Is there a best naming tactic or set of tactics? No. Tactics don’t automatically predict outcomes. But there is a best naming strategy. That’s because a name is an invented concept that integrates all salient aspects of the brand and that ultimately comes to represent them (sometimes together with a tagline).
The best naming strategy, therefore, begins by looking at what a product/company does, what benefits it offers, whom/what it competes with, to whom it offers them, and what (in an ideal world) one wants evoked in a customer’s mind when he is exposed to that name.
That is, you start with the end, not the means; with the evoked, not the evoker.
Once you’ve got a handle on those issues, you can codify the attributes of a brand, develop a positioning/messaging statement and see the big picture. Only then can you start developing names and establishing measurable criteria against which to measure prospective names.
Google is a great name for a number of reasons (for Google, if you’ll excuse my reflexivity), but it’s not a very good name for an ultra-conservative bank. The lesson: Context is everything. Without it, there are no great names or awful names, just boring, phonologically weak, meaningless ones.