Question 1

What questions should I ask an EPM implementation firm before signing?

Accepted Answer

Ask three. Has anyone been into the data model before producing a number? What happens if the data turns out worse than expected, and who pays for that? Is the AI foundation being built, or assumed? Each one exposes where the data risk sits, and data risk is what turns a fixed quote into a change order in month four.

Question 2

Why does it matter whether a firm has read the data model before quoting?

Accepted Answer

Most quotes are built from a discovery deck and a few workshops, not from the data. A firm that has read your entities, mappings, and intercompany history before it prices knows what it is agreeing to. A firm that has not is quoting a guess, and a guess only moves one way once the build starts.

Question 3

Who pays if the finance data turns out worse than expected?

Accepted Answer

On a time-and-materials contract the answer is you: every surprise in the data is billable, so the firm has no reason to look before it starts. A fixed price only protects you if the firm priced the data first. Constancia reads the data model in a paid diagnosis and fixes the price after it, so worse-than-expected data sits with us, not you.

Question 4

How do I tell if an AI foundation is real or assumed?

Accepted Answer

Ask what makes your data ready for AI use and how the firm will prove it at go-live. If the answer is only that the platform supports AI, the foundation is assumed, not built. Automating reporting on data you cannot trust produces the wrong answer faster.