Put each row into a category.

Start with the decision your categories will support

A useful category column answers a recurring question about your rows. If you want to understand what customers contact you about, use issue categories such as Billing, Technical, Account, and Shipping. If you want to route work, use the teams or queues that will receive it. These are different questions, even when they use the same source messages.

Choose labels at a similar level of detail. Billing, Refund, and Everything else create an overlap because a refund is also a billing issue. Either use broad issue categories throughout, or define a more specific set with a clear rule for choosing among them. Describe boundaries in the category definitions rather than relying on the label alone.

Decide how to handle a row with several topics. A customer might report a late shipment and request a refund in the same message. Your instructions can prioritize the action requested, the first issue mentioned, or another rule that suits the task. Apply that rule consistently so a category has the same meaning across the table.

Where a category column helps

For support reporting, a category column gives you a consistent field to filter or count after exporting. You can compare the volume of billing questions with delivery issues without asking someone to read every message again. Keep the original message alongside the label so reviewers can check the evidence.

For delivery operations, categorize the kind of exception: delay, missing shipment, damage, or address issue. The category describes what happened. It does not, by itself, determine the remedy or establish who is responsible. Those questions may need separate columns or human review.

For research and feedback, categories can group the topic of a response. Define the set around the question your research asks, and include an explicit way to handle responses that do not fit. If one response needs several independent labels, separate Flag columns may be a better choice than forcing all the information into one category.

Use the preview to find gaps in your categories

Review a mix of clear, ambiguous, and incomplete rows before processing the full table. Look for pairs of categories that keep getting confused. If reviewers cannot explain the distinction from the definitions, revise the definitions first. A different confidence threshold cannot repair categories that ask overlapping questions.

Treat No match as useful information. It can reveal a missing category, an irrelevant row, or too little evidence in the selected columns. Do not rename No match to a normal category simply to make the results look complete. Inspect why those rows did not fit and decide whether the category set needs to change.

A confidence value is a review signal, not proof that the category is correct. Check some apparently confident results as well as rows marked for review. Correct individual mistakes in the results; if many rows share the same mistake, revise the definition and preview it again before continuing.

Should you Categorize, Score, or Flag?

The same message can support different questions. Choose the output you need before choosing a tool.

Categorize

Which kind is it? Choose one label from a set, such as Billing or Shipping.

Score

To what degree does it meet the rubric? Assign a defined level, such as 1 through 5 for reproduction detail.

Flag

Does it meet this condition? Return Yes or No, with uncertain results marked for review.

A message about a broken export might be categorized as Technical, scored for the detail of its reproduction steps, and flagged for an explicit refund request. Keep those questions separate so each result remains interpretable.

Try Categorize on your spreadsheet

1. Add and preview your data. Open a tool below, then paste from Excel or Google Sheets or upload a CSV. Check the header row and column names.
2. Choose the source columns. Select the fields that contain evidence for your question. Leave out unrelated information and any existing answer column.
3. Define the result. Edit the starting definition and name the new column. Make sure the rules cover the examples and edge cases you expect.
4. Preview, review, and export. Preview a sample before processing the remaining rows. Inspect uncertain results, correct mistakes, then copy or download the results for your spreadsheet.