1. Home
  2. ERP Resources
  3. ERP Demo & Vendor Evaluation
ERP Buyer’s Guide

ERP Demo & Vendor Evaluation Checklist: What to Ask Vendors to Show

An ERP vendor can make almost any product look good in a prepared demonstration. That does not make the demonstration dishonest. It simply means a standard product tour answers a different question from the one the buyer actually needs answered.

An evidence-led ERP evaluation, not a product tour
On this page
The short answer

The useful question is not “Can you show us your ERP?” It is “Can you show us how your ERP handles the way our business actually works?”

The data is clean. The user knows exactly where to click. The difficult exception does not occur. The integration is available. The report has already been designed.

A good ERP demonstration should leave the buying team with evidence about:

  • what worked;
  • what did not;
  • what is standard;
  • what needs configuration;
  • what depends on an extension or another product;
  • what requires development;
  • what still needs proving;
  • what the implementation will require.

That is very different from leaving the meeting saying:

“The demo looked impressive.”

An ERP demo is not a product tour

A product tour

Starts with the software.

A buyer-led evaluation

Starts with the business.

The vendor may naturally want to show:

  • dashboards;
  • menus;
  • features;
  • automation;
  • mobile screens;
  • new technology.

Some of that is useful.

But none of it proves your critical operating workflow.

Instead, begin with situations your organisation must be able to handle.

For example:

A customer places an order that exceeds its credit limit.

A receipt arrives short and part of the material fails inspection.

Stock moves between locations and one site reports a difference.

A connected system becomes temporarily unavailable.

Management asks for a report based on the transactions entered during the demonstration.

Now the software has something to prove.

Decide what must be proved before vendors arrive

Do not write the demo script after seeing Vendor A.

And do not let Vendor B choose a completely different demonstration because its strengths happen to lie somewhere else.

Before the first vendor arrives, classify requirements.

Must prove in the demonstration

These are requirements where seeing the actual workflow materially changes the decision.

Usually they include:

  • critical operating processes;
  • difficult exceptions;
  • financial controls;
  • traceability;
  • integrations;
  • management reporting;
  • industry-specific workflows.

Can be verified another way

Some requirements are better established through:

  • written specification;
  • architecture documentation;
  • commercial proposal;
  • security documentation;
  • contract;
  • implementation plan;
  • reference call.

Do not waste demonstration time watching someone prove that a standard field exists if the real buying risk is elsewhere.

Give every vendor the same script

Fair comparison requires comparable evidence.

Give shortlisted vendors:

  • the same business scenarios;
  • the same assumptions;
  • the same representative data;
  • the same difficult exceptions;
  • the same reports to produce;
  • the same evidence questions.

They do not need to demonstrate everything in exactly the same way.

Different ERP architectures may solve the same business problem differently.

That is useful information.

But they should be solving the same problem.

Use realistic data without creating a data-security problem

Buyer data makes a demonstration more useful.

That does not mean handing a sales team your entire production database.

Use enough realism to expose the process.

Depending on the scenario, that might include:

  • representative items;
  • customer terms;
  • approval levels;
  • warehouse structure;
  • units of measure;
  • sample BOM;
  • reporting dimensions;
  • realistic transaction quantities.

Where actual records contain confidential, personal or commercially sensitive information, use:

  • de-identified records;
  • representative substitutes;
  • synthetic examples that preserve the difficult operating condition.

Before uploading any actual company data into a vendor environment, understand where it will go, who can access it and what happens afterwards.

The objective is

Realistic testing.

Not

Unnecessary data exposure.

Make the vendor identify exactly what you are seeing

After every important demonstration step, ask:

How is this capability being delivered?

Use a consistent classification.

Standard

The capability is part of the normal product.

Configuration

Supported settings, masters or workflow tools adapt the standard product to the requirement.

Extension or additional application

Another component works with the ERP to provide the capability.

That can be a perfectly valid architecture.

Ask who owns it, supports it and maintains compatibility.

Custom development

New code is required specifically for the requirement.

That can also be a valid decision.

But somebody must build, test, document and maintain it.

