A supplement protocol without a goal is a shopping list. A protocol with a goal has a success criterion, a timeline, and a stop rule. The difference matters because it determines whether you can tell if the supplement worked.
Most people run supplement stacks with vague intentions ("feel better," "have more energy," "get healthier") and then evaluate them with vague impressions ("I think it's helping"). That process cannot produce a useful answer. A goal-based protocol forces specificity at every step: what you are measuring, when you will evaluate, and what result would make you keep, adjust, or stop.
The structure of a goal-based protocol
Every protocol needs five defined elements before you take the first dose:4
| Element | What you decide | Example (sleep goal) |
|---|---|---|
| Goal | One sentence describing the outcome you want to change | "Reduce sleep onset latency from ~40 minutes to under 20 minutes" |
| Primary metric | The measurement you will use to evaluate the goal | Sleep onset latency (self-reported, logged nightly) |
| Secondary metric | A supporting measurement that checks for trade-offs or side effects | Morning energy rating (1-10, logged at 8 AM) |
| Trial duration | How long you will run the protocol before making a decision | 14 nights (melatonin has a short time-to-effect) 1 |
| Success threshold | The specific result that would make this "worth it" | Average sleep onset under 20 minutes in the final 7 nights, with no persistent morning grogginess |
If you cannot fill in this table, your protocol is not ready to start. Go back and sharpen the goal.
Worked example: A cognitive focus protocol
Here is a complete protocol for someone whose goal is better sustained focus during afternoon work blocks.
Protocol definition
| Element | Decision |
|---|---|
| Goal | Improve sustained focus during 2-4 PM work block |
| Primary metric | Deep-work minutes completed (tracked by timer) |
| Secondary metric | Subjective focus rating (1-10, logged at 4 PM) |
| Trial duration | 14 days (caffeine + L-theanine effects are acute) 2 3 |
| Success threshold | Average deep-work minutes increase by 20%+ over baseline, with no sleep disruption |
Baseline phase (days 1-7)
No supplement changes. Log deep-work minutes and focus rating every weekday for one week. Also log sleep onset and morning energy to establish a baseline for potential side effects.
Result: Baseline average is 45 minutes of deep work and a focus rating of 5/10 during the 2-4 PM block.
Intervention phase (days 8-21)
- Caffeine 100mg + L-theanine 200mg, taken at 1:30 PM (30 minutes before the work block).
- No other supplement changes.
- Continue logging the same metrics daily.
Review (day 22)
Compare the last 7 days of the intervention to the 7-day baseline:
| Metric | Baseline avg | Intervention avg (last 7 days) | Change |
|---|---|---|---|
| Deep-work minutes | 45 | 62 | +38% |
| Focus rating (1-10) | 5.0 | 7.2 | +2.2 |
| Sleep onset (minutes) | 15 | 14 | No change |
Decision: Primary metric exceeded the 20% threshold. Sleep was unaffected. Keep the protocol. Lock it in for the next 4 weeks and re-evaluate at that point for tolerance effects.
What if the results were different?
| Scenario | Decision | Next step |
|---|---|---|
| Deep-work improved but sleep onset worsened (>25 min) | The protocol has a trade-off | Move the dose earlier (12:30 PM) and retest. If sleep still suffers, try L-theanine alone without caffeine. |
| No change in deep-work minutes | Null result | Remove the supplement. Return to baseline for 1 week. Consider whether the issue is actually attention or something else (sleep debt, task design, distraction environment). |
| Focus improved but you skipped the supplement 4 of 14 days | Adherence problem | The protocol may work but the delivery does not fit your schedule. Simplify (pre-pack doses, set a timer). |
Stack drift: the most common protocol failure
Stack drift is what happens when you start changing your stack mid-trial based on something you read, a recommendation from a friend, or a feeling that "maybe I should add X too."
Here is how drift happens in practice:
- Week 1: You start L-theanine + caffeine for focus. Solid plan.
