Should i use a single API endpoint or split into profiles, orders, and history?

[Eric35]

I'm trying to decide if I should use a single, all-encompassing endpoint for our customer data or break it into separate ones for profiles, orders, and history. My gut says separate calls are cleaner, but I keep wondering if the performance hit from multiple network requests outweighs the benefits of a well-designed aggregate resource.

[Aubrey27]

I went through this last year and started with a single resource for all customer data It felt clean at first but the payload got unwieldy and we ended up splitting by profiles and orders later

[Steven.R]

I wonder if the real issue is how we model the data not the number of calls Maybe a good cache and fetch strategy would hide the extra traffic

[Paisley.J]

I measured latency after moving to separate calls and saw a small jump in per call time But the total time for a full customer view dropped when we could reuse cached profile data

[AaronR]

Either way im not convinced we are solving the right problem yet I want a quick pilot but is this really the core issue?