The first step is to have Parrot and Ponie hit maturity. Then sell Ponie as a free speed increase with an amazing ability to reuse libraries written for other languages. Yes, there will be conversion bugs, but they're managable.

Once people have converted to Parrot, then begins the Perl 6 conversion process. If it is as productive as people think it can be, that process can start with people writing modules in Perl 6 rather than Perl 5. Companies may find themselves using the Perl 6 modules without needing to use Perl 6 directly. Once you have enough modules like that floating around, there is a direct incentive for people to learn Perl 6. Or to use it for internal projects. And then the language can begin to grow.

However the details of how that goes can wait. The first step is getting Parrot and Ponie to a point where we can convert people to Ponie. Then we can start creating a clear value proposition for Perl 6.