Not demonstrated / still proposed

Do not turn a roadmap statement, planned development or verbal assurance into demonstrated capability.

Record it separately.

The important distinction is not:

The question is not “Does your ERP support this?”
It is “How does your proposed solution support this?”

Build scenarios from trigger to business result

Do not stop the scenario when the screen saves.

A useful scenario has a beginning, middle and operational result.

For example:

Trigger

Customer places an order.

Operating flow

  1. Credit check
  2. stock availability
  3. picking
  4. dispatch

Financial consequence

  1. Invoice
  2. receivable
  3. accounting entry

Management consequence

  1. Order status
  2. margin or fulfilment information
  3. reporting

That forces the ERP to carry the process across departments.

It also exposes the places where:

  • users re-enter data;
  • spreadsheets appear;
  • interfaces take over;
  • manual approvals remain;
  • reports are rebuilt somewhere else.

The hand-offs are often more revealing than the individual screens.

Test the exception, not only the normal flow

A system should obviously handle its normal transaction.

The difficult question is what happens when reality becomes untidy.

For every critical scenario, introduce at least one realistic exception.

For example:

Sales

Customer exceeds credit limit.

Or part of the order cannot be dispatched.

Procurement

Quantity received does not match the purchase order.

Or part of the material fails inspection.

Inventory

A transfer reaches the receiving location with a discrepancy.

Manufacturing

Actual output or consumption differs from expectation.

Integration

The connected service does not respond.

Then ask:

  • Who sees the exception?
  • What is blocked?
  • What can proceed?
  • Who can override it?
  • Is approval recorded?
  • What happens downstream?
  • How does management know it occurred?

A “yes” to a feature question does not answer those questions.

Follow the transaction into Finance and reporting

Operational functionality and financial control should not be evaluated in separate universes.

When an operational scenario changes stock, cost, receivable, payable or another financial position, follow it.

Ask:

What accounting consequence did that transaction create?

Then show it.

If a quantity is rejected, returned or reversed, show what changes.

If a transaction is corrected, show the audit trail.

If stock moves between locations, show the resulting record.

The purpose is not to turn every demonstration into an accounting lecture.

It is to confirm that the operational and financial versions of the business do not quietly diverge.

Test controls, roles and auditability

Do not run the whole demonstration as an administrator.

Ask vendors to show the workflow through realistic roles.

For example:

operator supervisor finance user approver management user

Then test:

  • who can create;
  • who can approve;
  • who can change;
  • who can override;
  • who can reverse;
  • who can see;
  • what is logged.

Ask:

Show usShow us what happens when this user does not have authority.
Show usShow us the record of what the authorised user changed.

A workflow is not only what the ERP lets somebody do.

It is also what it prevents the wrong person from doing.

Test integrations when something goes wrong

A successful integration message proves only the successful path.

Introduce failure.

For example:

The banking service is temporarily unavailable.

The warehouse interface rejects a message.

The same message arrives twice.

A connected system sends invalid data.

Then ask:

  • How does the user know?
  • Is processing blocked or queued?
  • Can the transaction be retried?
  • How is duplicate processing prevented?
  • Who monitors the failure?
  • What information is logged?
  • How does recovery happen?
  • How do the two systems reconcile afterwards?
The question is not only “Is there an API?”
It is “How does this integration operate when the surrounding world is imperfect?”

Make management reporting part of the demo

Do not accept a beautiful dashboard that has no relationship to the transactions you have just watched.

Ask the vendor to:

  1. enter or process the agreed demonstration transactions;
  2. open the management report or dashboard;
  3. show the result;
  4. drill back towards the originating records where appropriate.

Then change something.

Process an exception.

Reverse a transaction.

Approve something.

Ask:

What does management now see?

This tests whether reporting is connected to the operating record rather than existing as an isolated presentation layer.

Test one signature workflow your industry depends on

Every ERP buyer has common requirements.

Finance. Purchasing. Sales. Inventory. Reporting.

Those do not tell you everything about industry fit.

Choose at least one workflow where your industry becomes operationally distinctive.

The purpose is not to ask:

“Do you have an industry edition?”

