How to Choose a New Business System: A 9-Step Selection Process

The short answer

To choose a new business system, define the business outcomes first, agree who makes the decision and how options will be scored, then test a shortlist of three or four vendors against scripted scenarios based on your own processes and data. Compare total cost of ownership over at least five years, check references, and negotiate the contract before committing.

Why does system selection go wrong?

Many troubled implementations can be traced back to decisions made during selection. Common causes include:

• choosing on features and polished demonstrations rather than business fit
• requirements lists that are long, unprioritised, and copied from the current system
• underestimating data migration, integration, and change effort
• evaluating the software but not the implementation partner
• no single accountable decision owner

The nine steps below are designed to prevent those problems. They apply to ERP, CRM, finance, HR, practice management, and other core platforms.

Step 1: Define the problem and the outcomes

Write down why you are changing systems and what must be better afterwards. Express outcomes in business terms: a faster month-end close, fewer manual reconciliations, a single view of customers, reporting the board trusts, or the ability to open a new site without adding administrative headcount. These outcomes become the benchmark for every later decision, including the business case.

Step 2: Understand current processes and data

Map the core processes the system will support, where data lives today, and which reports and integrations depend on it. Identify which processes are genuinely distinctive and which are standard. Standard processes are usually better adapted to the software than customised to match old habits.

Step 3: Set up decision-making before talking to vendors

Agree:

• a single accountable business owner for the decision
• a small steering group with finance, operations, and technology represented
• the evaluation criteria and weightings, for example business fit, total cost, implementation risk, vendor viability, and security
• how scores will be recorded and how disagreements will be resolved

Setting criteria before demonstrations stops the decision being swayed by the most persuasive presentation.

Step 4: Write prioritised requirements

Keep requirements short and ranked:

• Must have: the system is unacceptable without it
• Should have: important, but a workaround is tolerable
• Could have: useful if cost and effort allow

Include non-functional requirements: security controls, integration methods, data ownership and export, reporting, performance, support arrangements, and where data is hosted. Confirm how each product supports your privacy and data-handling obligations under Australian law.

Step 5: Scan the market and shortlist

Identify products used by organisations of similar size and industry. Screen out options that fail must-have requirements or fall well outside budget. Aim for a shortlist of three or four. A longer list rarely improves the decision and slows everyone down.

Step 6: Run scripted demonstrations

Give each vendor the same scenarios, based on your real processes, and ask them to demonstrate those scenarios, ideally with sample data you provide. Score each scenario against your criteria straight after the session, while it is fresh. Scripted demonstrations show how the product handles your work, not just its best features.

Step 7: Evaluate the implementation partner, not just the product

The same software can succeed or fail depending on who implements it. Assess the partner's experience with organisations like yours, the named people who would work on your project, their approach to data migration and testing, and how they manage change requests. Speak to reference customers without the vendor present and ask what they would do differently.

Step 8: Compare total cost of ownership

Licence fees are only part of the cost. Model at least five years:

The cheapest licence is often not the lowest total cost.

Step 9: Negotiate the contract and make the decision

Before signing, confirm scope, milestones, acceptance criteria, service levels, limits on price increases, data ownership, and exit terms. Then take the recommendation, the evaluation evidence, and the business case to a formal decision gate. The Australian Government's business case guidance, listed in the sources below, is a useful reference for structuring options, costs, benefits, and risks, even for private organisations.

How long does system selection take?

It depends on complexity and how quickly decisions are made. For a mid-sized organisation replacing a core platform, a disciplined selection commonly takes a few months. Rushing selection to save a few weeks often costs far more during implementation.

Who should run the selection?

The business should own the decision; technology should shape and test it. An independent adviser can run the process, keep every vendor to the same rules, and challenge assumptions. That is particularly useful when there is no internal CIO, or when the vendor also proposes to implement.

CIO-led technology transformation explains how independent governance works alongside the implementation partner. For what happens after the contract is signed, see how to govern a technology transformation project. If you are still deciding whether to replace your current system at all, start with replace or modernise a legacy system.