What's the fastest way to know which marketing channels work?

I'm reading the book «Traction» by Gabriel Weinberg and Justin Mares now. There is a cool framework for it. 1. Brainstorm with a team and pick several traction channels that seemed promising. (ex. targeting blogs, publicity, search engine marketing»). 2. Then run a series of cheap tests in each (ex. sponsor a small newsletter, contact celebrities, place some Google ads) to see what worked and what didn’t. Keep track of the test results in a spreadsheet. 3. After running these experiments focus on the traction channel that seemed most promising and that «could move the needle» for its traction goal. For example, targeting blogs is your core channel. In the early days, the channel strategies of sponsoring mid-level bloggers in your niche and guest posting can allow you to acquire your first fifty thousand customers.

I would start with a couple of things: 1. Identify the main KPIs (that you can accurately measure) and make sure you are measuring them on each platform. 2. Try to create as level of a playing field as possible to compare platforms. - Every platform has it's own unique way of doing things and you need to make sure you have an equal presence on all before trying to evaluate (and none is OK) - Ex: How you market on Reddit is VERY different than Facebook. - Make sure you are only looking at new customer acquisition and not existing - you don't want to compare apples to kumquats. - Make sure you are comparing platforms that are not pay-to-play to be successful in the organic growth.