It is to ask:

“Can your proposed solution carry this business object and its consequences through the process?”

The detailed industry scenarios belong in the relevant industry evaluation.

For the general method, use the same discipline:

  1. trigger
  2. difficult workflow
  3. exception
  4. control
  5. downstream result
  6. evidence
Which workflows deserve this treatment

The Industry-Specific ERP vs Generic ERP guide explains how to identify which workflows deserve this treatment.

Ask what it takes to reproduce what you just saw

A demonstration can prove that something is possible.

It does not automatically prove what it takes to implement.

After an important scenario, ask:

What has to exist for us to reproduce exactly what you just showed?

Record:

  • product/module;
  • configuration;
  • master data;
  • roles;
  • integration;
  • third-party product;
  • extension;
  • custom development;
  • licence;
  • infrastructure/environment;
  • implementation service;
  • manual step;
  • buyer-side responsibility.

Then ask:

The commercial consequence

Is all of that included in the proposed scope and commercial quotation?

This is where Guide 6 connects directly with the pricing and implementation guides.

Record evidence while it is still fresh

After several ERP demonstrations, memory becomes unreliable.

Do not rely on:

“I think Vendor B handled that better.”

Use an evidence record.

For each important scenario capture:

What to capture for each important demonstration scenario
ItemRecord
ScenarioWhat the vendor was asked to prove
ResultWhat actually happened
EvidenceWhat was demonstrated
Delivery methodStandard / Configuration / Extension / Custom / Not demonstrated
Manual workWhat remains outside ERP
AssumptionsWhat the demonstration depended on
GapWhat did not meet the requirement
Follow-upEvidence still required
OwnerWho will obtain/verify it
Commercial impactWhether scope or price may change

The purpose of the record is not bureaucracy.

It is to stop a polished presentation from becoming stronger evidence in memory than what actually happened.

Use critical gates before weighted scores

Not every requirement should be averaged.

Suppose Vendor A scores very well across:

  • navigation;
  • dashboards;
  • reporting;
  • user experience.

But fails a business requirement without which operations cannot run.

A high average should not make that failure disappear.

Before weighted scoring, identify genuine gates.

These might include requirements where failure means:

  • the agreed operating model cannot run;
  • a required control is absent;
  • a critical integration cannot be supported;
  • a required data or legal condition cannot be satisfied.

Do not create dozens of “must-haves” merely because every department prefers its own requirements.

A gate should mean what it says.

Score independently before discussing the vendor

If the team discusses the demonstration before anybody records an assessment, the loudest opinion can become everybody’s memory.

Ask participants to complete their evidence notes and scores independently first.

Then compare.

Differences are useful.

If Operations scores a workflow highly and Finance scores it poorly, do not simply average the two numbers.

Ask why.

Perhaps:

  • Operations saw an efficient process;
  • Finance saw a weak control;
  • IT saw an integration dependency;
  • the power user saw excessive manual effort.

The disagreement may contain more decision value than the average.

Set weights before the demonstrations

If you use weighted scoring, agree the weighting before the first vendor presents.

Do not decide afterwards that:

reporting is actually more important;

user experience should receive twice the weight

because the preferred vendor happens to be strongest there.

The weights should reflect

Business priorities.

Not

The demonstration result.

Do not let the score replace judgement

A scorecard is useful.

It is not the decision-maker.

A score of 4.2 versus 4.1 does not automatically establish that one ERP is better.

Ask:

  • Did either vendor fail a critical gate?
  • What assumptions remain?
  • Which gaps affect implementation?
  • Which scores have weak evidence?
  • Which workflow differences actually matter?
  • Which risks can the organisation absorb?

The score should make the evidence easier to discuss.

It should not manufacture mathematical certainty that the evidence does not contain.

Evaluate the product and implementation partner separately

A strong ERP product can be poorly implemented.

A strong implementation team cannot manufacture missing product capability without consequences.

If a partner will implement the system, evaluate two things.

Product fit

  • workflow capability;
  • controls;
  • reporting;
  • architecture;
  • integration;
  • data model;
  • usability.

