those are more credible.
• Ask about vendor responsiveness. Years 2-5 after implementation, vendor support and updates matter more than initial implementation. How responsive is vendor support? How often do they update the system? How are they driving customer success?

Warning signs from reference calls:

• Reference is defensive; gets frustrated with tough questions
• Implementation took significantly longer than vendor claims ("We were told 6 months; it took 14")
• Implementation costs exceeded budget by 30%+ and references don't understand why
• Adoption is lower than expected; many users still use legacy workarounds
• ROI delayed by 12+ months due to issues unrelated to change management
• Vendor has been slow to respond to feature requests or critical issues

Step 6: Negotiate Contracts and Pricing

Most companies negotiate only the license price. That's a mistake. The license is 20-25% of TCO. You should negotiate the total contract value: licenses, implementation, support, and terms.

Common pricing pitfalls to avoid:

• Per-user per-month licensing without volume discounts. If a system charges $100/user/month with no discounts, 100 users costs $1.2M/year. Negotiate volume tiers ($80/user/month for 100+, $60 for 200+).
• Implementation charges separate from license. Some vendors quote low license costs but charge substantial implementation fees. Negotiate bundled pricing that includes implementation support, data migration, and training.
• Support costs not included in license. Some vendors charge 20% of license cost annually for maintenance + support. Negotiate whether that's included in Year 1 or separate.
• Add-on modules charged separately. Analytics, advanced manufacturing, compliance modules sometimes cost extra. Clarify all add-ons upfront so you understand true TCO.
• Implementation partner markup. Some vendor partner programs don't allow discounting; partners mark up services 30-40%. Negotiate directly with the vendor for favorable partner rates if possible.

Negotiation leverage points:

• Multi-year commitment: "If we commit to 3 years upfront with no early termination, will you reduce annual costs 10-15%?"
• Volume: "We're evaluating 3 vendors for this contract. If you drop to $X, we're signing with you today."
• Upsell potential: "We're $50M revenue today; we plan to be $150M in 5 years. If you give us favorable pricing now, we'll buy additional modules and users as we grow."
• Reference case: "If we become a strong reference case for companies like us, will you reduce implementation costs by 15%?"
• Flexible implementation timeline: "We don't need go-live until Q4. If we go-live in off-season, will you discount implementation?"

Contract terms to review:

• Performance guarantees (SLAs). What uptime does the vendor guarantee? 99.9%? 99.5%? What's the penalty if they miss it?
• Support response times. Critical issues should have 1-2 hour response time, not 24 hours. Clarify severity definitions and response times for each.
• Data export rights. Can you export your data in standard formats if you leave? Avoid lock-in clauses.
• Customization ownership. If you pay for custom code development, who owns it? (You should.) Can you modify it later?
• Update obligations. How often does the vendor release updates? Are you required to update? (Most cloud vendors require quarterly updates.)
• Termination clause. What's the termination notice period? Can you terminate for cause or only convenience? What are the penalties?

Step 7: Plan the Implementation

Selection and implementation are separate projects that must align perfectly. The best ERP selection creates conditions for implementation success. Planning should begin during selection so that implementation starts the day contracts are signed.

Key implementation planning decisions to make during selection phase:

• Big Bang vs. Phased: Deploy everything at once (faster, riskier) or in phases by department/region (slower, safer). Phased implementations rarely save total time or cost; they just spread pain over longer periods. Big Bang is typically recommended for clarity, unless organizational change capacity is truly limited.
• Partner selection: Will you use the vendor's direct services, a partner, or a hybrid? Vendor direct services offer deep product knowledge but higher costs. Partners offer local presence and relationships but variable quality. Many companies do hybrid: vendor for architecture/design; partner for configuration/deployment.
• Internal staffing: Who from your organization will be involved? Are they dedicated 100% or shared with other projects? How will you backfill their normal roles? 30-50% of implementation delays are due to organizational inability to staff the project adequately.
• Change management approach: How will you manage organizational resistance? Will you have a dedicated change manager? How will you communicate vision and progress? Will there be incentives for adoption? Change management budgets are often 8-12% of project cost but are sometimes seen as "nice to have" when budgets get tight. That's wrong; change management is critical.
• Timeline integration: How does implementation timeline align with business cycles? Implementing an ERP during your busiest season (like month-end close or peak sales periods) creates conflict. Plan for quieter periods if possible. Some companies implement post-acquisition when there's less operational pressure.

