In this article:
Benefits of collecting from Onna
An increasing number of our customers are connecting to Zendesk for different use cases (ex: legal hold, litigation collections and pii detection). This article provides tips and tricks based on the experiences of Onna customers with large Zendesk instances. For a more detailed, technical overview of collecting from Zendesk, please refer to How to collect from Zendesk.
How to Optimize Your Zendesk Collection
We always recommend speaking to your Customer Success representative before starting a Zendesk collection. They can provide specific advice on how to optimize your collection strategy depending on your use case and desired outcome. The following are prerequisite questions that will help during this strategic discussion:
- How many Zendesk instances does your organization have?
- If multiple instances, does each one have its own Admin?
- How many tickets are in each instance (and overall)?
- What type of subscription do you have?
- Although no particular subscription level is required, lower-tier subscriptions have a lower API rate limit which could impact the speed of a collection.
- Order of ingestion:
- Onna ingests Zendesk data by the date last modified - not creation date.
- Selection criteria
- Onna enables you to create a Zendesk source based on date (ex: Jan 1 - Jan 31), but not by organization or custodian. One will need to ingest the targeted Zendesk data before you can search for requester, key terms, etc…
- Onna also allows you to collect deleted or suspended tickets that are still available in your Zendesk instance (not permanently deleted).
Optimizing the ingestion speed
Because it can take weeks or even months to ingest large Zendesk instances, we always recommend speaking to your Customer Success representative before initiating your first Zendesk collection. However, the following describes our general recommendation for customers ingesting large volumes of Zendesk data.
- We recommend creating 1 workspace for your Zendesk data and creating multiple Zendesk sources that are broken up by a period of time (Year, Quarter or Month).
- Depending on your subscription level, we have found that each source collects around 30,000 resources (tickets or attachments) per day. So, if you have a source for 2021 that would ingest around 30,000 resources in one day, but if you had 12 sources (Jan-Dec) for 2021, you’d expect to ingest around 360,000 resources in one day. Thus, we often recommend monthly syncs for large Zendesk instances.
- We recommend one-time syncs for each month, and creating an auto-sync & archive sync with an open-ended “sync end date) for the beginning of the next month. So, if it’s July 15, 2022, create monthly syncs for every month prior to July 2022 and create 1 sync that is July 1-July 30, then the last source would be August 1 - with no end date and in auto-sync and archive mode so that it will collect on an ongoing basis. Eventually, all the one-time syncs will complete and you will only have 1 ongoing sync (the Aug 2022 onwards source).
This represents a general collection strategy that has been successful for our customers, but our specific recommendations might differ depending on your subscription level, number of instances, and number of Zendesk tickets, so please reach out to your Customer Success representative to discuss your specific use case. If you are not sure who your Customer Success representative is, please contact firstname.lastname@example.org.