ERP Selection Process: 6-Month Plan From Requirements to Contract
8 min read ยท 2026-10-25
The ERP selection process is the series of steps you follow to choose an ERP system and an implementation partner: set goals, document requirements, shortlist vendors, run scripted demos, score them, check references and negotiate the contract. Done well, it ends with a decision the steering committee can defend and an implementation plan that starts on solid ground.
The plan below runs 24 weeks in six phases, a realistic pace for a mid-size company with a few business units. It is written for CFOs, COOs, IT directors, finance and operations leads, and the project managers who run the selection. It covers what vendor-written guides tend to skip: who should sit on the team, how to write demo scripts that expose weak fits, how to score without bias, and what to lock in before you sign.
The roadmap at a glance
Goal: A signed ERP software contract and implementation partner agreement, backed by a scored evaluation, reference checks and a draft implementation plan approved by the steering committee. Duration: 24 weeks
Business case and team (Weeks 1-3)
Agree why you are changing ERP and who will decide.
- Write the business case: problems with today's systems, goals, scope and constraints such as contract end dates.
- Name the executive sponsor, the selection lead and a steering committee from finance, operations, IT and sales.
- Pick process owners for each area, such as order to cash, procure to pay, record to report and inventory.
- Agree the decision rules: who scores, who decides, and how ties are broken.
Milestone: Business case and selection charter approved by the steering committee.
Current state and requirements (Weeks 3-9)
Document how the business works today and what the new system must do.
- Map the main end-to-end processes with each process owner, including pain points and workarounds.
- List requirements by process and rank each one as must-have, important or nice-to-have.
- Record integration needs: e-commerce, CRM, payroll, banks, warehouse and reporting tools.
- Capture non-functional needs: users, entities, currencies, languages, security, hosting and reporting.
- Note what makes your business different, since those processes decide the shortlist.
Milestone: A ranked requirements list signed off by every process owner.
Market scan and shortlist (Weeks 8-12)
Go from the whole market to three or four serious options.
- Build a long list of ERP products that fit your industry, size and deployment model.
- Send a short request for information covering fit, industry customers, partners and hosting.
- Cut to three or four vendors using must-have requirements and the business case.
- Ask each shortlisted vendor to name the implementation partner and team they would propose.
Milestone: A shortlist of three or four vendors, each with a named implementation partner.
Scripted demos and scoring (Weeks 12-18)
See each system run your real processes and score them the same way.
- Write demo scripts based on your top processes, with your own sample data and edge cases.
- Send scripts two to three weeks before the demos and ask vendors to follow them.
- Run demos by process with the right process owners in the room.
- Score each script step on a shared scorecard, right after each session.
- Hold a follow-up session on gaps and on how each gap would be handled.
Milestone: Completed scorecards for every vendor, with a top two agreed by the steering committee.
Due diligence and references (Weeks 17-21)
Test the finalists beyond the demo.
- Call at least two reference customers per finalist, in your industry and size if possible.
- Ask for a total cost estimate covering licenses or subscriptions, implementation, integrations, training, support and internal effort.
- Request a draft implementation plan with phases, roles, assumptions and timeline.
- Review security, hosting, data ownership and exit terms with IT and legal.
Milestone: A final recommendation with scores, total cost, references and risks, sent to the steering committee.
Negotiation and decision (Weeks 20-24)
Sign contracts that protect you during and after implementation.
- Negotiate the software contract: scope, user types, price increases at renewal, support levels and exit rights.
- Negotiate the partner statement of work: deliverables, named team, milestones, change control and acceptance.
- Get the steering committee decision and record the reasons.
- Plan the implementation kickoff, including internal team backfill.
Milestone: Signed software contract and statement of work, with a kickoff date set.
Who should be on the selection team
An ERP system touches almost every department, so the selection team must include the people who run those processes, not just IT. Process owners know the edge cases, and they will lead the implementation later. If they are not part of the choice, they may not own the result.
Keep the decision group small. A steering committee of four to six people is enough. Everyone else scores and advises.
- Executive sponsor (CFO or COO): owns the business case and the final decision.
- Selection lead or project manager: runs the plan, the vendor contacts and the scorecards.
- Process owners: finance, procurement, operations, warehouse, sales and HR as in scope.
- IT lead: integrations, security, hosting and data migration.
- Key users: a few people who do the daily work, to test usability.
- Procurement and legal: contract terms, pricing and risk.
- Independent advisor (optional): helps with market knowledge and negotiation.
How to write demo scripts that reveal fit
A vendor-led demo shows the strongest parts of a product. A scripted demo shows how it handles your work. Write scripts around five to eight processes that matter most and make your business different, such as configure-to-order quoting, multi-entity consolidation or lot tracking.
Use your own sample data, include an exception in each script, and ask the vendor to show it in the standard product. When they say it can be configured, ask them to show the configuration. When they say it needs development, log it as a gap with an estimated effort.
