By David Roberts MBA FCCA. Former PLC CFO who previously worked at PwC, with hands-on international establishment, transformation, carve-out and integration experience. GAP Consultants was established in 2001. Published 8 October 2026.
Slow reporting and spreadsheet workarounds can make a new finance system look like the obvious answer. Sometimes it is. Sometimes the bigger problem is poor data, unclear ownership or processes that have never been designed to work consistently.
An improve-or-replace assessment starts with the outcome the finance team needs, then compares realistic options. It helps avoid carrying the same weaknesses into a more expensive platform.
1. Describe the problem in operational terms
Replace broad complaints with examples: how long does month-end take, which reconciliations remain unresolved, where is information rekeyed and which decisions lack reliable data? Establish a baseline before evaluating suppliers.
- List the reports and controls the organisation needs.
- Map the processes creating delay, errors or repeated work.
- Identify data gaps and responsibilities that technology cannot decide.
- Distinguish essential requirements from preferences.
2. Test what the current system can do
Review configuration, available functionality, integrations and user capability. Consider whether better use of the existing platform, a process change or a limited addition would resolve the main issues. Obtain evidence through a practical demonstration rather than assuming a feature will solve the problem.
3. Compare the full cost and delivery demands
A licence price is only one element. Include implementation, data preparation, integrations, testing, training, internal cover, continuing support and the cost of disruption. Compare these with the realistic benefit and the organisation’s capacity to deliver the change.
4. Ask suppliers to demonstrate your work
Use a small set of representative transactions and reporting scenarios. Include an exception, a correction and a reconciliation. Score the demonstrations against agreed requirements. Record gaps, assumptions and responsibilities so the recommendation can be understood and challenged.
5. Set readiness criteria before a go-live date
- Data reconciled and migration exceptions resolved or explicitly accepted.
- Core processes and controls tested by the people who will operate them.
- Reporting checked against known results.
- Users trained and support responsibilities agreed.
- A workable transition and contingency plan.
6. Give AI a bounded job
AI may help with tasks such as drafting commentary, preparing proposals or extracting invoice information. Begin with one workflow, approved data and a named human reviewer. Test accuracy, exception handling, time saved and the cost of checking the output. A persuasive demonstration is not evidence that a workflow can run reliably at scale.
GAP’s manufacturing and service-business AI applications illustrate how a pilot could be structured and measured. They are illustrative scenarios, not completed client projects.
What a useful decision paper contains
The current gaps, improvement and replacement options, a requirements assessment, full-cost assumptions, delivery risks and a recommended next step. The paper should explain what changes in the way work is done, not merely which software is preferred.
Independence matters
GAP remains independent of software suppliers and does not take software reseller commissions. Our role is to lead the business case, finance requirements, controls and delivery decisions, with technical specialists assigned where needed.
See delivery evidence from regulated-utility systems and billing assurance and multi-entity automation and control transformation.
How GAP can help
GAP’s independent finance systems and AI-enabled growth support can begin with a defined review and progress into implementation or recurring part-time finance leadership. Scope, fees, capacity and start dates are agreed before work begins. Discuss your priorities with David Roberts.