Create a Report by Uploading a Custom Dataset

min read - Updated on Nov 15, 2026

You can create a Report from a file you already have by uploading it to Kimola. Kimola reads the column that holds your text, groups every row under the themes it mentions, labels its sentiment and, if you choose, answers research questions about it with Dimensions. A file of survey answers, reviews, call center notes or CRM comments becomes a Report you can filter, compare and export. Kimola accepts .xls, .xlsx, .csv and .tsv files, one file at a time, up to 30 MB.

Uploading a file suits feedback that doesn't come from a page Kimola can read. To analyze reviews of a product or business on a supported platform, create a Report by adding links instead. To analyze a spreadsheet that lives in Google Drive without downloading it, connect Google Sheets.

Getting Ready

Sign in to your Kimola account and open Reports in the left menu. Have your file ready, with the column names in its first row and the text you want to analyze in one column.

Step 1: Upload the File

The creation area of the Reports page has an upload area for files. Drag your file onto it, or click it and choose the file from your computer. On an account without Reports, the page opens with Analyze Customer Feedback and offers Upload a Dataset.

[Screenshot: The upload area on the Reports page]

Kimola accepts one file per Report, in .xls, .xlsx, .csv or .tsv format, up to 30 MB. After you drop the file, Kimola uploads it and checks its format, size, readability and data structure, and then opens the steps that follow.

The way the file is laid out decides what Kimola can do with it. The first row holds the column names, and every row below it is one piece of feedback. Three columns matter to Kimola.

Text column (required): the column with the content to analyze, such as reviews, comments or survey answers. Keep the text you want to analyze in this one column.

Date column (optional): the date and time of each row. With a date column, the Report can show how topics change over time.

URL column (optional): the web address of each row's source, such as a review page. With a URL column, you can open the original source from the Report.

Tip

Columns other than these three aren't analyzed, but they become filters in the Report. Keep columns such as product, region or plan in the file if you want to filter or compare the feedback by them later.

Step 2: Choose the Columns

After the upload, the wizard opens on Sheets & Columns, with a preview of your file. Choose the columns Kimola should use. The wizard moves on only after you have chosen a Text Column.

When an Excel file has more than one sheet, the preview lists the sheets by name. Click a sheet to work with it, and choose its columns again.

Choose the Text Column, which holds the text content to analyze. If your file has them, also choose the Url Column, which holds each row's web address, and the Date Column, which holds its date and time. You pick each column by the name in your file's first row.

Example

An online course platform exports its end-of-course survey with the columns Student Comment, Course, Submitted On and Rating. Student Comment is the Text Column and Submitted On is the Date Column. The Course column stays in the file and becomes a filter, so the team can read the comments of one course at a time.

Step 3: Choose Classifiers and Structure

In Classifiers, you choose the models that label every row of your file. The Automatic Classifier discovers the themes from your data, and the Sentiment Classifier labels the sentiment of each row. Choose a content classifier and the Sentiment Classifier together to see the best and worst performing themes in the Report.

You can add other pretrained models for a specific kind of feedback. In the paid packages, you can also use your own custom models, for example a model with the topics of your survey. Starter doesn't include custom models.

In Structure, you choose how feedback is labeled. With Enable Multi-label Classification, one row can get several labels when it mentions different aspects, which gives a more complete view of strengths and weaknesses. Without it, each row is grouped under one main theme. See Create a Report Using Multi-Label Classification.

[Screenshot: The Classifiers step with the Automatic Classifier and the Sentiment Classifier chosen]

Step 4: Select Dimensions (Optional)

Dimensions are research questions that Kimola asks of every row, such as Pain Points ("What goes wrong or creates difficulty?") or Motivations ("Why is this chosen or preferred?"). Their answers come on top of the themes and sentiment that every Report has. The list is grouped into Drivers, Issues, Actors, Context and Comparisons, and All shows every Dimension.

Choose the Dimensions you want, or skip the step if you don't need them. See Add Dimensions to a Report for what each Dimension asks.

Tip

Selected Dimensions appear under My List, so you can review them before you create the Report.

Note

Each Dimension adds to the queries a Report uses: queries = rows × models × Dimensions. See How to Calculate Queries.

Step 5: Review and Create the Report

The Review step shows your choices before anything is analyzed. Check the title, the language and the number of queries, then create the Report.

Report Title

The Report Title is optional. A clear, descriptive title helps you find the Report in the Reports list later, especially when you upload several files.

Report Language

The Report Language is the language Kimola writes the Report in: theme names, sentiment labels, Dimension answers and summaries. Choose English, Spanish, French, German, Portuguese, Dutch or Turkish. Your file can be in any of the 30+ languages Kimola analyzes, whatever the Report Language. See Language Options in Reports.

Dataset and Required Query

Dataset shows your file and its rows. When some rows have an empty Text Column, only the rows with text are analyzed and counted in queries.

Required Query shows how many queries the analysis will use. Queries are the unit Kimola counts usage in, and the number follows the rule queries = rows × models × Dimensions: every analyzed row is multiplied by the number of models you chose and the number of Dimensions.

Note

If your package doesn't have enough queries left for every row, the step says so before you continue, and only part of the file is analyzed if you proceed. See kimola.com/pricing for the packages.

Create the Report

When the review looks right, click Create Report to start the analysis. Kimola processes the file with the classifiers, structure and Dimensions you chose.

What Happens Next

When the analysis is complete, Kimola shows Here we go!, and the Report is ready to view, download and manage. It appears in your Reports list and opens on Overview, which shows the number of records analyzed, the sentiment, the best and worst performing themes and your Dimensions.

From there, you can filter records by theme, sentiment or any column from your file, and export the results. For the whole workflow with survey data, from export to action, see How to Analyze Open-Ended Survey Responses. For more on what Kimola does with survey data, see Survey Analysis.

Was this article helpful?

Tell us about your thoughts and experiences regarding the article.