Anyone using Zoho People as their core HRIS? Pros and cons

Having evaluated numerous HRIS platforms for mid-sized tech organizations (200-1500 employees), I've found that discussions around Zoho People often lack the operational depth required for a serious technical evaluation. While it's frequently positioned as a cost-effective alternative to platforms like Workday or BambooHR, the critical question is whether its architecture and feature set can reliably serve as the **system of record** for all people data, which then feeds downstream payroll, finance, and analytics systems.

My team conducted a three-month proof-of-concept, integrating Zoho People with a bi-directional payroll sync (via API) and a separate data warehouse. The core findings, broken down by operational category, are as follows:

**Reported Advantages (The "Pros"):**
* **API Coverage & Flexibility:** The REST API is sufficiently comprehensive for core HR objects (employees, departments, time-off). We were able to build a custom orchestration layer (in Prefect) to sync employee lifecycle events.
```python
# Example: Fetching updated records via Zoho People API (Python pseudocode)
import requests
def get_updated_employees(since_time):
url = "https://people.zoho.com/api/employees"
params = {
"index": 1,
"pageSize": 200,
"lastModifiedTime": since_time
}
headers = {"Authorization": "Zoho-oauthtoken "}
response = requests.get(url, params=params, headers=headers)
return response.json().get("data")
```
* **Cost-to-Feature Ratio:** For basic HRIS functions (employee directory, org charts, document management, simple workflows), the platform is competitively priced. The built-in modules (like Performance Management) provide baseline functionality without additional per-module costs common in other suites.
* **Zoho Ecosystem Integration:** If your stack is already within the Zoho ecosystem (Books, CRM, Desk), the pre-built connectors reduce integration latency and vendor management overhead significantly.

**Operational Constraints & Grievances (The "Cons"):**
* **Audit Log Limitations:** The granularity of audit logs for administrative changes is inadequate for strict compliance regimes (e.g., SOX). We encountered difficulty tracing the lineage of specific field changes (like compensation or job title) over time through their API, requiring cumbersome manual exports for audits.
* **Workflow Engine Bottlenecks:** While customizable, the workflow engine for approvals (e.g., promotions, salary revisions) does not handle complex, multi-branch logic well. It becomes brittle at scale, and there is no native way to export workflow metrics (cycle time, bottleneck analysis) for observability.
* **Data Export & Observability:** Scheduled, automated exports of full datasets for backup or data warehousing are not as robust as needed. The system lacks native webhook support for critical events (e.g., termination), forcing a polling architecture that introduces data freshness latency.
* **Support SLAs for Critical Issues:** During a payroll-cutoff simulation where our sync job failed due to an API schema change, the time-to-resolution for a support ticket escalated to engineering exceeded our required SLA. The documentation did not reflect the change, leading to production downtime.

**Benchmark Considerations:**
When measured against a framework of data integrity, pipeline reliability, and operational visibility, Zoho People scores adequately for organizations with straightforward HR processes and no complex international payroll requirements. However, for entities requiring rigorous data lineage, advanced workflow analytics, and predictable integration SLAs, the platform introduces non-trivial overhead that must be managed by internal data engineering resources.

I am keen to hear from other data or IT leads who have managed Zoho People as a core system. Specifically:
* What has been your experience with the API's stability and rate limits during bulk operations?
* How have you architected external monitoring for the platform's health and data freshness?
* Are you leveraging the native analytics, or have you piped all data into a separate warehouse for reporting?

-- elliot