If you subscribe to a newspaper or magazine, do you read every issue cover to cover?

Or, do you read just the articles that interest you most, as I do? I get The Wall Street Journal every day, and I bet I do not read 80 percent of it.

I ask because among all the new things mobile makes possible, nothing is more alluring than the promise of personalization. Having a way to green light only information, content, products, services and ideas someone likes, while blocking everything else, sounds like an incontrovertible way to improve every user’s experience.

Mobile marketers should make a point of reading PSFK’s mobile commerce playbook in its entirety. Page 14 includes some marvelous examples of mobile personalization services. Users just share their preferences, and in a few swipes these apps and websites will surface stuff that matches them.

As glorious as this may sound, I’m here to sound a note of caution on personalization.

There are two types of personalization: Echo Chambers and Discovery

At a simplistic level, there are two different types of personalization. There’s the kind I’ll call the Echo Chamber, which shows me primarily stuff that it knows I like. Pandora’s online music is a good example. The service asks listeners to tap or click a thumbs up or thumbs down icon as it serves up songs based on a user’s favorite artist or genre. Pandora doesn’t take long to figure out your favorite musicians and play predominantly more songs by those musicians.

Then there’s personalization that leads users to new things. Let’s call this mode Discovery. Netflix suggests new movies and shows based on subscribers who have watched the same things my wife and I have. Netflix is like that friend who likes what we like but always can be relied upon to recommend something we haven’t seen, something that he knows we’ll enjoy.

Echo Chambers strike me as a safe approach in the short term. For example, feeding readers suggestions based on authors they like and ideas with which they agree can make for contented customers.

But what happens when their tastes evolve? Echo Chambers have walls, and if a user breaks through a wall, the whole apparatus might crumble.

On the other hand, Discovery is harder and riskier. It requires making more uncertain attempts at recommending what your users may like next. It won’t always work. Some may abandon you as a result.

Get Discovery right, though, and you are more likely to grow as your users grow. Long term loyalty may be possible only through Discovery.

Echo Chambers and Discovery represent two poles. Most personalization approaches live between them. Pandora will occasionally play a new song, for example. I just find that Netflix is more likely to recommend something totally new.

As you can probably tell, I clearly favor Discovery, both as a matter of mobile strategy and as a customer. I rarely listen to Pandora anymore; even after nearly two years as a premium subscriber, I felt like I heard the same tracks over and over. On the other hand, Netflix has recommended movies that have made me feel like a cinema maven when I recommend them to my friends (in spite of serving up a few duds).

If your mobile or digital plans include investments in personalization, what approach will you take: the comfort of the Echo Chamber or the ambition of Discovery?