Sometimes, very large exports can become unstable and time out. It's very common to have exports that are several hundreds fields wide and most of the time those exports will run without any issues. However, there are cases when the data volumes in combination with a very wide schema can cause exports to consistently fail.
Fortunately, there are ways to improve the stability.
Option 1: Reduce the number of fields in the export
An easy fix is to reduce the number of fields in the export. Ask yourself if you need all the chosen fields. If not, take them out of the selection.
Calculated fields can be extra compute heavy, for example those based on Date, like Week number (ISO), Year, Week number, Month, Day of week, Week and Month number. Removing those can help.
Option 2: Split the export into smaller parts
If you need all of the chosen fields, you can get it to work by splitting up the export.
One common and recommended way to split up your export is to do it by Connection type code. For this you would setup one export per Connection types (one for Twitter, one for Facebook etc).
To do this, create a new export and select only the fields you want from the specific Connection Type. Then add a filter like; Connection type code exactly matches (e.g. Facebook ads) i filter out values from common fields such as Cost.
Duplicate the export and make one for each Connection type that you want to export.
This also gives you the benefits of a more clean schema and resulting dataset where there are less empty rows.
If you still have issues with your export, please reach out to the Support Chat or send an email to firstname.lastname@example.org.