Controllers and CFOs in mid-sized companies are carrying more technology decisions than they were five years ago. ERP upgrades, reporting automation, data clean-up, AI pilots and security reviews all land on the finance leader's desk, often with a small team and no spare capacity.
That is why the question has shifted. It is less about which platform to buy and more about who will actually get the work finished. Here is how we would test any outside partner, whether it is a consultancy, an implementation shop or a staffing firm.
Ask what they own after go-live
A recommendation deck is cheap. A working month-end close on the new system is not.
Before signing, ask:
- Who is accountable if the timeline slips, and in writing?
- What does done mean? A launched tool, or a measurable change in close time, error rates or cost?
- What happens in the 90 days after launch?
Partners who answer in terms of business results are worth more than partners who answer in terms of deliverables. Many technology projects stall not because the software fails but because nobody owned adoption.
Make them explain it in financial terms
Your board and your owners will ask the same things every time: what does it cost, what does it return, and what does it take to keep running.
A good partner can give you a plain answer on:
- Total cost, including internal time and ongoing support
- Timeline to first measurable benefit
- Headcount effects, both the roles you will need and the work that changes
- Risks, and what you do if one occurs
If a partner talks only in features and architecture, you will end up translating for them in every steering meeting. That is your time, not theirs.
Check the people plan before the technology plan
Most stalled projects we see come down to staffing. The data analyst who understands your chart of accounts is also running payroll reporting. The one cloud engineer is on holiday during cutover. The project lead is a controller doing the job on top of a full-time role.
Ask any partner how they will fill these gaps:
- Who does the work, and are they employees, contractors or subcontractors?
- Can the team scale up for a migration and back down afterwards?
- How is knowledge handed over to your own staff?
- Who covers if a key person leaves mid-project?
Contract talent is often the right answer for a defined project such as a system conversion, a data migration or a reporting build. Permanent hires make sense for work that will never end, like owning the data model or running the AI tools once they are live. The mistake is using one when you need the other.
Fix the basics before buying anything advanced
AI tools are only as good as the data under them. If your general ledger coding is inconsistent, your customer records live in three systems and your reports are rebuilt by hand each month, a new tool will not fix it. It will make the mess faster.
A partner worth hiring will tell you that, even if it delays the larger engagement. Look for one who starts with:
- A review of data quality and system gaps
- Access controls and security basics
- A realistic view of what your legacy systems can and cannot support
Older systems also hurt hiring. Skilled analysts and engineers notice when your stack is 15 years old, and some will pass on the offer.
Build compliance in from the start
In Canada, that means privacy law, provincial requirements and, for regulated firms, OSFI expectations. In the US, it means SOX, state privacy rules and sector regulators. Private mid-sized companies are not exempt from audits, lender covenants or customer security questionnaires.
Ask partners how controls, documentation and approvals are handled during the project, not at the end. Leaving compliance as a final checkpoint is how a two-month delay turns into six.
Look for a partner who can change shape
Your needs this year are not your needs next year. A project team in spring may become a support arrangement in autumn, then a permanent hire in winter. Partners who can move between those modes save you from restarting the search each time.
That said, be wary of lock-in. Keep documentation in your hands, keep your own staff involved, and make sure you could bring the work in-house or move it elsewhere.
A short checklist
- Written definition of success tied to a business measure
- Named people and a plan for coverage
- Clear split between contract and permanent roles
- Data and security review before any AI work
- Compliance steps built into the schedule
- Exit and handover terms agreed up front
What to do next
Take your biggest finance technology project this year and run it through the checklist. Wherever the answer is vague, that is where the risk sits, and it is usually a people gap. If the gap is a contract analyst, a data engineer or a permanent finance systems hire, a Traverse recruiter can help you scope the role and find the right person.