83% of SharePoint PDFs not indexed despite EAP enrollment — possible folder depth or crawler issue
Despite being enrolled in the PDF Content Allowance EAP, a large number of PDFs from our SharePoint connector are not being indexed by Zendesk.
Note: We have already confirmed that our SharePoint connection is listed as a search source under Knowledge Admin > Search Settings > Search Sources (237 items for the main portal, 5 items for the inquiry directory).
Here is a summary of our investigation.
== PDF Sync Status ==
- Total PDFs in SharePoint: 591
- Synced to Zendesk: 99
- Not synced: 492 (~83%)
== Additional Context ==
The majority of unsynced PDFs already existed in SharePoint before we enabled the EAP (approximately 3 weeks ago). This suggests the issue is not limited to newly added files, but also affects pre-existing files that should have been indexed during the initial sync.
== Comparison of Synced vs. Unsynced Folder Structures ==
Folders with higher sync rates (confirmed on Zendesk side)
- DocLib1 (legal document library): 17 files
- Shared Documents/006_業務ガイド/情報セキュリティ/...: 41 files
- Shared Documents/004_社内システム/...: 13 files
- Shared Documents/005_制度/...: 14 files
- Other: 14 files
Folders with low or zero sync rates (confirmed on SharePoint side)
- Shared Documents/人事発令/ (FY2022–2026): 188 files
- Shared Documents/200_規程・内規: 48 files
- Shared Documents/INFOMATION添付資料: 29 files
- Other unsynced folders: 227 files
== Questions ==
(1) Folder depth limit — most likely cause Synced PDFs appear to be concentrated in folders with a maximum depth of 5 levels. We have 29 PDFs located 6–7 levels deep in the SharePoint folder hierarchy, and none of them appear to be synced. Is there a folder depth limit for PDF ingestion?
(2) Scope of synced libraries and folders Large folders such as 人事発令 (personnel announcements), 規程・内規 (company regulations), and INFOMATION添付資料 (information attachments) are entirely unsynced. Are there any restrictions on which libraries or folders are included in the sync scope? Is there a way to specify individual folders in the connector settings?
(3) Impact of unsupported file types We have noticed that some folders contain unsupported file types (e.g., .png, .rar, .zip, .pptx). Based on a similar report in this community, it seems that the presence of unsupported files may cause the crawler to stop mid-process, preventing other files (including PDFs) in the same folder from being indexed. Could this be a factor in our case? Is there any logging available to identify which files were skipped or caused the crawler to abort?
(4) How to verify EAP activation status We have confirmed that we are enrolled in the EAP, but we are not sure how to verify whether the PDF ingestion feature is actually active and functioning correctly. Is there a way to check this?
== Background / How We Got Here ==
We were initially directed to post in the "Zendesk EAP: Knowledge: PDF Ingestion" group by Zendesk Support, who referenced a similar thread there. However, when we attempted to post, we received the message: "This user does not have access to the category." So here we are.
As for the support process itself — we reached out to Zendesk Support first, and after some back-and-forth, the response we received was essentially: "Please post in the community." The irony of a company that sells customer support tooling responding to support requests by redirecting users to a community forum — with no troubleshooting steps offered — was not lost on us. We understand EAPs come with limited SLA coverage, but a templated deflection response is a rather on-the-nose demonstration of exactly the problem Zendesk's own AI is supposed to solve.
We're posting here in good faith and hope someone from the product team can shed some light.