Can consumer behavior be changed? Of course. Is it easy to do? Not at all.
As a general reaction, I’ve found that when VCs and other non-marketing business people look at consumer-facing services offered by new startups, they often fall into one of two polar camps:

1. Overly skeptical about the ability to facilitate the adoption of a new service whatsoever. “It’s too tough to change consumer behavior” is a mantra which I’ve heard numerous times.
2. Overly positive about the chances for the adoption of a new service based on a small dataset of reactions from familiar people (like their own, their families, or the early adopter / techie TechCrunch crowd).

Of course, healthy skepticism should be applied to all new endeavors (especially those with a consumer angle), and someone’s own gut is always a strong datapoint. However, over-generalizing in either direction leaves little opportunity to discover interesting opportunities which will be successful. All consumer-facing services change behavior, at least slightly, by definition. The key is to understand the context of this modification.
So in looking at consumer-facing companies I try to place a framework around evaluating the willingness of people to adopt a new service. To me there are fundamentally two levers:

1. The nature of the steps which deviate from existing behavior. The two primary questions are determining how many steps consumers must take in adopting a new service and how incrementally different those steps are from current behavior. I’ve written in the past that the friction between desired consumer actions is multiplicative; in other words, the more difficult it is to progress from one step to the next, the likelihood of a consumer completing this progression compounds.
2. The visibility of the value of taking those steps for the target consumers. Of course, there has to be value for the consumers to take each individual step towards fully adopting a new service. But it goes beyond what the true value of the service is to the end-user. In reality, adoption is driven by what s/he perceives the benefit of taking each individual next step in the process. And evaluation should be around consumers in the target market of the service, not the founder’s, VC’s, early adopters’, or other constituent’s mindset.

An interesting example is contrasting the adoption of TiVo and YouTube after their introduction – two different ways that consumers could (and now do) watch video vs. “traditional television.” First, TiVo users had to make a huge leap of purchasing an expensive item with a subscription fee. And while already adopted users sang the praises of the service, those who hadn’t already had a difficult idea perceiving those benefits because they were nebulously communicated. By contrast, YouTube users were already online and comfortable with accessing photos and text in its respective forum. (Once the barrier of adequate bandwidth had fallen,) the ability for users to experiment with watching one video or even uploading a video was relatively low. While this comparison is perhaps an oversimplification, looking at these cases through the above lens does provide some perspective about the number of years it took / is taking for these services to become adopted.
At the end of the day, my point is not to promote the above framework as a specific tool. Rather, my message here is that consumers are just like any other business’s customers – they have a set of needs and preferences, and we can apply analytical thinking around their potential behavior beyond just our gut feel. It’s difficult to separate our own feelings when we ourselves could be a possible customer. In reality, we’re only one among a market of (hopefully) millions.