One sentence conclusion: When the platform uses points for accounting, there is no direct comparison between the two bills. ——You have to convert points into money first, and the conversion relationship is often not public or unstable.
This is not a problem of which company, but a natural result of different business models. But it's a real obstacle:Many people cannot calculate their own AI costs and are stuck at this step.
Phenomenon
Export a usage details from the platform background. There are three common situations:
- Inside isAmount, can be understood directly;
- Inside isPoints, I can’t see how much it corresponds to;
- InsideBoth, but the ratio between the two is different from what you think.
When comparing bills from different platforms together, the third type is the most misleading - they all appear to be "numbers", but in fact the units are different.
Disassembly: What are the three difficulties?
The first point: The billing methods are not uniform.Some are calculated based on AI billing units, some are calculated based on times, some are calculated based on subscriptions, and some are billed based on points.The "billing methods" are different, so the numbers on the books are not comparable. - This is not a precision issue, it is a unit issue.
The second point: The difference between points and money is a platform-side parameter.In a point-based platform, "How much is 1 point worth" is a conversion relationship on the platform side, which will change with packages, activities, and time.The points you see do not constitute a price that can be directly disclosed to the outside world.; Any number that is to be disclosed to the outside world must be accompanied by "which day".
The third point: The exported files often only have a summary and no details.This is the most practical one: without detailed fields in the exported file, it is impossible to calculate which link the cost is concentrated in, nor how much duplicate overhead is accounted for.The thickness of the exported fields of the bill directly determines the lower limit of the conclusion.
Conclusion
When facing a points-based bill, you need to distinguish between what you can and cannot do:
| Can it be counted? | How to do it | |
|---|---|---|
| Total Spending and Trends | Can | Give priority to the vertical comparison of historical data on the same platform and the same billing method. |
| Where is the cost concentrated? | Look at the fields | It can only be calculated if there are detailed items; if not, clearly write "This section cannot be calculated" |
| Cross-platform horizontal ratio | Usually cannot be directly compared to | The amount must be unified first, and the date on which the conversion is based must be marked. |
Calculate what can be calculated and clearly mark what cannot be calculated - Not giving a number that looks like a conclusion is the easiest and most credible way to write on this road.
Boundary (under what circumstances does this article not hold)
- If your bill export contains both the amount and complete details, then the above three obstacles do not exist and you can calculate it directly. This article does not apply.
- "Can't figure it out" does not mean "no way": it usually meansNeed to fill in a field, rather than changing to another service provider.
- This article does not contain the specific conversion ratio of any platform, nor does it evaluate the billing model of any platform - it only explains the differences in methods when comparing.
How can you verify it yourself?
Group the bills you exported into three categories:With money / with points / with both。
Then only make a cross-platform comparison for the part with "amount", and put the rest aside for now - you will find that most of the "uncalculated" parts before are from putting the numbers of different units together.
Source and Description
- This article does not cite the pricing and points conversion ratio of any platform.
- Source: Tunenomics Tunenomics unified external measurement basis (compare and review date: 2026-09-18).
- This article does not contain original customer data, actual measured figures and savings commitments.
This article is from Tunenomics’ “AI Cost Observation” series. The description of methods and the measurement basis of conclusions in the article can be checked; the numbers are marked with the source, measurement basis and date. Sample data does not represent actual results for any customer.