Survey & Review data defined:
Survey Data: Survey data is collected through questionnaires distributed to a specific population to gather information about various topics, opinions, behaviors, or characteristics.
Review Data: Review data comes from evaluations or critiques, often of products, services, or experiences, where individuals share their feedback, ratings, and comments. When uploaded in a CSV file, review data may include:
When uploaded in a CSV file, Survey and/or Review data may include:
Closed-ended responses: Columns for ratings on a predefined scale (e.g., 1 to 5 stars)., Yes/No answers, numerical ratings, etc.
Open-ended responses: Columns filled with free-text fields where respondents provide their answers in their own words.
Demographic data/Metadata: Columns capturing information about respondents such as age, gender, location, and other relevant attributes.
Getting Started
If you have survey and/or review program data that you would like to upload to Beehive AI's platform, there are some steps you need to take beforehand to ensure that your data is ingested properly. Using this guide will enable you to more swiftly upload your data and initiate advanced AI analysis!
Additionally, attached at the bottom of this page is a downloadable CSV template that you can use to help you during the cleanup process!
Cleaning Up Your CSV
Essential Elements
In order to properly ingest your data into our system, there are some essential elements your CSV must have.
Each CSV must have a field that will act as the Customer_ID data source type.
There can only be one data source per upload that is marked as the Customer_ID. This field is used to map responses to Beehive generated IDs that are present, one per customer, in our platform. This Customer_ID field allows for us to track all of a customer's responses across different data collections, so long as it is a consistent ID for each customer.
Again, this is a required element for each CSV. If there is no such field in the original file, you would need to create an ID column, and use a function to randomly generate IDs.
Choose a field to act as the Response_Date (the datetime “anchor”) for the upload.
This field usually marks the date that the response was collected, and allows for you to filter by “Date” in the platform.
This is a required element for each CSV. If a Response_date column is not included in your uploaded CSV, you will not be able to filter by date in your Beehive dashboard(s).
If you do not have a column that could function as the Response_date from your third-party export file, you would need to create a Response_date column, and use a function to assign a date to each response. You could also fabricate the Response_date by assigning all responses to the date of your CSV upload. Although all responses will have the same Response_date, this will ensure that you can distinguish between this upload and any future uploaded responses when filtering by date in Beehive dashboard(s).
Essential Formatting
First Two Columns
The order and format of the first two columns in your CSV are very important:
The first column should always be the randomly generated Customer_ID(s).
The second column should always be the Response_date.
Important Formatting Adjustments
Required Rows
Several rows are required to provide Beehive AI's system with sufficient information to properly ingest and format your data.
First row: Include a header row for the names of each data source. Data-source names could be the question asked, the name of a hidden field, or any other identifying feature. The data sources as they are named in the header row will be how they are identified throughout the Beehive AI platform—for example, in dashboard widgets—so we recommend using names that are easily identifiable to you and your team.
Second row: Add a new row under the header row for the clarification of data-source types. This type row will help the data have the correct format when uploaded to the Beehive AI system. There are a finite list of data-source types that you can map the columns to. Please note that these data-source types are spelling and case sensitive. If they do not match the formats in the list below, the data will not be properly ingested or formatted.
Compatible Data-Source Types:
datetime: A timestamp (must follow the formatting specified below)
choice: A single-choice question
multiple_choice: A multiple-choice question (must follow the formatting specified below)
text: An open-ended or free-response question
other: A catch-all data type
Third row: For date-time data sources, insert another row underneath the data-source type row and indicate the format used. Each date-time data source must be formatted the following way: YYYY-MM-DD HH:MM:SS
To properly indicate this format in the CSV, include
{"format": "2006-01-02 15:04:05"}in all date-time columns.
Multiple-Choice Questions
A lot of programs will export multiple-choice questions in a rather unique way. Adjust all multiple-choice questions to ensure correct formatting and ingestion.
If the program you exported your CSV from separates multiple-choice selections into different columns, merge/consolidate the repeat columns for multiple-choice questions
The formatting in the red table below would be incorrect:
customer_id | Which of the following colors do you like? (select all that apply) - Red | Which of the following colors do you like? (select all that apply) - Blue | Which of the following colors do you like? (select all that apply) - Green | Which of the following colors do you like? (select all that apply) - Yellow |
1 | Red | Blue |
|
|
2 | Red |
| Green | Yellow |
The formatting in the green table below would be correct:
customer_id | Which of the following colors do you like? (select all that apply) |
| multiple_choice |
1 | Red|Blue |
2 | Red|Green|Yellow |
Some multiple-choice questions will separate each option with a comma. Change all commas (“,”) to pipes (“|”). Be careful if some of your options are sentences that include commas—you should change commas only if they separate one option from another.
The formatting in the red tables below would be incorrect:
customer_id | Which of the following colors do you like? (select all that apply) |
| multiple_choice |
1 | Red, Blue, I like other colors, but they are not on this list |
2 | Red, Green, Yellow |
customer_id | Which of the following colors do you like? (select all that apply) |
| multiple_choice |
1 | Red|Blue|I like other colors| but they are not on this list |
2 | Red|Green|Yellow |
The formatting in the green table would be correct:
customer_id | Which of the following colors do you like? (select all that apply) |
| multiple_choice |
1 | Red|Blue|I like other colors, but they are not on this list |
2 | Red|Green|Yellow |
Other Essential Checks
The following checks apply to any and all rows and columns in your CSV.
