Pick One Channel and Test It: Lessons From TractionPick One Channel 21 October 2026 · 4 min read · 84 reads
84readsMarketing · Business · Decisions
Pick One Channel and Test It: Lessons From Traction
Weinberg and Mares list 19 ways to reach customers. Their advice is not to try them all, but to test a few and commit to the one that works.
A founder with a growing business usually does not lack ideas for getting customers. She has too many. Someone suggests an influencer. Someone else swears by trade shows. A cousin sells WhatsApp broadcast software.
The result
Earn Embers for what you read
Sign in, answer a few quick questions about this piece, and earn Embers you can spend in the app.
is
familiar:
a
little
of
everything,
and
no
clear
lesson
about
any
of
it.
The 19 channels
In Traction, Gabriel Weinberg and Justin Mares set out 19 channels for reaching customers. They include search engine optimisation, search ads, public relations, unconventional PR, offline ads, email marketing, viral marketing and engineering as marketing, among others. The point of the list is range. It stops you from assuming that social media is the only door.
The authors built the book on interviews with founders who had found traction, and one theme repeats: the right channel for one business is often a poor fit for another. You cannot copy a competitor's channel and expect their result.
The Bullseye idea
The authors offer a framework they call Bullseye. Summaries disagree slightly on how many steps it has, so treat the following as the spirit rather than a script. The idea is to work through the channels in rounds:
Flow: List every channel > Rank by promise > Test the top few cheaply > Commit to the winner | narrow from many to one
Brainstorm what each channel could look like for you, even the ones that seem silly. Rank them by how likely they are to reach your buyer. Run small, cheap tests on the top few. Then put your effort into the one that clearly performs, and ignore the rest for now.
How to run a cheap test
A test is not a campaign. It is a question with a deadline.
For a manufacturer in Pune, say, it might read: "Can email outreach to 50 purchase managers produce five replies in two weeks?" For a restaurant it might be: "Can a tie-up with two nearby offices bring 20 lunch orders a week?"
Set three things before you begin:
- The cost you will allow, in ₹ and in hours.
- The number that would tell you it is working.
- The date you will judge it.
When the date comes, look at the number honestly. A channel that fails the test has still taught you something cheaply.
Why you should resist spreading out
Each channel has its own learning curve. Writing for search takes months. Outreach calls take weeks to get good. If you divide your attention among six, none reaches the point where it works.
Focus here is a decision to say no. It is easier to say no when you have a clear test result than when you rely on a hunch.
Watch the ceiling
A channel that works today may saturate. Ads get costlier as more advertisers arrive. A partner may reach only so many people. When results flatten, you go back to the list and test another, while keeping the working one running. This is a slow rhythm rather than a one-time choice.
A single good week does not prove a channel. Look for results that repeat across a few weeks before you commit your budget.
What the pros do differently
Experienced owners keep a short written log of tests: channel, cost, result, decision. Within a year, that log is worth more than any consultant's opinion because it describes your buyer in your market.
Fit the channel to your economics
Before picking, check the arithmetic. If an average sale earns you ₹2,000 of profit, a channel that costs ₹3,000 per customer cannot be your main road, however popular it is. If a customer stays for years and orders again, the same ₹3,000 might make sense. This is why knowing what a customer is worth comes before choosing where to spend.
Take one honest hour with your numbers and write down the most you could pay to win a customer. That figure will rule out some channels immediately.
Try this
- Write all 19 channels you can recall or look up, and mark the five where your buyer is most likely to be found.
- Choose two. Design one test for each with a budget, a target number and a date.
- Run both for two weeks, then keep one and drop the other.
- Record the result in a simple sheet.
Comments
Sign up free, read the story and mark it finished to join the conversation.
Loading comments…