Forum Discussion
Camel Case Displaying as Lowercase When Exporting to Excel
Responsive ResidentHii kteja
This is a known behavior in Power BI exports, especially when using DirectQuery with sources like Snowflake. The issue is not actually Excel, but how Power BI’s export engine pulls the data from the underlying model.
In DirectQuery mode, what you see in Power BI visuals ≠ what gets exported.
Excel receives the raw Snowflake values—hence “camelCase” becomes all lowercase.
To ensure proper casing in export, you must:
✔ store the formatted text in the model (Import mode or calculated column),
OR
✔ fix casing at source (Snowflake).
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
Happy to help!
Hi Shivu-2000,
Thanks for the explanation — I understand your point about how the export engine pulls data differently in DirectQuery.
However, in my case the values in Snowflake are already in the correct camel-case format. The issue is that only a few specific rows are being converted to lowercase during export, while the rest of the dataset exports correctly.
I also can’t apply any text-formatting logic in Power BI and snowflake, because some rows are intentionally meant to remain lowercase based on the client’s requirement. So applying a transformation would modify valid lowercase values, which I need to avoid.
This inconsistency for just a few rows is what I’m trying to troubleshoot. If you have any suggestions on how to handle selective casing issues in DirectQuery exports, I’d really appreciate it.
v-hashadapuCommunity Support
Hi kteja , Thank you for reaching out to the Microsoft Community Forum.
What you’re seeing usually happens when the export engine sends a slightly different SQL request than the one used to render the visual. Even though the values in Snowflake are stored in proper camel case, DirectQuery exports don’t reuse the same query Power BI uses for visuals. Instead, they issue a simplified query that can bypass metadata, lose collation hints or hit a different expression path in a view or join. When that happens, only certain rows, typically the ones coming from a branch of the query that applies an implicit LOWER(), a cast or a case insensitive collation, come back in lowercase. That’s why the visual looks perfect, but the export doesn’t match for only a handful of rows.
Run the actual export query directly in Snowflake using Performance Analyzer. If those few rows also return lowercase there, then you know the export engine is triggering a different code path upstream. If they return camel case in Snowflake but still export lowercase, then the issue is on the Power BI side and it’s an inconsistency in how DirectQuery exports handle text fields. In either scenario, the only reliable fix without touching valid lowercase rows is a source side conditional expression that corrects casing only for the specific problematic patterns, leaving intentional lowercase values untouched.
Thank you Shivu-2000 for your valuable response.