I studied marketing (with an emphasis on e-commerce) at Uni. In a professional capacity I've put together and executed campaigns for various online and offline

I studied marketing (with an emphasis on e-commerce) at Uni. In a professional capacity I've put together and executed campaigns for various online and offline businesses. Tl;dr - Start by mimicking your competitors (or the next best thing), test whatever you do, and document whatever you do. If you're looking for a book I'd check out Steve Blank's "The Four Steps to the Epiphany," which contains useful worksheets, including one to develop channel strategy and pricing model. You can preview the first few chapters of this book for free at Stanford's website [1]. An aside: Blank also published a sequel to this book called, "The Startup Owner's Manual," which I haven't read, but which I've heard includes updated and electronic versions of the worksheets [2]. Marketing is similar to engineering in the sense that you need well-honed problem solving skills if you want to drive predictable and repeatable results. The process by which you achieve these results will vary depending upon your product, your competitors (if you have competitors -- or substitutes or compliments, if you don't), your target market, your industry (or industry segment), the economic climate, the tools at your disposal, among other things. More concretely here are a few things I like to keep in mind when planning and executing a campaign: 1. Quoting Stravinsky, "Lesser artists borrow; great artists steal." If you're not first to market study your competitors. Become their customer. Find out how they market to their customers. Hypothesize what might [be|not be] working for them. Mimic what you think might work. If you /are/ first to market use the above approach to study complimentary products or services, or substitutes. 2. Always be testing Once you've found a promising idea, test it. Ideally tests would achieve statistical significance, but this is not always possible. Furthermore, commit yourself to multiple tests (e.g. A/B tests, multivariate tests), but don't be afraid to trust your gut if something is obviously [working|not working]. Use whatever analytics tools are appropriate for the job and create reports for yourself based on these data (e.g. Google AdWords, Optimizely, etc.). See below for more on the latter half of this point. 3. Keep data organized Have a written plan when you begin. And think of marketing your product like you would the code base for a well-written and long-lasting program. Comment your steps along the way so your "future self" will know exactly how you achieved certain results. (Note: documenting what didn't work can be as important as documenting what did (think of a bug tracker as an analogy).) In essence, your marketing documentation and data should serve as a compendium of what worked and didn't work so you can build on successes and prevent unnecessary mistakes. And like code comments it can help bring someone else up to speed if your product is successful and you add to the team. Links: [1] http://www.stanford.edu/group/e145/cgi-bin/winter/drupal/upload/handouts/Four_Steps.pdf http://www.stanford.edu/group/e145/cgi-bin/winter/drupal/upl... [2] http://www.stevenblank.com/startup_index_qty.html http://www.stevenblank.com/startup_index_qty.html