Common ERP Selection Mistakes

Avoid these 10 mistakes that derail selection decisions:

1. Not involving end users in selection. Department heads and IT select a system that executives like, but end users find hard to use. End-user adoption is the #1 predictor of ERP success, so end users should have a voice in selection.
2. Focusing only on features, ignoring usability. A system with 100 features you might use is worse than a system with 50 features you'll definitely use and love. Ease of use should weight 20-30% of your selection decision.
3. Ignoring total cost of ownership. Comparing only per-user costs or annual license cost, ignoring implementation, training, and data migration. That's like comparing airplane ticket prices without considering destination. TCO is what matters.
4. Rushing requirements gathering. Skipping straight to vendor demos because you're impatient. Rushing this phase costs months later when you realize the shortlist doesn't match your requirements.
5. Choosing based on price alone. Selecting the cheapest option without evaluating fit, support quality, or implementation timeline. Saving $100K on licenses often costs $500K in implementation delays and customization.
6. Not validating vendor claims with references. Trusting vendor marketing instead of talking to real customers. References often reveal issues that vendor salespeople never mention.
7. Underestimating implementation complexity. Assuming "it's just software" and that implementation will be smooth. ERP implementations are organizational transformations, not software deployments. They're complex and time-consuming.
8. Ignoring integration requirements. Selecting an ERP without evaluating whether it integrates smoothly with other critical systems (accounting, CRM, HR, e-commerce). Integration problems surface 6-9 months into implementation when it's expensive to fix.
9. Not planning change management early. Waiting until deployment phase to think about how you'll get users to adopt the new system. Change management planning should start during selection so you can bake adoption thinking into implementation from day one.
10. Underestimating timeline and cost buffers. Planning to the minute without padding. ERP projects have unknowns; 10-15% contingency buffers are essential and should be built in from day one.

ERP Selection Checklist

Use this checklist to ensure you haven't missed critical selection steps:

Phase	Checklist Item	Completed?
Requirements	Cross-functional team assembled (Finance, Ops, Supply Chain, Manufacturing, IT, HR)	☐
	Current state processes documented	☐
	Future state requirements defined (3-5 year vision)	☐
	Must-haves vs. nice-to-haves clearly identified	☐
	Requirements matrix by business function completed	☐
Budget & Timeline	Total cost of ownership (TCO) calculated by category	☐
	Implementation timeline estimated and approved by leadership	☐
	10-15% contingency buffer built into budget and timeline	☐
	Internal resource availability confirmed	☐
Shortlist	Research completed; 50+ candidate systems narrowed to 3-5	☐
	Each finalist meets all must-haves	☐
	Industry fit and company-size fit validated	☐
	Partner ecosystem and vendor stability assessed	☐
Demos	Scoring matrix defined before demos begin	☐
	Scripted business scenarios created (not vendor-selected demos)	☐
	3-5 demo sessions per vendor completed	☐
	Multi-functional observers participated and scored each demo	☐
	Demo scoring consolidated and finalists narrowed to 2-3	☐
References	Minimum 3 reference calls per finalist completed	☐
	References from similar industry, size, and implementation timeline	☐
	Real questions asked (timeline, budget, adoption, lessons learned)	☐
	Red flags or concerns documented and discussed with finalist	☐
Negotiation	License pricing negotiated (volume discounts, multi-year terms)	☐
	Implementation, training, support bundled into proposal	☐
	Contract reviewed by legal (SLAs, data rights, customization ownership)	☐
	True total cost of ownership confirmed in final proposal	☐
	Final contract signed with clear terms and timeline	☐
Implementation Planning	Implementation partner selected or internal team staffed	☐
	Detailed implementation plan created	☐
	Change management strategy and budget allocated	☐
	Kickoff meeting scheduled; implementation project begins	☐