Advanced Tools: N-Tiles, Volumetric, Multiple Add, Custom Audiences, Large Row Analysis

N-Tiles

N-Tiles, also known as quantiles, are a statistical concept within Explore used to divide a dataset or distribution (like age, income, or a satisfaction score) into equal intervals or sections. The core idea is to partition the data into 'n' equal-sized groups, based on the data's values arranged in ascending order. This allows for a deeper analysis of how different segments of the continuous variable behave relative to the intersecting variable in the crosstab

Key Use Cases for N-Tiles

N-Tiles are a powerful tool for segmentation and benchmarking in data analysis, particularly in market research and statistical reporting:

1. Segmenting Responses:
2. • Goal: To understand how respondents with low, medium, and high values on a metric (e.g., spending, usage frequency) differ in their attitudes or behaviors.
   • Example: Using Quartiles (n=4) on a "Total Monthly Spend" variable to create four groups: Lowest Spenders (Q1), Lower-Mid Spenders (Q2), Upper-Mid Spenders (Q3), and Highest Spenders (Q4). These groups are then crosstabulated against brand preference to see if the highest spenders prefer a different brand than the lowest spenders.
2. Benchmarking and Ranking:
3. • Goal: To establish performance benchmarks by dividing the population into top and bottom performers.
   • Example: Using Deciles (n=10) on a "Customer Satisfaction Score" (CSAT) to identify the Top 10% (10th Decile) and Bottom 10% (1st Decile) of customers. The characteristics, feedback, or demographics of these two extreme groups can then be compared to inform strategic action.
3. Handling Skewed Data:
4. • Goal: To analyze data that may be heavily skewed (e.g., income, website visits) by creating balanced comparison groups.
   • Reasoning: N-Tiles ensure that each resulting segment has an equal number of observations, providing reliable comparison points even if the raw data distribution is irregular.
4. Simplifying Analysis and Reporting:
5. • Goal: To reduce the complexity of continuous data into a manageable number of categories for easier interpretation and presentation.
   • Benefit: Instead of reporting results for every single value of a scale (e.g., ages 18 through 65), the data is summarized across a few meaningful N-Tile groups (e.g., three Tertiles).

Types of N-Tiles Available in Explore

Explore provides several options for partitioning data, generally based on splitting percentage values:

When using the percentile option, users can control the range by adjusting the plus or minus icons up to 10 increments and down to a minimum of one increment.

Process of Using N-Tiles in Explore

The process involves accessing the functionality either through the Crosstab settings or the Coding Grid. In both methods, the N-Tiles are generated and added directly to the table.

The process requires the user to select the item they want to serve as the basis for creating the N-tiles. They can also define a custom title for the resulting coding statement. The generated tiles are then added right below the target item in the table.

Method 1: Using the Crosstab Settings

1. Click on the Crosstab settings button located on the top left of the crosstab table.
2. Click on N-Tiles.
3. The N-Tiles popup will appear.
4. Choose the target type (table, rows, or columns), targets, and the specific N-Tiles type (Tertiles, Quartiles, Quintiles, Deciles, Percentiles).
5. Click 'Generate' to create the N-tiles.

Method 2: Using the Coding Grid

1. Navigate to the Coding Grid.
2. Click on the Coding Grid Kebab dropdown.
3. Click on N-Tiles and choose the relevant settings in the N-Tiles pop-up.
4. The N-Tiles will be generated.

Volumetrics

The Volumetric Modal in Explore is a dedicated tool accessible from the Codebook section, designed to allow users to apply specific numerical and statistical calculations to selected codes before adding them to a report's cross-tab.

The Volumetric Modal supports calculations typically involving numerical values, such as volume, time, or money spent, or the number of visits or amount of product consumed.

The specific statistical calculations available through this modal include:

• TOTAL
• MEAN
• MEANZ
• MEDIAN
• MEDIANZ

The Volumetric Modal retains the same layout and functionality as previous versions of Explore.

Process for Using the Volumetric Modal

The Volumetric Modal is accessed when a user chooses to add items from the Codebook to the report and selects the Volumetric method of addition.

The process to use the Volumetric Modal is as follows:

1. Select Codes: The user must first select the desired codes, categories, or sub-categories from the Codebook by clicking the corresponding checkbox.
2. Initiate Addition: Click on the Add To buttons for the desired location in the report (rows, columns, or table bases).
3. Choose Volumetric Method: Select the Volumetric method from the options presented. This action opens the Volumetric Modal.
4. Define Calculations and Output: Within the modal, users can choose which calculations (TOTAL, MEAN, MEDIAN, etc.) they wish to apply. Users can also select whether the results should be:
5. • Combined into one row, column, or table base.
   • Separately calculated, with each calculated code appearing on its own line.

The modal also includes options for statement titles and coding previews, allowing the user to assign a custom name to the resulting coding statement and view the actual code that will be applied to the cross-tab.

Once the settings are configured, the user clicks 'Generate' to apply the calculation and create the resulting row, column, or table base item.

The DEC() option also allows the users to define the decimal places of the final calculation

Add to Multiple

The "Add to Multiple" Modal in Explore is a specialized function within the Codebook designed to apply selected codes simultaneously (en masse) to multiple existing items within the codebook.

