Caching needs careful planning. How many times your application needs a specific data or sets of data? As amnah mentioned, you need to also consider the caching object size.
Basically, evaluate the costs of performance impact to understand what to cache first?
One other thing very very important is USER Experience. Caching can be great for performance, but may impact user experience, if you do not put the dependency right to reset the cache.
Kartik V gave some high level advice, but I think you’d do better with some real, practical advice.
Stop caching. Completely. Stop it right now, because you don’t need it.
I know what you’re thinking. I know how you feel, I’ve been there. It’s so inefficient to be making the same call on every page, surely this can be done better!
That’s true. But here’s the reality of it: Those basic select statements? They’re nothing. Short, simple, few, efficient, and optimized via your db. They don’t need to be cached.
So when should I cache then?
It’s simple! Just follow this great checklist my mentor gave me.
1) Don't cache anything
2) Wait until website becomes slow
3) Identify bottlenecks
4) Fix bottlenecks by optimizing the algorithms
5) Algorithms optimized? Now you start caching!
In your case, I’d say you’re still at #1; you don’t need to do anything yet.
Starting to get slow? Then fix your algorithm. For example, why are you making three ajax requests? Make one.