Goal-Led Carbohydrate Experiment
Adjust one dietary variable, measure the target outcome, then decide
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 91%
The Goal-Led Carbohydrate Experiment begins by asking what outcome the person wants, such as a change in weight, blood-glucose control, cognition, or another defined goal. Next, locate their current eating pattern on a carbohydrate spectrum and choose one realistic reduction, such as reducing bread rather than automatically adopting a ketogenic diet. Select a measure tied to the purpose—Unwin mentions weight, blood work, and glucose feedback—then observe what changes. Review whether the person is satisfied, whether the measure improved, and whether a further adjustment is desirable. The mechanism is iterative personalization: purpose determines the intervention, measurement produces feedback, and feedback guides the next choice. Unwin presents this as experimentation, not a guarantee, and the transcript does not establish that every claimed outcome will occur for every person.
Origin
Extracted from The Diary of a CEO
Core principles
- 01Begin with the outcome the person actually wants
- 02Dietary intensity should follow purpose rather than fashion
- 03Change can move along a spectrum instead of jumping to an extreme
- 04Measurement should match the stated goal
- 05Feedback determines whether to maintain or adjust
How to run it
- 1
Define the purpose
Name the result you want before selecting a diet. Make the goal specific enough to determine what evidence would count as progress.
Pro tip Separate goals such as glucose control, cognition, weight, and strength instead of assuming one measure represents all of them.
Watch out Do not select a restrictive diet only because it is popular.
- 2
Locate the baseline
Estimate the current carbohydrate pattern and identify a food or habit that could realistically change. The baseline provides the comparison for later feedback.
Pro tip Choose a change concrete enough to observe, such as reducing a recurring bread serving.
Watch out Medication can alter the safety of carbohydrate reduction; involve a qualified clinician when relevant.
- 3
Choose the measure
Match a relevant measure to the goal, such as clinician-ordered blood work, weight, or another agreed outcome. Record enough baseline information to avoid relying only on memory.
Pro tip Use more than a momentary reading when the outcome naturally changes over a longer period.
Watch out A consumer measurement can be incomplete or misleading without clinical context.
- 4
Run one adjustment
Make the selected change while keeping the purpose and measurement plan stable. Observe effects over a period appropriate to the chosen outcome.
Pro tip A modest first adjustment makes it easier to learn what changed and to sustain it.
Watch out Do not interpret coincidence or a single reading as proof of causation.
- 5
Review and decide
Compare the feedback with the goal and ask whether the current level is satisfactory. Maintain it, change course, or test a further reduction based on the result and how the person feels.
Pro tip Record the decision and the reason so the next cycle starts with a clear baseline.
Watch out Stop and seek medical advice if the experiment causes concerning symptoms or unsafe readings.
In the wild
When asked about ketogenic eating, Unwin first asks the host what he wants from it. He then describes carbohydrate as a spectrum: establish the current position, try reducing a food such as bread, measure the chosen parameter, and ask whether the result is satisfactory before going lower.
→ The dietary choice is tied to the host’s stated goals and reviewed through feedback rather than treated as a universal prescription.
Common mistakes
Starting with the diet label
Beginning with “keto” or another label skips the question of what outcome the person wants and how it will be assessed.
Changing everything at once
A large bundle of simultaneous changes makes feedback harder to interpret and may be more difficult to sustain.
Measuring without a decision rule
Data is not useful by itself. Decide in advance how the result will inform maintaining, reversing, or extending the change.
Is it for you?
Best for
It is best for people exploring a dietary change with a clear outcome and an appropriate way to monitor it.
Not ideal for
It is not ideal for unsupervised changes when diabetes medication, pregnancy, an eating disorder, or another medical condition creates risk.
From the transcript
“we need to begin with your goals and hope.”
“there is a spectrum of carbohydrate that you're on.”
“Let's measure whatever parameter we want, which might be blood work or weight or whatever.”
From the episode
Fatty Liver Expert: Your Liver Is Filling With Fat Right Now - Dr David Unwin