Data build tool (dbt) has introduced packages specific for Fivetran that enriches the data for refined goals in when a connection is established with the source. When these packages are executed, additional tables are added in the destination which otherwise requires intensive manual effort to generate them. These packages are available for the sources GitHub, Salesforce, NetSuite, Mailchimp and Marketo.
With these packages, Fivetran data from your sources also gets enriched:
- To have descriptions over tables and columns
- To have recent data from the source to be considered by conducting freshness tests
- To have model staging tables to be used in transform package
- To have primary key tests for unique and non-null values
In a collaborative environment like GitHub contextual information like assignees and history for issues and pull requests are maintained in such a way that it is easier to define API extracts for a specific one. As an organization, if you are interested in analyzing all your underlying issues and pull requests you need to manually collect details all the issues and pull requests
With the GitHub package, you can create a separate table to track issues and pull requests which creates visibility over the velocity of the codebases over time and avoids the need to create separate process logs for the same. This helps to track assignee – assigned issue ratios, set a time out value for a pull request, and to track time taken in each stage of a pull request. It maintains information about the assignees and time of completion for the issues and tracks these metrics for daily, weekly, monthly, and quarterly levels.
When data is extracted from the sales force API it is often difficult to track the relationship between the account owner, their manager, their accounts, and related opportunities as the data retrieved is not categorized properly
Salesforce package creates separate tables of performance metrics automatically and monitors the sales team’s performance health and assists the managers/executives to track the sales funnel of their team members while saving development time. Opportunity table stores contextual account information about an opportunity and its relevant details including its owner, snapshot table creates a view of sales funnel, owner and manager table tracks booking counts, amounts, win-loss percentages about individuals and teams respectively by leveraging the opportunity table.
Currency conversion or generating yearly balance sheets and income statements are time-consuming tasks in any application.
NetSuite package, if you select the tables in Fivetran, it automatically generates balance sheets and income statements, transactions with converted amounts and in addition to this, it also fetches all related tables that match accounting and reporting periods to create commonly used data models. It also provides an opportunity to combine NetSuite data with several other data like sales or account data to gain a full view of the business.
Mailchimp’s API extracts data like opens, clicks, and unsubscribe but it doesn’t identify to which granularity these items belongs to. For example, all recipient information is extracted at the campaign level but it is not available to segment levels.
Mailchimp package transforms recipient and activity tables into analytics-ready models and provides aggregate metrics for separate granular levels including campaigns, lists, segments, members, and even automation. It also maintains an activities table which carries information about each campaign like when the email was sent and the lag between send and the activity.
With Marketo, there are several options to track user activities but it’s a complex process to associate it with every single activity generated to aggregate a parent activity like the campaign.
Marketo package provides separate tables for the campaign, email sends leads, and programs to be enriched with metrics about email performance, such as opens, clicks, and unsubscribes. It also maintains a lead history table that shows the state of leads on every day thus provides historical records of daily changes.
When Fivetran dbt package is executed, several automatic tables are created, you can directly create reports over the resultant data generated from this package by connecting it with the reporting tool. The packages enable the standardization of tables and join the tables for the data models.
Such packages clearly help the user a great deal in terms of simplifying their effort and untangles some common challenges they might face in terms of accessing their data from their source. We are expecting more packages from Fivetran dbt which will be covered in our upcoming blogs.