@TourEiffel, I agree that action is better than overthinking. I find that these models can be used as diagnostic tools to quickly identify general areas for research and design.
Suppose you’ve identified a specific problem you want to solve; let’s say you have a “sharing economy” site to get parfaits delivered on demand and the buyers aren’t coming back to rate the deliverer’s service. For people to return, they have to be motivated when they get the trigger to return. A quick glance at the motivation part of the model says that the people most likely to respond to an external trigger are the achievers, who require challenge. So to bring them back, you could test out sending an email with some sort of challenge, perhaps tied to an extrinsic reward, such as site credit. “Rate your parfait delivery and win $1-3 in credit for your next purchase.” Once you’ve used the model to identify something to test, the thinking’s done and the action begins.
Mind you, Achievers (with their extrinsic motivation) are probably the easiest to reach (which I think accounts for the prevalence of points, badges, and leaderboards). So let’s take a moment to wonder what would have happened in the example above if instead of targeting Achievers you decided to aim at the opposite end of the motivation spectrum and target Explorers. Explorers are the most likely to engage with the content for the content’s sake. So to bring them back, you could build curiosity into your service. With the parfait’s delivery, you might give the customer a puzzle or a question about an obscure fact from parfait-making history. To find the answer, they could click on a link that takes them to the answer on a page to review the service. Once again, the model took only a moment to think about, and you have a new hypothesis to test.