The more distributed the system landscape, the more important it is for our customers to have sensible and, above all, comprehensive reporting.The operational reporting for the customer service organization takes place directly within Zendesk (Zendesk Explore). Business Intelligence (BI)-Tools are used for scenarios where the data from Zendesk alone is not sufficient (such as Tableau, Looker, Power BI, etc.).
You need ticket content from Zendesk in your BI database or in your data warehouse (DWH) in order to relate it to data from other sources (such as CRM, ERP, shop …).However, the data must first be transported so that it is also available there. There are basically two approaches, push & pull.
Push vs. Pull
Push – means something like event-based sending of data from Zendesk to a backend system. In the case of activities on a ticket, this can be sent to an external system as soon as an event occurs (webhook). Zendesk configures a trigger in such a way that required information is sent to an API endpoint. This is ideally located in the same environment as the database itself and receives a JSON, which then creates a data record in the BI database.
Pull – on the other hand, means that you “pick up” the data. Such tasks are usually carried out e.g. once / several times a day and thus get the latest information from the Zendesk API. Zendesk has its own API endpoints for this purpose (incremental endpoints). If the data is not time-critical, we prefer “pull”, as this gives you access to a large number of other data (users, organizations, group names, metrics, etc.). Pull is also ideal for on-premise databases.
We can easily find out which approach is best for you and which fits the IT architecture.
Best-Practice / Entities
- Tickets (including custom fields)
- Ticket comments
- Ticket metrics
- User (inquirers and agents including custom fields)
- Organizations (including custom fields)
At “PULL” we usually create low-maintenance scripts in Python, which then fetches the data from Zendesk e.g. 1x a night and writes it to the designated database.
It doesn’t really matter which database is used here (MSSQL, Oracle, Postgres, Snowflake, AWS Redshift, Google BigQuery …).
- It is possible to report data from your ticket system across companies
- Questions such as “How does total sales relate to ticket volume” can thus be answered (ticket statistics from Zendesk / sales information from the ERP or shop system)
- Technically, there are quite simple methods for achieving the goal
- No more manual reports when it comes to creating cross-source evaluations