Here’s three factors to keep in mind when running a CVM campaign.
Reach
Reach is obvious. The more people you target, the more you can potentially convert.
There is a cost involved and so you will want to limit to those who most closely match your target profile.
But given that even the best run campaigns will yield a conversion of no more than a few percentage points, reach is definitely worth considering carefully.
Repetition
How often do you send out a followup message to a target customer ? The answer is usually: not many - but repetition does help. The entire advertising industry is built around it
Once you have gone to the trouble of profiling and targeting a prospect, it is worth trying at least once more time before moving on - and perhaps more depending on the product you are selling.
But this needs to be handled with care. You must have a system in place to ensure that people who have opted are not annoyed with another followup and that the followups are paced appropriately to rekindle recall, but not overwhelm.
Relevance
Relevance to the customer - not to you.
Relevant to you is simple: you want to target customers who have not tried X before, or who have been dormant for 3 months, or who are high spenders - that's called a qualifying criteria
Relevant to customer is different: Does a customer actually want X ? It depends on what's going on in their minds and their lives when the message reaches them. And even Google with all their data collecting and advanced algorithms doesn’t know enough at this level.
So absolutely spend a lot of time fine tuning the target segment, but remember that this is an inexact science. It is also domain specific and market specific - so domain experience will be handy here.
Practically, the best way to build relevance is experiment and post-analysis. The more campaigns you run, the more systematically you analyze the results of each campaign, the better your relevance will become.
One caveat: It is always tempting to increase relevance by offering freebies - and if you're trying to generate trial or to bring back dormant customers that's OK - but its not sustainable. The whole point is to get people to make a purchase decision because they discover something they would have missed before.