Delivery fit

  • proposed team;
  • implementation understanding;
  • migration approach;
  • integration capability;
  • governance;
  • training;
  • support;
  • judgement about customisation;
  • ability to explain trade-offs.

Do not let a strong salesperson substitute for either.

And do not assume the senior people in the demonstration are automatically the people who will implement the project.

Ask:

The delivery question

Who from the team in this room will actually work on our implementation?

Turn every unresolved claim into a follow-up item

During a demo you will hear phrases such as:

“Yes, that is possible.”

“There is an API for that.”

“We can configure it.”

“Our partner has done that.”

“That is on the roadmap.”

Do not argue.

Record the claim.

Classify the evidence still required.

For example:

  • live demonstration;
  • screenshot/documentation;
  • architecture note;
  • implementation estimate;
  • commercial clarification;
  • reference validation;
  • contractual confirmation.

Give it:

  • an owner;
  • due date;
  • decision impact.

Then leave the item unproven until the evidence arrives.

Use customer references to validate specific claims

Do not begin a reference call by asking:

“Are you happy with the ERP?”

You will learn very little.

Instead, choose the claim you need to test.

For example:

How does the system handle the workflow we just saw?

Which parts were standard and which needed additional work?

What surprised you during implementation?

Which integrations were harder than expected?

How much internal effort did your team need?

What happens when you need support?

What would you evaluate differently if you were buying again?

A relevant reference is not simply a company in the same industry.

It is a customer whose experience reduces uncertainty about the decision you are making.

Ask for the evidence closest to your actual risk

Suppose you are evaluating batch traceability.

A famous customer logo tells you very little.

Ask:

Which customer actually uses the traceability workflow you demonstrated?

If your concern is a multi-location operation:

Which customer gives us evidence closest to that operating structure?

If your risk is integration:

Which customer uses the connected system or integration pattern closest to ours?

The useful customer proof is

The proof closest to the uncertainty.

Not

The biggest logo.

Reconcile the demo with scope, price and implementation

A demonstration creates obligations for the rest of the evaluation.

If a vendor demonstrated:

  • an extension;
  • a third-party component;
  • a custom workflow;
  • an integration;
  • a special report;
  • an additional environment;

ask whether it appears in:

The proposed solution

Is that exact component part of the architecture being sold?

Implementation scope

Who configures/builds it?

Commercial proposal

Is the necessary licence/service included?

TCO

What recurs after Year 1?

Support model

Who supports it after go-live?

The demonstrated solution and the contracted solution should not quietly become different systems.

Build the final ERP evaluation evidence pack

Before the final decision, consolidate the evidence.

For each major requirement or scenario, the decision team should be able to see:

business requirement priority scenario tested what was demonstrated standard/configuration/extension/custom status remaining manual work implementation dependency integration dependency commercial impact customer/reference evidence unresolved assumption risk owner final decision treatment

This is the record that should support the recommendation.

Not a collection of vendor brochures.

ERP demo and vendor-evaluation red flags

The vendor refuses to follow the buyer’s scenario

A standard product tour is easier to control than a difficult buyer workflow.

The vendor repeatedly says “yes” without showing it

A verbal answer is not demonstration evidence.

The difficult exception is skipped

The exception may be where the real fit difference lies.

Everything is described as standard

Ask what was configured specifically for the demonstration.

“We can customise that” answers every gap

Individual customisations may be reasonable.

A pattern of them changes implementation risk and lifecycle cost.

The vendor cannot identify which product or add-on produced the capability

The proposed architecture is not yet clear.

The demo uses a perfect administrator account

Real users operate under roles and controls.

Reporting is shown from prepared dashboards but not from demonstrated transactions

The relationship between operating record and management information remains unproven.

The integration succeeds once but failure handling cannot be shown

Operational ownership is incomplete.

A roadmap item is scored as current functionality

Future intent is not present capability.

The presenter is excellent but the proposed delivery team is unknown

Presentation quality and implementation capability are different things.

Follow-up evidence never arrives

An unanswered question after the demo remains an unanswered question.

Customer references are unrelated to the workflow being evaluated

A logo is not evidence of the requirement.

