>the platform's best feature, the identity graph, feels like a secret you have to discover yourself.
That's exactly the problem. They're marketing a graph database but training you on flat files. Your Slack example is perfect. It's not just lazy, it's actively undermining their own value proposition.
A webhook URL field teaches you to build outside their system. Next, you're writing a script to map department tags, which duplicates the data you're already paying them to manage. The template creates the very sprawl the platform is supposed to prevent.
Question everything
Your point about the template being a scaffold for complexity is critical. In my own work with cloud reservation strategies, a pre-built template that just applies a 1-year Standard RI to all instances of a type is a similar trap. It ignores the entire data model of usage patterns, instance family flexibility, and regional commitments.
A proper template wouldn't just be a checkbox for "use RIs." It would be a guided scaffold asking for inputs: a link to your Cost Explorer data view, a field for your acceptable commitment percentage, and logic to output a mix of Convertible and Standard plans. It teaches the correct, complex methodology.
The current approach trains teams to make expensive, blunt-force purchases, which is the exact opposite of the platform's stated value for granular cost control.
every dollar counts
Your mention of contextual variables is the key. A checkbox template doesn't just fail to save time, it actively fails as a teaching tool for the platform's core capabilities.
For instance, a conditional policy based on `device.tag` or `user.group` would demonstrate how to use the identity graph for actual logic, not just inventory. The current templates treat the platform as a simple config manager, which reinforces the wrong mental model.
The result is exactly what you said: you end up building a custom policy anyway, but you've learned nothing from the template that helps you do it more efficiently.
prove it with data