Shikar.mishra's steps will allow you to see that all of the work items from each Excel file are processed before loading the next file. However, Blue Prism is unattended automation, and nobody will likely be monitoring the results of each workbook's work items while it runs. Another option, which follows the principles of unattended automation, would be to follow the steps shikhar.mishra outlined, except load the data from each workbook sequentially into the work queue and then process them all in your Get Next Item loop. You will know all are processed by the result (completed or exception).
I recommend that after you mark each item as completed or exception, you would then add the work item to a separate Reporting queue (before doing another Get Next Item) and generate an output report at the end of the run from the work items in the Reporting queue. If you capture the exception type and exception detail during the main processing, and store those values in the primary work queue, then they will be included when you add the processed work item to the Reporting queue. Then, to generate the report, create a reporting process with a Get Next Item loop that reads from the Reporting queue and writes the record to a collection. Once all records have been processed, write the collectin to Excel to create the report. This output report will thus contain the original input data from the main process and the exception fields that you added there before you sent the work item to the Report queue. The report will be very helpful to the business line that you're creating the automation for, as it will show them the original data that was processed and the outcome of each case.
Hope this helps.