Segmentation is one of the cornerstones of marketing in almost all contexts. If you are buying TV time, you are buying time on a particular network, for a particular time slot and for particular content. Print ads are placed in specific publications that have their own demographics and their own regional editions. Direct mail goes to specific geographical areas or demographic groups. These are all segments.

Each of these choices is driven by the principle that a campaign will be more successful if it reaches a receptive audience. If you know your audience, your marketing can be targeted more accurately and your message will be more relevant to its recipients.

Email marketing is no different, but segmentation is something that can be overlooked with email, in part because a massive email campaign does not require the deep pockets that are a prerequisite for a large-scale TV campaign. The temptation, then, is to send everything to everyone.

At the other extreme, the ideal campaign would be truly individualized. Each customer or prospect would receive a completely customized message that spoke to his unique interests, preferably with a personal aside asking after little Timmy’s progress on the soccer team.

Between these two extremes lies the kind of list segmentation that is both practical and valuable.

On the practical side of the scale, lists do not have to be segmented according to each distinguishing point that can be extracted from the data. You do not have to analyze every click.

Of course, a mailing list will not segment itself, and you need to do the work to arrive at segments that are meaningful. Simple parameters are best, especially if you can devote some time and energy to testing those parameters against user response.

Segmenting and testing take time and energy, but they do not need to be complicated. In fact, too much complication can be counterproductive. For example, if you test two different subject lines against two segments, you have four variables to study. With four segments, the same test has eight variables, a smaller population per segment and, generally, results that are less statistically reliable.

The best approach is to keep it simple. Start small with the data you have on hand. Prospects versus customers is a good place to start, but you can expand along other parameters like clicks, opens and deferrals.

On the value side of the scale, Lyris, Inc. surveyed marketers in 2010 and found that 39 percent reported increased open rates, 28 percent reported lower unsubscribe rates and 15 percent reported both lower spam complaints and increased customer acquisition.

There are costs associated with email segmentation, but they can be more than offset by gains in many areas that impact your bottom line. If you keep segmentation simple, you can control costs and still achieve meaningful improvements in your marketing.