The following frameworks have been designed to help startups to choose the right traction channel to start with.

Brass Framework

David Arnoux of Head of Growth at Growth Tribe
http://growthtribe.io

This framework helps you prioritize which customer acquisition channels you should test first. It is based on two premises:
1)  the answer to the question “What should we try to drive users to our website?” lies within the startup. It is simply a matter of asking the right questions.
2) the more people you get involved in the acquisition channel brainstorming and voting, the better your chances of choosing the right channels to test.

Step 1. Get people together
The more people you get involved in the process of selecting customer acquisition channels, the better. Diversify the types of people (marketing, technical, CEO, analysts)

Step 2. Brainstorm user acquisition techniques
Mix traditional methods with “outside of the box” ideas. Try to be as detailed as possible.

Step 3. Apply the BRASS criteria (get the template here)
You create a spreadsheet with all your ideas listed in the first column and a brief description in the second column. Then you are going to give a score (1 (worst) to 5 (best)) voting the following:

  • Blink. Gut Feeling. It’s your intuitive score. What is your initial intuitive feeling about this channel?
  • Relevance. Product/Channel Fit . Think about how relevant is this technique to your target audience and to your product?
  • Availability. Ease and Cost. How easy is it to set up this channel and how long it will take? How much will it cost? Do we have the necessary resources?
  • Scalability. How scalable is this channel?
  • Score. Multiply your scores together.

Usually, you will come out with 3 or 4 clear winners, and these are the channels you should test first.

Step 4. Design experiments to test the clear winners you have obtained in step 3. The experiments should be time-boxed, have a quantifiable goal, minimum success criteria to validate or not the experiment, and be clearly detailed.

ICE Framework

Sean Ellis founder GrowthHackers, growth at Dropbox
Building a Company-Wide Growth Culture: SaaStr Annual 2016

ICE can help you assess and prioritize acquisition channels. Like the BRASS framework, you start by getting ideas from everyone on your team. Then, instead of using Blink, Relevance, Availability, and Scalability, here you score based on:

  • Impact. The expected outcome. If this idea works, how big will be the impact on your company?
  • Confidence. The level of confidence you have that the idea will work
  • Ease. How easy is it to test this idea?

You give every idea an ICE score that goes from 1 to 10. Rank your ideas according to the ICE score so you can easily see which ones you should start testing. The best ones are those that are easy to test and will have a high impact on your business.

The Bullseye Framework

Justin Mares and Gabriel Weinberg authors of the book “Traction: How Any Startup Can Achieve Explosive Customer Growth”

The bullseye framework forces you to take all the traction channels more seriously and to uncover strategies for getting traction that you wouldn’t have found using other approaches. As the other frameworks, this one also starts with brainstorming.

Step 1. Start by brainstorming every possible traction channel you can think about without dismissing any channel.  Think of at least one idea for every channel. The objective of this step is to identify one decent channel strategy that has a chance of moving the needle.

Step 2. Run cheap traction tests in the channels that seem most promising to determine if the idea really is good or not. Spend no more than $1,000 and a month of time. These tests should answer the following questions:

How much will it cost to acquire customers through this channel?
How many customers are available through this channel?
Are the customers you are getting the ones you want?

These tests are not meant to get a lot of traction. They are simply trying to determine if it’s a channel that could move the needle for your startup.

To get some test ideas read: 32 Fast cheap marketing test ideas across all 19 traction channels

Step 3. For this step, you have already determined a core traction channel. Now, you need to focus on that channel and avoid distraction with other marketing efforts. If after testing, no channel seems promising you need to repeat the whole process.

Brian Balfour Framework

Follow these 5 steps to choose a traction channel.
http://www.coelevate.com

Step 1. Define what you are optimizing for. Common possibilities are:
Learning. You are trying to learn something. This is a priority when the startup is pre-product-market fit or when you are expanding to a new target audience.
Volume. You want to increase the amount of customers. It is a priority if you need a minimum amount of users/data to reach a broader goal.
Cost. You want to optimize your cost per acquisition.

Step 2. Define the constraints. Common constraints are:

  • Time (Do you have a specific optimal window to reach your audience?)
  • Money
  • Target Audience (Do you need to reach someone really specific?)
  • Legal

Step 3. Set up your channel matrix
This can help you identify the pros and cons of each marketing channel. Using a spreadsheet, list out all potential channels vertically (include the alternative channels, like StumbleUpon, Reddit, email sponsorships, app stores, etc).

Certain channels can be broken down into detailed tactics. For example, content marketing can be broken down into blogging, guess posting, webinars, ebook, whitepapers, etc.

In the header row fill in the attributes for each channel (targeting, cost, input time, output time, control, scale)

  • Targeting. The depth of ability to target different audiences
  • Cost. The upfront and ongoing cost to acquire users in this channel
  • Input time.The upfront time required to start running experiments with this channel
  • Output time. The time it takes to start getting data from your experiments
  • Control. The control you have over turning the channel on/off at will. The more control you have over the channel, the easiest and fastest to run experiments.
  • Scale. The size/reach of the channel

Step 4. Fill in values of channel matrix with low, med, high
Do research to get enough information to make educated comparisons of each channel relative to each other.
Some research methods you can use: Channel specific tools (For example, Google Keyword Tool (SEM), SEMrush (all PPC), Moz (SEO), MixRank (Display)), talk to people that have experimented with the channel, use  Google for channels like blog sponsorship.

Help to value each channel/attribute combination when doing your research:
Targeting. Does the channel have the type of targeting you need? How close to your exact target audience can you get with this channel? How many different ways to target users are there?
(For example, Facebook ads is “high”, PR is “low”)
Cost. Are there upfront costs? Minimum spends? Pay as you go?
Input Time. How much time is required to get to the first experiment? How much time to run each additional experiment? Is there any technical work involved to get started? Creative/content creation?
Output Time. How much time will it take to start getting results? How many experiments do we have to run to get meaningful data?
Control. Can you turn the channel on and off at your will? (For example, PR is low control, Facebook or Google ads are high control)
Scale. How large is the channel in overall volume? How scalable is it in terms of your time/resources?

Step 5. Choose 1 – 2 Channels as hypotheses
In a perfect world you want something like this:

Targeting – High, Cost – Low, Input Time – Low, Output Time – Low, Control – High, Scale – High

Once you have the matrix completed, how do you prioritize?
Use your constraints (step 2) as a guide. If money is a constraint, look at channels with low cost.

At this point, you have a hypothesis about a customer acquisition channel. You need to set up and run a number of experiments to prove out the viability of the channel in relation to your goals.