- Script title and process owner.
- Business scenario in plain words, with your data attached.
- Steps the vendor must show, in order.
- The exception or edge case to handle.
- What good looks like, so scorers know what to expect.
How to score vendors fairly
Build the scorecard before the demos, not after. Weight the criteria using the requirement ranks: must-haves carry more weight than nice-to-haves. Ask each scorer to rate only the scripts they are qualified to judge, and to score right after each session while memory is fresh.
Combine functional scores with the other factors the steering committee cares about. A common split is fit to requirements, implementation partner, total cost, technology and vendor viability. Agree the weights in phase one so nobody adjusts them later to favor a product.
- Functional fit: scores from the scripted demos, weighted by requirement rank.
- Implementation partner: team experience, method, references and named people.
- Total cost: software, implementation, integrations, support and internal effort over several years.
- Technology: integration options, reporting, security and upgrade approach.
- Vendor and partner viability: customer base in your industry, roadmap and support model.
Example: a filled-in ERP selection plan
Picture a 400-person manufacturer with three entities, running an old on-premise system and spreadsheets. The CFO sponsors the project, and the current support contract ends in 18 months.
Every line has an owner and a check. The steering committee only meets at the end of each phase, which keeps the decision moving.
- Weeks 1-3 (CFO and IT director): business case and charter approved, seven process owners named. Check: steering committee sign-off.
- Weeks 3-9 (selection lead and process owners): 11 processes mapped, 240 requirements ranked. Check: sign-off by each owner.
- Weeks 8-12 (selection lead): long list of nine products cut to four after a request for information. Check: shortlist approved.
- Weeks 12-18 (process owners and key users): four scripted demo rounds, scorecards complete. Check: top two agreed.
- Weeks 17-21 (CFO and procurement): five reference calls, multi-year cost estimates, draft plans. Check: recommendation sent.
- Weeks 20-24 (CFO, legal and procurement): contract and statement of work negotiated. Check: signed contracts and kickoff date.
How to present the recommendation to the board
Put the recommendation on one slide: the chosen vendor and partner, the top reasons tied to the business case, the multi-year cost, the implementation timeline and the top three risks with their fallbacks. Add a line on why the runner-up lost, because board members will ask.
Keep the full scorecards, demo notes and reference call notes in an appendix. Show that the process was fair: same scripts, same scorers, weights agreed before the demos. That record protects the decision when the implementation hits its first problem.
Risks in the selection itself
Selection projects slip for predictable reasons. Write them down at the start, give each an owner and review them at each steering committee.
The largest risk is internal capacity. Process owners still have day jobs. Agree in phase one how much of their time the selection needs and who covers their work.
- Process owners too busy: owner is the sponsor. Fallback: backfill or reduce their other duties for the demo weeks.
- Requirements list too long: owner is the selection lead. Fallback: rank ruthlessly and demo only the top processes.
- Vendor drives the agenda: owner is the selection lead. Fallback: send scripts early and stop demos that leave them.
- Partner team changes after signing: owner is procurement. Fallback: name key people in the statement of work.
- Hidden costs: owner is finance. Fallback: ask every finalist for the same cost template.
Common mistakes to avoid
- Starting with vendor demos before you know your requirements lets the vendor set the agenda, so map processes and rank requirements first.
- Writing hundreds of equal-weight requirements hides what matters, so rank each one and score the must-haves more heavily.
- Letting vendors run free-form demos shows only their strengths, so send scripted demos with your own data and exceptions.
- Choosing the software and ignoring the implementation partner puts the project at risk, so evaluate the partner and its named team as carefully as the product.
- Comparing only license prices misses most of the cost, so compare a multi-year total including implementation, integrations, support and internal effort.
- Leaving process owners out of the decision weakens ownership later, so make them score the demos and sign off the requirements.
Frequently asked questions
What are the steps in the ERP selection process?
Most processes follow six steps: business case and team, requirements, market scan and shortlist, scripted demos and scoring, due diligence and references, then negotiation and decision. Each step should end with a sign-off before the next one starts.
How long does ERP selection take?
For a mid-size company, a structured selection often takes four to six months. Smaller companies with simple needs can move faster, and large multi-country groups often take longer. The pace usually depends on how quickly process owners can document requirements and attend demos.
How many ERP vendors should be on the shortlist?
Three or four is a common number for scripted demos. Fewer limits your choice and your negotiating position, more stretches the team too thin. Narrow to two finalists for reference checks and detailed cost estimates.
What should ERP selection criteria include?
Include functional fit to your ranked requirements, the implementation partner, multi-year total cost, technology and integration, and vendor viability. Agree the weights before the demos so the scoring stays fair.
Who should make the final ERP decision?
A small steering committee led by the executive sponsor, usually the CFO or COO, should decide. Process owners and key users score the demos and advise. Record the decision and the reasons so the choice can be explained later.