The commercial proposal does not contain components demonstrated as part of the solution

The demo and the quotation are describing different scopes.

A practical ERP demonstration script

Before each shortlisted vendor arrives, give them the same document.

Confirm the solution being demonstrated

Ask the vendor to identify:

  • ERP product/version;
  • modules;
  • extensions/add-ons;
  • implementation partner;
  • deployment assumptions;
  • capabilities not currently available;
  • anything mocked or simulated.

Run the agreed business scenarios

For each scenario:

  • begin with a real trigger;
  • complete the normal process;
  • introduce the difficult exception;
  • show controls/approvals;
  • show downstream operational impact;
  • show financial impact where relevant;
  • show reporting;
  • identify manual steps.

Classify the delivery

For each important requirement:

  • Standard
  • Configuration
  • Extension/additional application
  • Custom development
  • Not demonstrated/still proposed

Test operational resilience

Where relevant:

  • user lacks authority;
  • connected system fails;
  • message duplicates;
  • transaction needs reversal;
  • workflow requires an override.

Show management evidence

Use transactions created during the demonstration. Show:

  • operational status;
  • report/dashboard;
  • drill-down;
  • audit record where relevant.

Explain implementation impact

Ask:

  • What setup is required?
  • What data is required?
  • What integration is required?
  • What custom work is required?
  • What customer effort is assumed?
  • What additional licence/service is required?

Record unresolved evidence

For every open item:

  • evidence needed;
  • owner;
  • due date;
  • effect on decision.

Then stop.

Do not let the final thirty minutes become another product presentation.

A practical ERP demonstration evidence record

What to record for each evaluation item after a demonstration
Evaluation itemWhat to record
Business scenarioWhat we asked the vendor to prove
Critical requirementYes / No
What was shownConcise factual observation
DeliveryStandard / Configuration / Extension / Custom / Not demonstrated
Exception resultWhat happened
Control/auditWhat was demonstrated
ReportingWhat was demonstrated
Manual workWhat remains outside the ERP
Implementation dependencyWhat must be built/configured/migrated
Commercial consequenceScope/licence/service implication
Evidence qualityDemonstrated / Partial / Stated only / Follow-up required
Follow-up ownerNamed person
Reference validationRequired / Not required / Completed
Final statusAccepted / Gap / Risk / Unresolved

Do not fill the table with long meeting minutes.

Record the decision evidence.

How exactllyERP fits into this evaluation

Do not evaluate exactllyERP differently because you are reading this guide on Exactlly’s website.

Give us the same difficult scenarios you give every other shortlisted ERP vendor.

Bring:

  • the workflow;
  • the exception;
  • the control;
  • the systems that must remain;
  • the reports management needs;
  • relevant representative data;
  • industry-specific operating requirements.

Then ask us to show:

  • how the workflow runs;
  • what happens when the exception occurs;
  • who can approve or override;
  • what happens downstream;
  • where the management information comes from;
  • which part is standard;
  • which part needs configuration;
  • whether an extension, integration or development is required.

Where integration is relevant, ask us to explain the implementation requirement rather than accepting “API available” as the complete answer.

Where customer evidence is relevant, ask for the proof closest to the workflow you are validating.

And after the demonstration, ask:

The reproducibility question

What exactly must be included in the implementation and commercial scope for us to reproduce what you just showed?

That is the standard Exactlly should be held to.

Not a lower one.

ERP demo & vendor evaluation checklist

Before the demo

Solution clarity

Scenario evidence

Delivery method

Integration

Implementation

Commercial

Evaluation

References

Final decision

If those answers are clear

The demonstration has done its job.

Evaluating exactllyERP?

Do not ask us for our standard product demonstration.

Send us:

  • your difficult workflows;
  • the exceptions;
  • the controls;
  • the systems that must stay;
  • the reports management needs;
  • the industry-specific processes that matter.

Then ask us to show:

  • the real operating flow;
  • what happens when things go wrong;
  • what is standard;
  • what is configured;
  • what requires an extension, integration or development;
  • what implementation effort is required;
  • what evidence remains after the demo.