The Portfolio of Probes
Share core capabilities across small bets, then concentrate on proven demand
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 97%
Nicola Kilner explains DECIEM's decision to pursue multiple brands as a portfolio of market probes supported by one internal ecosystem. Manufacturing, communications, research, and buyer relationships could serve several concepts, spreading capability costs and allowing the team to present more than one brand in a meeting. Because consumer demand remained uncertain until products started selling, the company treated relatively cheap failure and another attempt as part of the design. The Ordinary, its eleventh concept, eventually supplied the strongest signal and absorbed most operational attention. That success also exposed the model's limit: weaker experiments were neglected when the winner consumed the shared team. Nicola's corrective is structural separation—a dedicated incubator team—rather than expecting an overloaded core operation to protect exploration by willpower alone.
Origin
DECIEM began around the idea of doing ten things at once. Nicola Kilner says its shared in-house ecosystem made repeated consumer-brand tests possible, while The Ordinary's later dominance revealed the need for a dedicated team to preserve incubation.
Core principles
- 01Market traction reveals what research cannot settle in advance
- 02Shared capabilities can make several experiments cheaper than isolated launches
- 03Small failures preserve the ability to run the next test
- 04A proven winner deserves disproportionate attention
- 05Incubation needs protected ownership once the winner dominates operations
How to run it
- 1
Bound the probes
Define several distinct customer propositions that can be tested without risking the whole company. Make each probe small enough to stop when evidence is weak.
Pro tip Vary the proposition, not every part of the operating system.
Watch out Parallel brands without clear boundaries create activity rather than useful experiments.
- 2
Build the shared core
Identify capabilities such as manufacturing, communications, research, or distribution that multiple probes can use. Bring them in-house only where shared use justifies the cost.
Pro tip Calculate the capability across the portfolio rather than charging its full value to one probe.
Watch out The transcript presents DECIEM's ecosystem as a privilege that required funding; it is not automatically cheaper for every startup.
- 3
Test with buyers
Launch each proposition far enough to observe real purchasing and retailer response. Treat forecasts and research as inputs, not substitutes for traction.
Pro tip Use the same evidence standard across probes so internal favourites do not receive special treatment.
- 4
Fail cheaply
Close or deprioritize weak probes before they consume the capacity needed for another attempt. Carry reusable learning and capabilities into the next launch.
Pro tip Decide the acceptable loss before enthusiasm and sunk costs grow.
Watch out Cheap failure still needs clear responsibilities to customers, staff, and partners.
- 5
Concentrate on traction
Direct disproportionate capacity toward the proposition with the strongest demonstrated demand. Accept that equal treatment is no longer rational once one signal is materially stronger.
Pro tip Use sales and repeat demand alongside operational feasibility.
Watch out A winner can quietly starve every exploratory project.
- 6
Separate exploitation from exploration
If new probes remain strategically important, give incubation a dedicated team and mandate. Do not ask the team operating the winner to protect experiments during peak workload.
Watch out Dedicated incubation adds cost and should follow a deliberate decision, not portfolio habit.
In the wild
DECIEM launched The Ordinary after ten earlier brand concepts. Nicola says the shared ecosystem let the company keep trying, while real demand made The Ordinary the clear winner. The brand then became so large that the other concepts were pushed aside, prompting plans for a dedicated incubator team.
→ Repeated bounded attempts found exceptional demand, then concentration and organizational separation became necessary.
Nicola contrasts flying to meet a buyer with one brand against presenting ten brands supported by the same internal communications and operating capabilities. The portfolio increased the usefulness of that trip and spread specialist costs across several concepts.
→ A common capability served multiple market probes instead of being duplicated for each one.
Common mistakes
Running unrelated businesses at once
The mechanism depends on genuinely reusable capabilities; a collection of unrelated ventures does not gain the same leverage.
Predicting demand instead of testing it
Research can inform a launch, but Nicola's model treats actual traction as the decisive signal.
Making the winner fund invisible hobbies
Once one brand dominates, exploration needs explicit ownership and limits rather than unexamined access to the winner's team.
Is it for you?
Best for
It is best for teams that can reuse manufacturing, communications, research, or distribution across several bounded product experiments.
Not ideal for
It is not ideal when each experiment requires a separate expensive infrastructure or when the team cannot contain the cost and complexity of parallel work.
From the transcript
“until something gets traction you don't really know what's going to work”
“we could fail relatively cheaply quickly and kind of start the next thing”
“the only way we can do that is by having a dedicated team”
From the episode
A Billion dollar NIGHTMARE! The Disturbing Drama Behind The Scenes Of A Billion Dollar Beauty Business. The Ordinary - Nicola Kilner