- Week 2: You read about lion's mane for cognitive function. You add it.
- Week 3: Someone mentions alpha-GPC. You add that too.
- Week 4: Your focus is better. But you changed three variables. You have no idea which one (or which combination) produced the result.
The fix is simple but requires discipline: do not change your protocol during a trial. If you learn about something promising, write it down for the next experiment. Finish the current one first.
Choosing metrics that work in real life
The best primary metric is one you can measure daily without special equipment, that directly reflects your goal, and that varies enough to detect a change but not so much that noise drowns the signal.
| Goal | Good primary metric | Why it works | Poor primary metric | Why it fails |
|---|---|---|---|---|
| Sleep improvement | Sleep onset latency (self-reported) | Daily, sensitive to change, directly measures the goal | "How well did I sleep?" (unanchored) | Too vague, no scale definition, shifts with mood |
| Strength | Training volume (sets x reps x weight) | Objective, logged in training app, directly measures performance | Body weight | Confounded by water, food timing, and is slow to change |
| Stress reduction | Daily stress rating (1-10, anchored) | Simple, repeatable, captures subjective experience | "Do I feel less stressed?" (yes/no) | Binary answers lose granularity and trend information |
| Fat loss | Waist circumference (weekly) | Practical, tracks visceral fat loss, less noisy than daily weight | Daily scale weight | Fluctuates 1-3 lbs daily from water and food, creates false signals |
Anchoring your subjective scales
If your primary or secondary metric is a self-rating, define the scale before you start:
Focus (1-10): 1 = Cannot sustain attention for more than 2 minutes. Constantly distracted. 5 = Normal working state. Some drift but can redirect. 10 = Deep, uninterrupted focus for the full session. Effortless concentration.
Write this down. Refer to it when you log. Without anchoring, a "7" on day 3 and a "7" on day 12 might mean completely different things.
When to adjust dose versus add a new supplement
When results are underwhelming, the instinct is to add something. Usually the better move is to adjust what you already have.
Adjust first when:
- You are below the studied dose range (room to increase).
- The timing might be wrong (taking caffeine too close to sleep, or too far from the work block to catch peak effect).
- Adherence was below 85% (the protocol did not get a fair test).
Consider adding only when:
- The current supplement was taken at a studied dose, with good adherence, for the appropriate duration, and the primary metric did not meaningfully change.
- The new addition targets a different mechanism from what is already in the stack.
- You are willing to run another full trial period after adding it.
Goal-based protocols in Unfair
Unfair structures every supplement stack around a defined goal. When you create a stack, you name your primary metric and set a success threshold before the trial starts. The app tracks your metric alongside adherence data, and prompts a structured review at the end of your defined trial period. Each adjustment you make creates a new version of the stack, so your decision history is preserved and you can compare performance across protocol changes.
Continue with Muscle Gain Stack Basics, Fat Loss Stack Basics, and Achieving Goals with Stack Cycling.
References
Ferracioli-Oda E, Qawasmi A, Bloch MH. Meta-Analysis: Melatonin for the Treatment of Primary Sleep Disorders. PLoS One. 2013;8(5):e63773. https://pubmed.ncbi.nlm.nih.gov/23691095/
↩Guest NS, VanDusseldorp TA, Nelson MT, et al. International society of sports nutrition position stand: caffeine and exercise performance. J Int Soc Sports Nutr. 2021;18:1. https://pubmed.ncbi.nlm.nih.gov/33388079/
↩Giesbrecht T, Rycroft JA, Rowson MJ, De Bruin EA. The combination of L-theanine and caffeine improves cognitive performance and increases subjective alertness. Nutr Neurosci. 2010. https://pubmed.ncbi.nlm.nih.gov/21040626/
↩Vohra S, Shamseer L, Sampson M, et al. CONSORT extension for reporting N-of-1 trials (CENT) 2015 Statement. BMJ. 2015;350:h1738. https://www.bmj.com/content/350/bmj.h1738
↩