Table of Contents
CaliberMind Insights - Application Concepts and Terminology
Updated by Cathy Funderburg
Data models may be files or databases located on servers on a local network, or remote locations and web services. CaliberMind supports a variety of data sources, among which are:
- High-Performance databases such as Snowflake, Redshift, Big Query
- Traditional relational databases, such as SQL Server, MySQL, or Oracle
- Online web services, such as Salesforce.com, Google AdWords, Google Analytics, Zendesk, and more
- File-based data sources, such as spreadsheets (Excel) and CSV files
These sources are supported through native connectors, customer REST connectors, and JDBC drivers, which are used to import data into the ElastiCube Server.
Common Source Connection Types
- Live Data Sources - Live connections connect directly to the data source. Use a Live connection if your data changes frequently and your dashboard must reflect this.
- ElastiCube - Import your data into the ElastiCube to take advantage of CaliberMind's high-performance database.
Technology Partner Connectors
Technology partners enable connections to data sources via the data pipeline.
CaliberMind technology connectors include:
Apple Search Ads
ElastiCube Data Hub
The ElastiCube Data Hub allows you to easily integrate data from multiple disparate data sources into a single source of truth, and enable viewing both near real-time and historical data from your sources.
The ElastiCube Data Hub provides the ability to deliver on your organization’s analytics strategy even when changes occur to underlying source systems. This includes not waiting for the completion of data warehousing projects to perform analysis. The data model can easily be adapted to changes in the underlying data architecture by easily reconnecting to changing sources systems ensuring continuity. Working with the ElastiCube Data Hub, you can fully utilize data from across your organization to ensure that you are seeing the full picture.
Data models are abstract entities that organize your data and determine how your tables relate to one another. CaliberMind has two types of models, Live models, and ElastiCube models. Which model you use depends on how you connect to your data source. If you want to run queries directly against the data source, you will use a Live model. If you are importing your data into CaliberMind, you will use an ElastiCube model.
Live and ElastiCubes models are created and managed from the Data page.
Live models are a type of data model used to manage the schema over your Live data source. Live connections are useful for getting real-time updates, with changes in your data reflected in your dashboard.
ElastiCubes are specifically designed to withstand the extensive querying typically required by business intelligence applications.
Each widget is a dynamic visualization of data. You pick the type of data to appear in a widget and you pick the type of visualization (chart type). A few examples of widgets are displayed below.
Other Widgets include: Accordion, Area Chart, Area map, Box and Whisker Plot, Calendar Heatmap, and Scatter Map.
Dashboards are composed of various widgets which visualize the data you select and customize. It can be used for a variety of purposes, but its primary purpose is to provide easy access to information such as KPIs.
- Calibermind provides you a starting point with several templated dashboards such as Attribution, Engagement, Performance, Return on Ad Spend, and System.
- You define which widgets appear in the dashboard, their design, how they are organized, and the filtering of the data that appears.
- Calibermind allows you to create as many dashboards as you need. No limitations! When you create a dashboard, you are the dashboard’s owner. Owners can share dashboards that they have created with other Calibermind users.
- Calibermind provides a variety of built-in automatic dashboard sharing features. You can easily share a dashboard with others in the Calibermind environment or have a dashboard automatically delivered to your chosen recipients by email (on a scheduled basis or upon each data update).
Widgets are composed of fields that represent the data in your data sources. You create widgets by simply selecting from the fields displayed in a Data Browser, which appears in various places across the product. Each field represents a column of data in the Live Models. Each Widget is a Dashboard component and a dynamic visualization of data.
Each Field represents a column of data in the ElastiCube.
Numeric Fields: Numbers (quantified data), such as salaries, sales, scores, number of clicks. This is data that you may want to aggregate or calculate. For example, the sum of sales or the average of costs.
Date Fields: Dates can describe both date and time values. Dates can be used to organize your data into hierarchies according to year, quarter, and month or into buckets of time such as by the hour or by 15-minute intervals.
Descriptive Fields: Items used to label and categorize, such as Products, Locations, and other categories.
The Filters panel on the right of the dashboard provides data filtering options that affect the data displayed in the widgets. You can change these filters to focus on more specific data. Filters are used dynamically to interact with dashboards
For more on Filters, visit our Filter documentation.
Aggregation Types can be used in dashboards when you want to toggle between views.
For example, our out-of-the-box Attribution Summary Dashboard allows you to select whether you want to view the underlying data by opportunity closed date or opportunity open date. These views have their own use cases, which you can find more information on here.
The attribution model outlines how sales and conversions should be credited along the conversion path.
If you have a sizable data set (200-400 Closed Won opportunities) or have worked with your customer success representative to train your model against meetings or pipeline generated, we recommend using the Chain-Based model. This machine learning model looks for the most common pattern or sequence of events and weights activities accordingly. This model removes human bias and reflects the most common winning scenario.
Finally, the Even Weighted model is good for organizations that don't want to dictate which touches get extra weight applied based on sequence or person. This model naturally gives more weight to the touches that more people gravitate to rather than applying any assumptions.
First touch: The First-touch attribution model attributes 100% of the credit to the first marketing touchpoint (the channel or activity through which the contact was acquired).
If you want an estimate of which campaigns are initially engaging prospects that convert into revenue or pipeline, we recommend using the First Touch model. This looks at all contacts and leads associated with an account and chooses the touch that happened up to 365 days prior to opportunity creation.
If you want an estimate of which campaigns are encouraging people to engage with sales or signal an opportunity should be created, we recommend using the Middle Touch model. This looks at all contacts and leads associated with an account and chooses the touch that happened immediately prior to opportunity creation.
If you don't have a lot of Closed Won opportunities and have confidence in your sales team's ability to consistently identify the primary contact on the opportunity, a W-Shaped model may be the right fit. This model takes into account all touches associated with the opportunity but applies additional weight to the first touch, middle touch, and any touches associated with the primary contact. If your sales team does not use contact roles, this model turns into what is known as a U-Shaped model and weights the first and middle touch more heavily, which can still be useful.
We have additional articles that give you more information on Engagement Terminology and Key Concepts, using Dashboard Templates, with details on drilling down and filtering data.