The modal is an addition method available when transferring selected items from the Codebook to the report's cross-tab. Its primary function is to simplify the application of selected survey answers or codes to several pre-existing items in the codebook.

Process and Use

The process of using the Add to Multiple Modal is initiated when adding items from the Codebook:

1. Selection: A user selects the desired codes, categories, or sub-categories from the Codebook.
2. Initiation: The user clicks the "Add To" buttons (for rows, columns, or table bases).
3. Method Choice: The user selects the "Add to Multiple" method from the options presented (which also include Separate, Combined, Volumetric, and Count). This action opens the "Add code to multiple modal".

Once the modal is open, it allows selected codes to be applied to multiple items within the report.

Custom Audiences

Custom Audiences are a powerful feature in Explore that allows users to save codings made on rows and columns to be reused later. They let users build out a defined audience or target profile and quickly apply that structure across other compatible surveys.

Process for Saving and Managing Custom Audiences

The process to save a custom audience involves creating the desired coding and then using the dedicated save feature:

1. Code the Segment: Create the crosstab and code the rows and columns that define the desired audience.
2. Initiate Save: Access the ‘Settings’ button in the top right and select ‘Save audiences’. Alternatively, the clicking on the kebab menu next to the row or column item and clicking Save Custom Audience option allows users to save any columns, rows, or table bases as a custom audience.
3. Define Location and Name: Assign a filename and choose the columns and rows to save. Users must select the drive, either their personal drive or a shared company drive.

Use Cases and Benefits

• Audience Reusability: The core use case is the ability to build an audience once and quickly reuse that audience across other surveys from the same provider.
• Centralized Management: The Manage Custom Audiences modal acts as a central location where users can view all custom audiences created across their entire account.
• Editing and Organization: Users can edit, duplicate, rename, export, move, or delete custom audiences.
• Compatibility Check: The system ensures that users can only edit or apply custom audiences that are compatible with the active survey. A filter option allows users to view only the audiences compatible with the currently active report.
• Collaboration: Saving custom audiences to a shared drive allows the entire agency access to use the same complex coding statement, provided they are using the same survey.
• Media Planning Integration: Custom audiences can be selected and sent to the Plan application using the "Send to Plan" tool, enabling the planning team to work with audiences created in Explore.

Snapshots

Snapshots are a tool within an open report designed to save a temporary version of the document for easy later reference.

The Snapshots functionality is accessed via a dedicated button located in the right section of Explore’s Top Bar.

Clicking this button reveals a drop-down menu that allows for the following actions:

1. Create a new snapshot: This saves the current state of the report.
2. Apply an existing snapshot: This allows users to choose from a list of previously saved snapshots to apply those settings back to the active report.
3. Manage Snapshots: This opens a modal where users can rename snapshots, delete them, and export selected snapshots into an Excel format.

Use Case for Snapshots

The primary use case of Snapshots is to provide a temporary version of a report that can easily be referenced at a later point in time.

Snapshots are useful for capturing different stages of an analysis or setting up specific views that a user might want to revisit without necessarily saving them as permanent, fully named Reports or duplicating the main report file.

Large Row Analysis

The Large Row Analysis (LRA) function in Explore is a specialized tool designed for efficiently handling and distilling extensive datasets, particularly when dealing with large categories of data in the cross-tab. It allows users to quickly analyze categories that might contain many thousands of items, extracting only the most relevant rows for further reporting.

Explore supports a maximum cross-tab size of 5,000 rows and 150 columns. The LRA feature is essential for working with data sets that approach or exceed this practical limit, as users typically only want to focus on the top-performing items.

Use Case for Large Row Analysis

The primary use case for LRA is to help users identify and focus on top or bottom items within large categories.

Specific utility includes:

• Quick Analysis of Large Categories: It provides a method to quickly analyze a large category or group of categories to find outliers or key trends without needing to scroll through or process the entire dataset in the default cross-tab view. For example, a category might contain 14,000 items, and LRA allows the user to efficiently select only the top 100.
• Data Filtration: It allows users to determine the top or bottom items based on specific criteria, such as index values or audience reach.
• Report Focus: It ensures that the user is left with a shorter list of the more important things (rows) that fit their analytical criteria, which is critical when analyzing extensive survey data.

Process of Using Large Row Analysis

1. Access the Function: LRA is accessed via the table settings icon within the cross-tab interface.
2. Select Data: Users select the specific large categories they wish to analyze.
3. Specify Quantity: Users specify the number of top or bottom items they wish to retain from the selection (e.g., top 100 items).
4. Define Criteria: Users select the specific metric to sort by, such as percent column.
5. Apply Filters: Users can apply filters (e.g., filtering out any items with an index less than a specific value, such as 30).
6. Preview: A preview tab allows users to review the items selected by the criteria before finalizing the export to the cross-tab.
7. Send to Cross-Tab: The user decides whether to add the results to existing rows or to clear and replace everything in the current cross-tab with the new, filtered selection.

Large Row Analysis does not work with volumetric items, and in cases where a category has pre-calculated means, medians, and totals this can potentially skew the analysis if a user attempts a blanket grab of that category.