New Marketing Channel Test Decision
"Should we add TikTok?" - the question can be answered with real budget or upfront on the model. A new channel doesn't bring customers right away: it has a ramp-up time, a saturation limit and a contribution that can be estimated before the first dollar flows. The model shows two things - whether the channel carries at all, and whether it beats the existing mix.
Why trying it out with real budget costs more
The live test costs before it gives answers
Simply letting a channel run ties up budget and time - and only delivers a first signal after the ramp-up phase. If it turns out badly, the money is already spent. The expensive part of learning comes before the insight.
What the upfront simulation provides
The model estimates the expected contribution from benchmark ranges, saturation and ramp-up - a picture of what the channel could deliver before it goes live. No substitute for the real test, but a filter that sorts out obviously weak candidates.
How the new-channel test works
Two questions for the new channel
First: does the channel carry enough at all - measured by its saturation limit and its ramp-up? Second: is the extra dollar better invested here than in the existing mix? The model answers both.
- Add the channel. Create it in a few clicks from one of the templates or as your own channel.
- Start with benchmarks. If you have no numbers of your own, take the industry ranges you can show as a starting point - the status per value shows whether it sits in a realistic range.
- Check feasibility. Look at the saturation limit (α) and ramp-up: a channel with a low limit saturates early, a channel with a long ramp-up pays in only late - both flow into the expected contribution.
- Compare against the existing mix. Calculate with and without the new channel and use "compare plans" to see whether it brings more than the same budget in the existing mix - which often already works closer to its saturation.
- Decide - and test small. If the channel carries in the model, confirm it with a limited real budget instead of going in big right away.
Work through a new channel upfront, before budget flows: try the tool
Talking points for the conversation
- "Calculate first, then test." Shifts the decision from gut feeling to an estimated order of magnitude.
- Make the ramp-up visible: a channel that only carries after months needs patience in the budget - that belongs cleared up before the start.
- Show the marginal benefit: a new channel can bring more than extra budget in a channel that's already working at its saturation.
- Don't forget the halo: some channels feed indirectly into others - that belongs in the comparison, or the new channel looks weaker than it is.
Common thinking traps
- Reading the simulated contribution as a promise. The model estimates forward from benchmarks - it doesn't guarantee the channel will perform that way. The real test stays necessary.
- Skipping over the ramp-up. Whoever treats a channel with a long ramp like a paid channel plans the break-even too early and disappoints themselves.
- Looking only at the new channel. The real question is the comparison: the new dollar competes with the existing mix, not with zero.
Frequently asked questions about the new-channel test
Can I test a channel that isn't in the templates?
Yes. Alongside the templates you can create your own channel with any name. For unknown types, a conservative default serves as a starting point that you sharpen with your own values.
How do I tell whether the new channel carries enough?
From the expected contribution under saturation and ramp-up. A low saturation limit caps the channel early, a long ramp-up delays its return - both flow into the projection and show whether entering is worth it.
How do I compare the new channel with the existing mix?
Through the plan comparison. Calculate once with and once without the channel and set the states against each other - the difference shows whether the new dollar brings more here than in the previous mix.
Does the simulation replace the real test?
No. It is an upfront filter that sorts out obviously weak candidates and estimates the order of magnitude. Whether the channel really carries is only shown by a limited real test - the simulation just makes it more targeted.
Do I need my own numbers for the test?
No. Without your own values, the benchmark ranges you can show provide a first direction; as soon as real or researched figures are available, the estimate becomes more robust.
