A new channel with strong promises always appears. Deciding whether it is worth testing requires criteria, a short test and an exit rule defined beforehand. Without them, each new channel becomes an expense without conclusions, or a distraction from what already works.
The essentials in a definition
Deciding on a new channel requires clear criteria, a 30-day test with defined limits and an exit rule. Instead of following trends, assess whether the audience is there, whether the message fits, and whether you can produce and measure. The approach of setting objectives, measuring and reviewing, close to NIST risk management, combined with conversion measurement and studies such as those from Think with Google, reduces fashion-driven decisions.
Before testing: criteria
Ask: is my audience there? Does the channel serve my objective (awareness, leads, sales)? Can I produce suitable creatives? Can I measure the result? Is the minimum cost manageable? Do I have capacity to serve if it goes well? If several answers are no, postpone.
Compare with the potential of channels you already use and can still improve.
The 30-day test
Define a small, fixed budget, a period (for example 30 days, or whatever is needed for a minimum volume), one or two hypotheses and success metrics. Create native creatives. Prepare measurement (parameters, pages, codes). Record everything.
Avoid changing many things during the test.
Deciding on a new channel in four steps
- Assess criteria before spending
- Run a short test with a fixed budget
- Define the exit rule in advance
- Record learnings and decide
The exit rule
Before starting, write down the conditions under which you continue, adjust or end. For example: “if cost per qualified lead is below X, I scale; if between X and Y, I adjust one variable; if above Y, I end”. This rule avoids emotional decisions afterwards.
Also set a maximum loss limit.
A new channel is a question; the test is the way of answering it.
Learn and decide
At the end, review the results against the defined metrics. Record what you learned, even if you end. If you decide to continue, plan scaling in steps. If you stop, use the learning (audiences, messages, creatives) in other channels.
Keep a list of channels to watch and review it quarterly.
Decision table
| Step | Question | Decision |
|---|---|---|
| Criteria | Is the audience there? Can I measure? Do I have content? | Proceed only if three answers are yes |
| 30-day test | Fixed budget, hypothesis and metric | Define before launching |
| Assessment | Does it meet the success criterion? | Continue, adjust or stop |
| Exit rule | When to abandon? | Stop without guilt if there are no signs |
Common mistakes to avoid
- Entering because it is fashionable. What serves others may not serve your audience.
- No test deadline. The test stays open and spends without a decision.
- Changing the criterion afterwards. Fix it beforehand to avoid after-the-fact justifications.
How to measure the result
Record criteria, budget, results and decision of each test, and compare the channels tested over time.
Limits and exceptions
A short test may not capture the effect of discovery channels that act over time, and initial performance is often worse. Adjust duration and metrics to the type of channel. The decision to stop also costs: consider opportunity cost. Review criteria with experience, keeping the discipline of defining them beforehand.
Illustrative example
Illustrative example: a company evaluates three new channels with the same criteria and picks one to test for 30 days with a fixed budget and a success metric. At the end, the result does not meet the criterion and the company decides to stop. The decision was easy because the rules were written, and the record serves future assessments.
Frequently asked questions
How much budget should I allocate to tests?
A small fraction of the total (for example 10 to 15%), with clear limits.
How long should the test last?
Long enough for a minimum volume of data, generally a few weeks.
What if the test is inconclusive?
Decide by the exit rule. You may extend once, with a criterion.
How long should the test last?
Long enough for useful data; 30 days is a reference, adjustable to the purchase cycle.
What if the result is ambiguous?
Define a rule: extend once with a clearer criterion or stop.
Assess new channels with criteria, a 30-day test with a fixed budget and an exit rule set beforehand, to decide by evidence and not fashion.
Sources and recommended reading
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.