From what I've heard, one unique/distinguishing feature of app marketing is the effectiveness of doing many apps. Angry Birds was Rovio's 57th game I believe.

Many successful app developers seem to follow this approach of doing lots of apps.

I think this is partly because there is a lot more uncertainty and unpredictability in the app market, as well as a lot more very highly arbitrary and industry-specific learning to get through, including the idiosyncrasies of the iOS and Android markets, weird features of the distribution channels, payment mechanisms and ratings models, etc. Compared to other kinds of marketing, there appear to be a lot more specialized hacks, special tricks etc. It's like SEO/SEM, but weirder.

One example: to get more positive reviews, many apps don't prompt the user to post a review until he/she has been using it for a certain period of time. There is even some sort of library function that implements this feature I think. This makes sense in hindsight, but I wouldn't have guessed the importance of this from a priori application of textbook marketing principles or porting from other domains. I learned this from a developer, not a marketer.

Design too seems idiosyncratic. Good design is of course always important in any kind of product marketing, but in most cases you can figure out what's important by porting concepts from other domains. Not so apps, it seems like. Again there seem to be a lot more specific design tricks, which you can only learn by diving into the specifics. For example, there are Apple-dictated norms around what loading screens should look like. How much you can push the norms is something you apparently learn via a few experiments.

To summarize, app marketing is highly specialized. In my experience (I've been involved in one as a consultant and am currently involved in one that I am developing with a few friends who all know more than me), it is the domain that functions the least like any other marketing domain, with the fewest portable lessons from other fields, and lowest applicability of general textbook theories.

Result: you have to learn this as a specialized skill. This is why I think so many app-focused individuals and companies go for the multiple-app route to iterate and learn fast from as many data points as possible.

Finally, I'd say that this is one of the areas where generic marketing and PR firms can add the least value, for all the reasons above (unless they have consciously specialized in apps and can point to a significant track record where they did more than just write press releases or contact bloggers: to hire well, ask them what they did that was specific to app marketing and different from other types). Extra hint: look for those who have been involved in East Asian markets. I am told the dark arts of app marketing have evolved much further there.

Among my own marketing consulting gigs, the one app project over a year ago was where I felt I was able to add the least amount of value. I could probably do a lot better now, but still nowhere close to the value I know I can add in more traditional domains (include Web products).

To summarize the summary, if you do only marketing thing, make it this: make an app just to learn the marketing game. It's the marketing equivalent of the software principle, "plan on throwing one away." Make up something simple and quick, perhaps even slightly silly. Keep the engineering and design as simple as possible. The goal is to use this first app as a sort of radioactive tracer to learn the game. Engineering and even design (at least simple design) is relatively cheap. Invest in a marketing-learning app.