Velocity scripting will cause your email to take at least 3 times longer to go out. It's not 3x, Jep. Any such penalty is extremely minor -- when it exists at all in practice. Remember that the assembly process parses Velocity even if you don't explicitly include any Email Script {{my.tokens}}. It's not like you avoid instantiating the Velocity context. We've done extensive benchmarking on this, and the TTFD (Time To First Delivery) and delivery spread (TTLD - TTFD) is effectively identical for an email asset with no tokens at all vs. an email with simple Velocity logic applied to a single field, like the above. Underlying differences in assembly time end up blending into the background noise. With in-depth testing, you'll see that 10,000 emails with a simple Velocity token may assemble, queue, send, and be 100% delivered in less than the time it takes 10,000 empty emails to even start to go out from the same instance. Depends on what else is going on at the time. 100,000 emails with Velocity can be faster, slower, or effectively the same as 100,000 non-Velocity emails. Of course you can do a wide range of things with Velocity. On the low end, you might just output static text. On the high end, you can inject lots of lead and CO properties into the context (naturally requiring more RAM) and perform sorts, filters, date math, object casts, et al. Those high-end features can add milliseconds or even full seconds onto the elapsed time, and eventually there is an effect with large sends, but using Velocity at all is not the worry. I agree it would be best to save this particular "cleaned up" field permanently rather than regenerating it every time, since it's a very slowly-moving data point. You could create the permanent field via webhook or via SFDC. Even though the Velocity overhead is negligible, the permanent field can be used in other places that don't support Velocity, like Landing Pages, SMS, etc.
... View more