In lighters, it converts mechanical into electrical energy. This doesn't consume the Pvdf, nor does it alter significantly - within engineering limits.
Pvdf is not strictly piezoelectric. It's ferroelectric - but after having been formed mechanically and electrically, it behaves like a piezoelectric material, and is much more efficient in that use than normal piezo materials like BaTiO3, PZT and others. You may also consider copolymers and terpolymers of Pvdf. Ceramics are less attractive: brittle, expensive, heavy metals...
The copolymer poly(vinylidenefluoride-trifluoroethylene) is often written as P(VDF-TrFE).
After stretching, a Pvdf chain has "all" (simplified!) fluorine on one side and all hydrogen on the other side, so negatively charged fluorine give the chain a side polarization. Mechanical deformation pushes these charges, resulting in high voltage and, for a short time, current - good for a spark.
I had vaguely hoped that polyvinyl dichloride would be ferroelectric as well, for similarity with Pvdf, and cheaper - but the bulky chlorine atoms must prevent straight chains. Stretching and trying wouldn't be difficult. I suppose it fails, because PVC also would be ferroelectric, but I never noticed that on electric cables, while it's easily observed with Pvdf insulation.
Better alternatives? I doubt! Pvdf is already tough and rather cheap. It makes a simple lighter.