ERP Evaluation Scorecard: Criteria, Weights, and Limits

An ERP evaluation scorecard makes the criteria behind a shortlist explicit. ERP Scorecard uses five weighted factors to compare 16 systems against your industry, size, complexity, requirements, and priorities. The score supports an initial shortlist; it does not replace a demonstration, reference checks, or an implementation plan.

This page explains the model running on this site, including a worked example and its limits. The weights are our methodology, not a universal standard for every selection project.

The fit score: five weighted factors

Every system in our catalog gets a 0 to 100 fit score against your answers. The score is a weighted sum of five factors, and the weights are fixed:

Key takeaway

As of July 2026, ERP Scorecard fit scores weight five factors: industry match (25%), company size (25%), complexity alignment (20%), capability coverage (20%), and stated priorities (10%). The same answers always produce the same scores.

Industry match (25%). Is your industry one the system demonstrably serves well? Not "could theoretically serve." Serves, today, with reference customers.

Company size (25%). Does your revenue band sit inside the system's proven sweet spot? This one can override everything else. A system that is too small for you stays "likely too small" no matter how pretty its other numbers look, and the reverse holds too.

Complexity alignment (20%). Your operational complexity, meaning entities, currencies, and operations, gets compared against the complexity range the system was built for. Too much system is a real failure mode. Buying enterprise software for a single-entity company is how you end up paying for a consolidation engine you never open.

Capability coverage (20%). Of the capabilities you actually need, based on your function checklist and answers, how many does the system cover natively? Gaps are shown, not hidden.

Your priorities (10%). Cost sensitivity, Microsoft-stack affinity, industry depth, scalability headroom. These nudge the score. They never dominate it.

The readiness track

Fit is only half the question. The readiness track scores the factors that decide whether implementations succeed at all: executive sponsorship, documented requirements, budget clarity, data quality, and change risk, alongside implementation complexity, data migration risk, and integration scope. Each is a 0 to 100 score built from explicit rules, so you can see exactly which answer would move it.

What the model refuses to do

Three rules are load-bearing, and they are worth stating as plainly as possible.

1. Scores follow published rules. The same answers produce the same scores, with the criteria and weights available for you to inspect.
2. No single "the answer." Output is always a tiered shortlist with the reasoning attached, because a one-name answer from a quiz is a sales pitch wearing a lab coat.
3. No score is generated by an LLM. AI writes narrative around results the scoring engine computed. It never computes a score, which means scores are reproducible and auditable.

Key takeaway

As of September 13, 2026, ERP Scorecard covers 16 ERP and accounting systems. Profiles rate 12 functional domains on a 1 to 5 scale. Profile research dates and pricing confidence vary by system and are shown on the relevant pages.

A worked ERP evaluation scorecard example

Suppose a fictional candidate has factor scores of 80 for industry, 60 for company size, 75 for complexity, 50 for capability coverage, and 90 for priorities. These are illustrative inputs, not a vendor's actual rating.

• ▪Industry contributes 80 × 25% = 20 points.
• ▪Company size contributes 60 × 25% = 15 points.
• ▪Complexity contributes 75 × 20% = 15 points.
• ▪Capability coverage contributes 50 × 20% = 10 points.
• ▪Priorities contribute 90 × 10% = 9 points.

The weighted sum is 69 out of 100. That number tells you how the candidate performed against these criteria. It does not say there is a 69 percent chance of a successful implementation, or that the system covers 69 percent of every requirement you might discover later.

The score also sits alongside recommendation rules. A size mismatch can keep a product out of a recommended tier even when the arithmetic looks attractive. Review the capability gaps and the reasoning, rather than sorting by score alone.

What to do with unknown requirements

An unanswered question is not evidence that a capability is unnecessary. The scoring model distinguishes unanswered inputs from requirements excluded by the questionnaire's relevance rules. Risk scores include evidence bounds and confidence indicators so incomplete input is visible.

Before choosing a vendor, revisit the unknowns that could change your shortlist. Give each one an owner and a way to test it. If a high-scoring candidate depends on a capability that has not been demonstrated, keep that dependency visible in the selection record.

Fit versus readiness

Use the fit questionnaire when you are choosing which systems to evaluate. Use the readiness questionnaire when you need to understand preparation gaps across sponsorship, requirements, data, resources, and change.

Neither questionnaire observes your live environment. The results reflect the answers you provide and the dated research behind the catalog. A scored shortlist is the start of due diligence.

Why determinism matters

A deterministic model can be wrong, but it is wrong in public and wrong consistently, which means it can be corrected. When a vendor disputes a factual claim, the about page documents the process: the underlying fact gets fixed and the model rescores from there. Rankings never change as part of a dispute. Only facts do.

If you want the longer version, the methodology page has it. If you want to see it run on your own company, the assessment takes about ten minutes and shows its work on every number.