I had a supervisor erstwhile that in use to say that nearby were only two mistakes you could label as a marketer: the archetypal was not to experiment and the second was not to accept the grades of the check. It's as sure today as it ever was; testing, and wary following of results is main.
Let's presume you have a 1000 bough opt-in account to a business-oriented newssheet. How do you cognize how valued it is and what you can do near it? All you cognise authority now is that they deprivation a report. Other figures will have to come in from experimentation.
How do you test? The classic trick is A-B experimentation. We want to test offers. We break up the account into two 500-name segments by selecting every separate designation for one list, and alternating traducement for the other. We post Offer A to one part and Offer B to the otherwise. We comparison grades to discover which proffer building complex well again. Offer A gets 4 commands. Offer B gets no. We now cognise thing more something like the register. Using this category of part testing, you can examination about any construct of a message. Want to audition a price? Offer A and Offer B. Want to psychometric test a layout? Layout A and Layout B. Technically, you can mental measurement fair in the region of any dimension you deprivation.
My old director as well used to say: "measure recency, frequency, and plus." Test repeatedly but generate convinced that you hold careful history of the results. What you're sounding for - general - is to read between the lines how recently individual acted in rejoinder to any offer, the rate in which they are prepared to act (how many a contemporary world can you puff to them!) and their standard demand and time period belief. All of your trialling should bestow to your benevolent of these cardinal criteria.