Blind Box™ · the 28-day self-trial subscription
Does it work for you? Find out blind.
Each month we post you a supplement and an identical placebo, randomised into a coded calendar blister. You log a few numbers a day. On day 29 you unblind and get your own statistics, not someone else's study.
- From £20.75 a month
- Pause or cancel any time
- Free UK tracked delivery
28 capsules. 14 active, 14 placebo, in an order nobody in the room knows. One sample member's real-format log sits behind it.
How a Blind Box works
Four weeks from question to answer
Pick a question
Choose a trial: focus, sleep or energy. Set the smallest change that would actually matter to you.
Take it blind
One pocket a day. Active and placebo capsules look, weigh and feel the same. The order is randomised in pairs.
Log in a minute
Rate a couple of measures in The Lab™. Some trials include a built-in reaction-time test.
Unblind on day 29
See which code was active, the size of the difference, an exact p-value and how well your blind held.
Sample result
One member, 28 days, one honest chart
Daily reaction time from the sample kit above. Before unblinding every dot is the same colour. Unblind the kit to see which days were active. Notice the slow downward drift: people get faster at any test with practice, which is exactly why the order is randomised in pairs.
Inside every Blind Box
A small clinical trial that fits through your letterbox
Kits are letterbox-sized and posted Royal Mail Tracked 48. Capsules are vegan HPMC, opaque, and identical between active and placebo.
Browse this month's trials- 28 ×Coded daily doses in a calendar blister, week rows and day columns, no labels that give the game away.
- 1 ×Sealed allocation. The randomisation lives in your Lab account and stays locked until you finish.
- 1 ×Protocol card with dose, timing, what to log and who should not take part.
- ∞The Lab™ app for logging, charts, exact statistics and your results archive.
- 0 ×Health claims we can't stand behind. We measure. You decide.
This month's trials
Start with one of these
Trials we refuse to run
Some supplements can't be tested blind. We say so.
The Control Group
One email a month: a study, a chart and what it can't tell you.
Trial library · October 2026
Five questions you can actually answer in 28 days.
We only offer compounds that act fast enough to test in short blocks, at doses used in published research, with a placebo that can't be told apart by taste or look.
Not offered, on purpose
Trials we refuse to run
If a blind can't hold or 28 days can't show an effect, a trial would just produce confident noise.
The method
A proper trial, with a sample size of you.
Each Blind Box is a randomised, placebo-controlled, double-blind N-of-1 trial. Here is exactly what that means and where its limits are.
Randomise
Days are grouped into blocks and blocks into pairs. A coin flip per pair decides which block gets active.
Blind
Identical capsules. Even our packing team sees only a kit number.
Measure
Pre-set measures, logged daily, plus a guess about each capsule.
Test
An exact permutation test over every allocation the coin flips could have made.
Interpret
Compare the difference against the change you said would matter.
Why blind?
Expectation changes how people feel and how they rate themselves. If you know you took the "focus" capsule, a 7 easily becomes an 8. A placebo that looks identical removes that path, so any difference left is more likely to come from the compound.
Why randomise in pairs?
People improve at any test with practice, sleep differs on weekends, and work stress comes in waves. If active days all fell in week 1, a learning effect would look like a supplement effect. Pairing blocks and flipping a coin within each pair spreads active and placebo evenly across the month.
Why block length differs
Some compounds wash out in hours, others take days. Block length is set per trial so that one block's effect has mostly faded before the next. Longer blocks protect against carryover but give fewer pairs, and fewer pairs means coarser statistics.
How we test the result
We calculate the difference between active and placebo days. Then we recompute it for every other way the coin flips could have landed. The p-value is the share of those allocations that give a difference at least as large as yours. No normality assumptions, no fudging.
That also sets a floor: with only a few pairs, even a perfect result can't produce a small p-value. We show that floor before you start.
| Trial | Block | Pairs | Smallest p |
|---|
Blind integrity
Each day you guess whether you took active, placebo or don't know. If your guesses are right far more often than chance, the blind probably broke and the result deserves less trust. We report this alongside every result.
What we never claim
- That a supplement treats, prevents or cures any disease.
- That your result applies to anyone else.
