Why 100% is a starting point, not always the answer
The clean, easy-to-explain default
100% annual offset is the most intuitive target: your system produces what you use, with no more math required. Where a utility still credits every exported kWh at the full retail rate, this target is usually the most cost-effective one, since there is no real economic reason to undershoot or overshoot it. How many places still work that way keeps changing, as states revise net metering rules year by year, so check your current rules rather than assuming one. The DSIRE database entry for net metering lists the policy in force where you live.
Why the real answer often isn't exactly 100%
Two forces push away from the clean 100% number in opposite directions. Reduced-export net metering policies, covered directly by the net billing vs net metering comparison, make the marginal panels between 90% and 100% less valuable, since exported power earns less than retail value, a case the 90% offset tool covers in detail. Meanwhile, planned electrification like an EV or heat pump, plus ordinary panel degradation over time, both push toward sizing above 100% instead. The right answer depends on which of these forces applies more to your specific situation.
How small the panel gap between tiers actually is
Because panel count scales directly with the offset percentage, the difference between 90%, 100%, and 110% is smaller than it might sound. For a home needing 20 panels at 100%, the full range across all three tiers spans only about 4 panels, from roughly 18 at the low end to roughly 22 at the high end. That is a modest cost difference to weigh against a real economic or planning question, one the loan payment calculator can translate into an exact monthly figure if financing is part of the plan.
What actually decides between the tiers
Your utility's net metering structure, your roof's available space, and whether an EV or heat pump is on the horizon all matter more than any general rule of thumb. Degradation matters too. NREL's compendium of photovoltaic degradation rates found median degradation of 0.5 to 0.6% a year for crystalline silicon modules, so a system sized at exactly 100% of today's usage produces about 95% of it after a decade. A 110% target roughly absorbs 20 years of median degradation, which is why the 110% offset tool frames it as the future-planning case. If your utility caps system size, the net metering cap tool works backward from that limit instead of a target percentage.