Alright folks, I'm diving headfirst into the world of IAM and looking seriously at Ping Identity. As someone who usually lives in the land of data pipelines and API gateways, I know enough to be dangerous but not enough to avoid all the potential pitfalls 😅.
We're a mid-sized shop, and our use case is pretty classic: need to unify access for a growing suite of internal apps (some home-grown, some SaaS), a couple of data analytics portals, and eventually maybe customer-facing apps. We're also heavy into event-driven architectures, so the thought of streaming auth events to our data lake for security analytics is already making me a bit giddy.
I've done the demos, read the datasheets, and I'm now at the "talking to sales/engineering" stage. But I know from bitter experience with other enterprise tools that the devil is in the details you *don't* think to ask about upfront.
**So, for those of you who've been through a Ping implementation (or even a PoC), what questions did you wish you'd asked *before* signing?** I'm thinking beyond the obvious feature-checklist comparisons.
Here's where my head is at, but I know I'm missing angles:
* **Deployment & Operational Reality:** They talk about "flexible deployment," but what does the day-to-day operational overhead *actually* look like? Is the containerized deployment (PingDirectory, etc.) truly cloud-native or does it feel like VMs in disguise? How painful are upgrades?
* **Developer Experience:** How seamless is it for our app teams to integrate? Is the documentation for the SDKs and APIs as good as the marketing claims? Can I easily mock the Ping services for local development?
* **The "Gotchas" in Scaling:** We might start with 500 users, but plan for 50k. What bottlenecks emerged for you as you scaled? Was it licensing, infrastructure, or something architectural?
* **Event Streaming & Data Integration:** I want to pipe all authentication and authorization events into our data warehouse. How easy is it to configure and maintain that stream? Are the events well-structured, and is the schema stable?
* **Hidden Cost Drivers:** Everyone knows about per-user licensing. But what about costs tied to MFA usage, API call volume to the directory, or support for certain protocols (SAML vs. OIDC)? Did any particular feature or requirement unexpectedly blow up your quote?
I'm especially keen to hear from anyone who's integrated Ping into a data-heavy or streaming ecosystem. Did it play nicely with your Kafka pipelines or your monitoring stacks (like Prometheus/Grafana)?
Any wisdom you can throw my way would be hugely appreciated. I'd rather look a bit naive here now than be blindsided six months into an implementation.
Data nerd out.
Data nerd out