- That "no reliable difference" proves something is useless. It means it didn't detectably shift your measure, at this dose, over 28 days.
- Any health claim beyond the Great Britain authorised list. Where one exists, we quote it word for word.
Standards we follow
Results are reported in line with the CONSORT extension for N-of-1 trials (CENT 2015). Claims are checked against the GB nutrition and health claims register.
Pricing · all prices include VAT at 20%
One question a month. Your answers, kept.
Compare plans
| Feature | Pilot | Lab Member | Principal Investigator |
|---|---|---|---|
| Blind Box kits | 1 | 1 every month | 12 a year + 2 repeat kits |
| The Lab™ logging, charts, exact statistics | Yes | Yes | Yes |
| Pooled results across repeat kits | — | Yes | Yes |
| Choose or queue your next trial | — | Yes | Yes |
| New trials before general release | — | — | Yes |
| Skip, pause or cancel | No subscription | Any time before dispatch | Pause any time; cancel within 14 days for a full refund |
| Delivery (Royal Mail Tracked 48) | Free | Free | Free |
Checkout
Set up your Blind Box
- Plan
- Safety
- Trial
- Delivery
- Review
The Lab™
Your lab notebook
Brand book · v1.0 · October 2026
The capsule with one hatched half.
Our mark is a capsule split in two: one half solid, one half hatched. One is real, one is placebo, and you don't get to know which until the data's in. Everything else in the identity follows from that idea.
Clear space
Keep clear space equal to half the capsule's height on every side. Nothing enters it: no text, edges or other logos.
Minimum size
Mark: 16 px or 5 mm tall. Horizontal lockup: 96 px or 28 mm wide. Below that, use the mark alone.
Construction
The wordmark sets "Nutra" condensed and heavy, "Geeks" expanded and light: two weights of one variable face, echoing the two halves of the capsule.
Colour
Two codes and the lab around them
Cobalt and Rose are reserved for Code A and Code B. They never mean "good" and "bad", and never imply which code is active.
Typography
One variable face, stretched with intent
Display · Anybody
Aa 0.05
Aa
Condensed for data and headlines. Expanded light for the one phrase that matters. Never more than one expanded phrase per heading.
Body · Atkinson Hyperlegible
Aa Il1 O0
Designed by the Braille Institute so similar characters stay distinct. We use it because results must be read correctly, not just quickly.
Codes · IBM Plex Mono
NG-0742
Kit numbers, lot codes, blister positions and every p-value. Uppercase with 6% tracking for labels.
Packaging
The Blind Box™ outer
Letterbox format, 230 × 160 × 22 mm. Front carries the kit number and trial name only. Mandatory food supplement statements sit on the back panel. Code colours never appear on a sealed kit.
- Recyclable board outer, single colour plus Cobalt offset shadow
- Blister rows are weeks, columns are days, matching The Lab™
- Kit number in Plex Mono on every component, so parts can be matched back
BLIND
BOX™
28-day self-trial · Caffeine + L-theanine vs placebo
Food supplement. 28 capsules, coded. Take one capsule daily with breakfast, in blister order. Do not exceed the recommended daily intake. Food supplements should not be used as a substitute for a varied and balanced diet and a healthy lifestyle. Store in a cool, dry place out of the reach of young children. Contains caffeine: not recommended for children or pregnant or breastfeeding women.
Voice
We sound like a careful scientist who likes you
Say
"Your data showed no reliable difference."
Not
"Sorry, it didn't work for you!"
Say
"Magnesium contributes to a reduction of tiredness and fatigue."
Not
"Banish fatigue with magnesium."
Say
"We measure. You decide."
Not
"Clinically proven to boost focus."
Say
"This trial can't go below p = 0.5 from one kit."
Not
Hiding a limitation in the small print.
Trade mark usage
NutraGeeks™
- Write it as one word with a capital N and a capital G: NutraGeeks. Never "Nutra Geeks", "Nutrageeks" or "NUTRAGEEKS".
- Use the ™ symbol at the first or most prominent use on each page, pack or screen.
- Do not use ®. NutraGeeks is an unregistered trade mark, and ® may only be used once a registration is granted.
- Use it as an adjective with a noun: "NutraGeeks Blind Box", not "a NutraGeeks".
