The old model
For years the arithmetic was simple. You generated more than you used during the day, exported the surplus, and were credited at a rate that made the payback straightforward. Storage was an expensive optional extra.
That model has changed. California now values exported solar considerably less than it once did, which shifts the advantage decisively toward consuming your own generation rather than selling it back.
Why that favours batteries
A Valley house generates most of its power in the middle of the day and uses most of it in the evening, when everyone gets home and the air conditioning is still working hard. Without storage, that mismatch means exporting cheap and buying back expensive.
A battery closes the gap. It shifts midday generation into the evening peak, which is precisely when electricity costs the most. The economics depend on your consumption pattern, but for a household that is out during the day and home in the evening, storage is doing real work.
The non-financial argument
Public safety power shutoffs and heat-driven grid strain are facts of life here. A battery keeps essential circuits running through an outage, which is a benefit that does not show up in a payback calculation but tends to matter a great deal the first time it is needed.
Worth being clear on scope though: a modest battery covers essentials, not an entire house running air conditioning. Sizing should be an honest conversation about which circuits you actually want backed up.
Sequence the roof first
One practical point that has nothing to do with economics. Panels outlast roof coverings. Mounting an array on a roof with a few years left means paying to remove and reinstall it when the roof is replaced.
If your covering is aging, re-roof first. It delays the solar project and saves considerably more than it costs.
Exporting is worth less than it was, so storage is now part of the conversation rather than an upsell - and the roof goes first if it is aging.