Make sure there are no repeated headers/columns. One data source should be attributed to one column, and one survey/customer response should be attributed to one row.
If some text is too long for a cell, ensure that the text does not continue onto a new line.
Other Formatting Cleanup
The following formatting changes are not required, though they are highly recommended for the sake of clarity in your Beehive AI dashboard(s). While cleaning up your CSV, you have the ability to make certain formatting changes and control how your data is ingested into the Beehive AI platform.
Special Characters/Fields
Certain programs will export data CSVs with improperly coded characters or provide field IDs for certain variables. If you would like these characters and fields cleaned up to make your dashboard easier to read and understand, please make sure to edit these elements.
CSV exports may have improper coding for accent marks, quotation marks, apostrophes, formatting (e.g., boldface, italics), or other non-alphanumeric characters. We recommend changing these encodings to reflect the correct characters.
Original Characters | Clean Characters |
¬øCu√°l es tu color favorito? | ¿Cuál es tu color favorito? |
I don't have a favorite color. | I don't have a favorite color. |
What is your <strong>favorite</strong> color? | What is your *favorite* color? |
If your survey has variable fields in its questions, those fields may have long IDs that increase the length of data source names and may make those names difficult to read in your dashboard. If you would like, you could shorten these fields so they are more easily identifiable.
Original Field | Clean Field |
Why is {field:a1b2c3d4e5f} your favorite color? | Why is [previous color selection] your favorite color? |
Typos
If you notice any typos in your data-source names or hard-coded answer options, please feel free to correct them before uploading your CSV. Otherwise, the typo will be ingested into the Beehive AI platform.
Survey Formatting
If there is certain survey formatting that you worry will be unclear when translated to your Beehive AI dashboard(s), you can change this formatting so your widgets are easier to understand.
For example, you may have an opinion scale question that has answer options from 1 to 5. If you would like to change these options so that 1 becomes "Very dissatisfied" and 5 becomes "Very satisfied," we highly recommend doing so before uploading your CSV.
CSV Template
Use the template attached to this article to guide your CSV cleanup process. This includes all of the essential information we need to properly ingest your data into the Beehive AI platform.
Exporting Your CSV
Each program that allows you to export CSVs of your data will have its own unique CSV format. Here are formatting and export differences to look out for before exporting your data.
Typeform
Export Settings
If exporting your data from Typeform, ensure that you are exporting all responses as a CSV file.
Formatting Warnings
Some of Typeform's CSV formatting is not compatible with Beehive AI's system. The following are some formatting adjustments you will need to make during the cleanup process. (For more information about formatting adjustments, please refer to the section(s) above.)
The ID that appears in the first column of the CSV will not be the same as a customer ID. If you have passed a hidden field that indicates the identity of a customer, please move that column to the beginning of your CSV and delete Typeform's first column.
The Submit Date column (toward the end of the CSV) should be moved to the second column of your CSV. Please ensure that the date is properly formatted.
Multiple-choice questions will separate each chosen option into a different column without a proper header that indicates the question asked. To read about Beehive AI's desired formatting for multiple-choice questions, please read the section above.
Opinion scale questions will not preserve the labels that you assign the numbers of the scale. If you would like to specify the labels of the scales in your survey, please read the section above.
Other columns Typeform includes in its CSV exports may be superfluous to your research (e.g., Network ID, Tags). If there are any columns in your Typeform export that you will not use in your analysis, or perhaps are empty, please delete them.
Survey Monkey
Export Settings
If exporting your data from Survey Monkey, ensure that you are exporting all complete responses as a CSV file. Alternatively, you could integrate your Survey Monkey survey(s) using ingestion tokens, as explained in this article.
You will first need to filter by all complete responses.
You will then need to export all responses data.
In the "Export Survey Data" window, you will need to make sure you are exporting a CSV of the current view of your data. (Please make sure the only filter you have applied is for complete responses.)
Formatting Warnings
Some of Survey Monkey's CSV formatting is not compatible with Beehive AI's system. The following are some formatting adjustments you will need to make during the cleanup process. (For more information about formatting adjustments, please refer to the section(s) above.)
The ID that appears in the first column of the CSV will not be the same as a customer ID. If you have passed a hidden field that indicates the identity of a customer, please move that column to the beginning of your CSV and delete Survey Monkey's first column.
The End Date column (toward the beginning of the CSV) should be moved to the second column of your CSV. Please ensure that the date is properly formatted.
Multiple-choice questions will separate each chosen option into a different column without a proper header that indicates the question asked. To read about Beehive AI's desired formatting for multiple-choice questions, please read the section above.
If there is an "Other (please specify)" option in any question, it may be separated into a different column. Please ensure that there is only one column per question.
The data-source types that Survey Monkey includes in its CSV exports are not equivalent to Beehive AI's accepted data-source types. Please adjust these types as needed.
Other columns Survey Monkey includes in its CSV exports may be superfluous to your research (e.g., Collector ID, IP Address, Device Type). If there are any columns in your Survey Monkey export that you will not use in your analysis, or perhaps are empty, please delete them.
CSV Template
For a guided template of what a clean CSV would look like, please download the following file. Please ensure that your browser settings are not preventing you from downloading any files.
(Note that depending on the third-party app you use to view this file, the format of the datetime columns may have changed automatically. Please ensure that during your cleanup process, you edit all datetimes to the correct format.)