How we decide to launch a new flavour

A new flavour at Color of Vietnam does not start with someone liking an idea. It passes three gates. Gate one is a real signal from eaters and buyers. Gate two is one small trial with a question written in advance. Gate three is four go or stop questions, and a single no means we stop and write down why. This piece walks through each gate with an example whose outcome was not yet.

Gate one: a signal, not a preference

Flavour ideas are never in short supply. Signals are. We treat something as a signal when:

• The same question arrives from people who do not know each other: a buyer at a trade fair, an eater at a tasting session, a shop assistant at a retail chain.
• A comment repeats across several weekly tastings, in different words but with the same meaning.
• The brand promises an eating moment and none of its products serves that moment yet.
• A new ingredient sample makes the whole tasting table stop and pay attention.

What is not a signal: one loud voice, a competitor launching something similar, an article about a trend. Those go on the waiting list, not the trial list.

Gate two: one small trial with a written question

Once a signal is there, we do not write a recipe. We write one question, a single sentence, such as: if we make a cashew bar in a savoury direction, can it hold its shape without a sweet binder. The question goes to our partner facilities together with the flavour direction, and we ask for one small batch of two or three samples that differ in exactly one variable.

When the samples arrive, the team tastes blind. Everyone fills in a sheet before anyone speaks. A few bars are kept as retained samples, because many problems only show up after weeks in the wrapper. The decision goes into the sample tracking sheet with the usual three outcomes: approved, rejected, or adjust and retry.

We would rather run three small rounds than one big one, because every small round teaches something clear.

Gate three: four go or stop questions

A sample that passes the tasting table is only half the way. Before we move to packaging and product registration, we ask four questions and need four yeses:

• Does it sound like the brand? A striking flavour that blurs the brand's voice is not worth it.
• Can it be made consistently? On the current line, with ingredients we can buy all year, not from one lucky lot.
• Is there room for it? On the shelf, in the price list and in the buyer's mind, without cannibalising a product that already sells well.
• Can the team finish it? Packaging, registration file, sales samples and content for the brand, this quarter, with the people we have.

One no means stop or wait. The reason is recorded, so that when circumstances change everyone knows why we stopped and what it would take to reopen the file.

An example: a savoury cashew bar for Natural & Healthy

The signal came from two directions in the same season. At weekly tastings, the repeated comment was that every nut bar is sweet, and people wanted one for the moments when they are not in the mood for sweet. At two separate meetings, two buyers asked whether we had a savoury nut bar. Two sources, one question. That was enough for gate one.

The written question: roasted salted cashew with seaweed, pressed into a bar, can it hold together without a sweet binding layer. Our partner facilities made one small batch with two samples: one with the binder cut sharply, one with the binder at the lowest level the press could still handle. The first sample fell apart as soon as the wrapper was opened. The second held its shape but tasted half savoury and half sweet, and the second-piece question drew a lot of shaking heads.

At gate three, the brand question was a yes: Natural & Healthy lets the cashew speak, and a savoury direction fits. The consistency question was a not yet, because there was no binder for a savoury bar that the current line could run. The sheet now reads: not proceeding, on the waiting list, reopen when a suitable non-sweet binder is found. The whole loop took a few weeks and one small batch. The signal is kept. Nobody had to throw it away.

Why three gates beat one good idea

This is slower than making whatever sounds good. But it keeps our four brands from swelling with flavours only we like, and it keeps our partner facilities from running batches that lead nowhere. Our value Small Trials lives at gate two. Speak Honestly lives at gate three, when a sample that tastes good still has to stop because it cannot yet be made consistently.

If you want to be the person who hears the signals and writes the trial questions, our Marketing Leader and Fresher Product R&D (FMCG) roles in Hanoi are open on the Careers page. If you are a buyer looking for a flavour we do not have yet, tell us through the Contact page. That is exactly the kind of signal we are waiting for.