Knowledge Center

Knowledge Center > Insights > Developer Reference > Guidelines for Optimal Insights Performance

Guidelines for Optimal Insights Performance

This article covers guidelines for getting optimal performance out of your Insights implementation -- everything from how much data can be brought into Insights to optimal usage of APIs and Connectors.

General Guidelines

This section outlines how much data you should plan to bring into Insights for optimal product performance and user experience.

 

 

Guideline

Number of data sources

Plan to bring in data from three to six data sources to get the most out of your Insights implementation.

 

At a minimum, bring in data from the following three types of sources:

  • Zuora: for financial and subscription data

  • Your CRM: for customer demographic data

  • Your product: for customer behavior data

 

Many customers also find great value by bringing in data from the following additional types of sources:

  • Your data warehouse: for additional customer data

  • Your web analytics tool: for additional customer behavior data

  • Your help desk system: for customer support cases data

  • Your marketing platform: for customer marketing interaction data

Number of accounts

500K or fewer total records

Number of users

500K or fewer total records

Number of attribute fields

(across account or user record types)

50 or fewer total fields

Number of metric fields

(across account or user record types)

100 or fewer total fields

Number of event types

100

Number of custom event attributes

(across all event types; customer event attributes are event attributes beyond those attributes required by Insights)

100

Number of new events per day

10M

Number of new sessions per day

10M

 

API Guidelines

Batch API

The Batch API is the primary API customers use to send data into Insights.

 

 

Guideline

Number of new batches in a 24 hour period

When calling the API, plan to send in no more than:

1 new batch per data source per day

Maximum batch size

(gzip compressed file format)

1 GB

Minimum batch date that will be processed by a dataflow

3 years prior to date of dataflow execution

Streaming API

The Streaming API is used for bringing events into Insights in cases where the Batch API is not a viable option. For example, some data sources do offer the ability to extract historical data. With such systems, Insights has to capture data from it in real-time. All event data captured via the API takes up to two hours for Insights to process before it becomes available for use in the product.

 

 

Guideline

Maximum number of requests in a 24 hour period

10M

Maximum number of requests in a one second period

200

Maximum number of records that can be included  in a request

1

Connector Guidelines

Zuora

Insights brings subscription data about your Accounts in from your Core tenant once per day.

 

 

Guideline

Fetch frequency

Runs once per day

(typically between 12am-1am in your local time zone)

Minimum batch date that will be loaded into Insights

3 years prior to date of Dataflow execution

Salesforce

Many Insights customers use Salesforce as a CRM. The Insights Salesforce Connector enables you to bring in data from Salesforce into Insights once per day.

 

 

Guideline

Fetch frequency

Runs once per day

(typically between 12am-1am in your local time zone)

Outbound segment sync frequency

2 hours

Outbound field sync frequency

2 hours

Number of Account Segments that can be synchronized

Limit to no more than 15

Number of Account Fields that can be synchronized

Limit to no more than 20

Number of User Segments that can be synchronized

Limit to no more than 15

Number of User Fields that can be synchronized

Limit to no more than 20

 

Last modified
19:23, 11 Oct 2017

Tags

This page has no custom tags.

Classifications

(not set)