- Never alter the mark, recolour the hatched half, or animate the halves apart except in the official unblinding motion.
Sub-brands
| Blind Box™ | The monthly kit |
| The Lab™ | Logging and results app |
| Unblinding Day | Day 29 of every kit; descriptive, no mark |
| The Control Group | Monthly newsletter; descriptive, no mark |
Planned UK IPO filing classes
| 5 | Food supplements; placebos |
| 9 | Downloadable apps |
| 35 | Online retail of supplements |
| 42 | Software as a service |
| 44 | Nutritional information services |
Questions
Frequently asked, carefully answered.
What is an N-of-1 trial?
A clinical trial with one participant who acts as their own control. You alternate between the real thing and a placebo in a random order and compare how you did. It answers "does this help me?" rather than "does this help people on average?"
Why would I want a placebo?
Because expectation shifts self-ratings. Without a placebo you can't separate the compound from your hope that it works. A blinded comparison is the only fair test.
What is the placebo made of?
Microcrystalline cellulose, a plant-fibre filler used in many capsules and tablets, in the same vegan HPMC capsule as the active dose.
Can I tell which capsule is which?
They look, weigh and feel identical. Some compounds, like caffeine, can still be felt. That's why you guess each day: your blind integrity score shows whether the blind held.
What if I miss a day?
Keep taking the capsules in order and simply skip logging that day. The analysis uses the days you logged. We recommend at least 21 logged days out of 28.
Can I unblind early?
Results unlock after day 28. If you have a side effect or need to know for medical reasons, stop taking the capsules and use "Safety unblind" in The Lab™. It reveals your codes immediately and marks that kit's result as compromised.
Does "no reliable difference" mean a supplement doesn't work?
No. It means it didn't detectably change your chosen measure, at this dose, over 28 days. Effects smaller than your measure's day-to-day noise can still exist. Repeating the kit and pooling results narrows that down.
Is this medical advice?
No. Blind Box is a self-experiment tool using food supplements. It doesn't diagnose, treat or prevent any condition. If you have a health concern, speak to your GP or pharmacist.
How do skipping, pausing and cancelling work?
In The Lab™ under Account. You can skip the next kit, pause for one to three months or cancel at any time before the next dispatch. We email a reminder three days before every charge. You have 14 days from receiving your first kit to cancel. Opened blister packs can't be returned for hygiene reasons.
Where do you deliver?
UK mainland and Northern Ireland, by Royal Mail Tracked 48 (free) or Tracked 24 (£2.95 per kit).
Where is my data stored?
In this preview build, everything you log stays on this device in your browser. Nothing is sent to us. Clearing your browser data erases it, so copy your CSV from The Lab™ if you want a backup.
Still curious?
Write to us and a human with a science degree will reply within two working days.
hello@nutrageeks.ukLegal
Terms, privacy and safety
Terms of sale
These terms cover orders placed at nutrageeks.uk. NutraGeeks is a trading name; full registered company details will be published here before launch.
- Pilot is a one-off purchase of one kit.
- Lab Member renews monthly and Principal Investigator renews yearly until you cancel. We email a reminder three days before each charge.
- You can cancel within 14 days of receiving your first kit for a refund of any unopened kits. Opened blister packs can't be returned for health and hygiene reasons.
- You can skip, pause or cancel a monthly plan at any time before the next dispatch in The Lab™.
- We deliver to UK addresses only.
Safety
Blind Box kits contain food supplements and placebo capsules. They are for adults aged 18 and over. Do not exceed the recommended daily intake. Food supplements should not be used as a substitute for a varied and balanced diet and a healthy lifestyle. Keep out of the reach of young children. Stop taking capsules and seek advice if you experience an adverse reaction.
Privacy
In this preview build your account, trial logs and results are stored only in your browser's local storage. We don't receive them, and the site sets no tracking or advertising cookies.
When payments and accounts go live, we will process your name, address and email to fulfil orders, as required under UK GDPR. Trial data will be stored separately and never sold. You'll be able to export or delete it at any time.
Trade marks
NutraGeeks™, Blind Box™ and The Lab™ are unregistered trade marks of NutraGeeks. All other trade marks belong to their owners.
Contact
hello@nutrageeks.uk