How Are You Accurately Counting Characters for Storyline 360 Localization?

We're working with Articulate Storyline 360 Localization and trying to find a reliable way to calculate the total number of source characters, including spaces, before translating a project.

Since Articulate charges translation credits based on the number of source characters multiplied by the number of target languages, getting an accurate count is important for budgeting.

We've explored exporting project content to Word and XLIFF, but we're unsure which method provides the most accurate count. 

For those of you using Storyline 360 Localization:

1. How are you calculating the total number of characters in a project before localization?
2. Is there a way to get a character count that closely matches what Articulate uses to calculate translation credits?
3. Are you using Word exports, XLIFF files, third-party tools, or another method?
4. Does Articulate count repeated text across slide content, notes, captions, and audio scripts separately?

Ideally, we're looking for a repeatable method that allows us to estimate translation credit costs accurately across dozens of existing Storyline projects.

I'd appreciate hearing what methods others have found reliable.