What 7,593 brain tests
tell us.
712 people tested in the CONKA app over 30 months. Here is what moves a score, and how sure we are. Tap a finding for the data.
Time of day
Your brain runs on a curve.
Your sharpest hours are 9am to 3pm. By 9pm, scores drop nearly a full point below your daily average.
Strong712 users · 7,593 tests · 30 months9am–3pm
Your sharpest hours
vs. your daily average
When to take each shot
06:00 – 12:00
Morning focus. Take with or after breakfast.
12:00 – 18:00
Afternoon reset. Take with or after lunch.
See the full dataHide the full data
Peak to trough: ~1.5 points
Roughly the difference between catching the small typo and missing it.
9pm: -0.82 vs your own average
The same brain that nailed it at 11am is now operating below its own baseline.
MORNING PEAK
+0.47
Most people are at their sharpest between 9am and noon — about half a point above their daily average.
n=1,650 tests · 09-12
EVENING DIP
-0.82
By 9pm, scores drop nearly a full point below the daily average — a real, measurable dip.
n=818 tests · 21-24
CONKA EFFECT
+1.09
On days CONKA was taken, scores held above the daily average even during the evening dip — 1.09 points above the non-CONKA curve.^^
n=74 CONKA tests · 18-21
Your cognitive performance naturally rises and falls throughout the day. Most people hit their sharpest point between 9am and 3pm, then slide into a noticeable dip by evening — lowest around 9pm.^^
What we see with CONKA
CONKA users hold their level when others drop.
When we look at tests logged on CONKA days, scores stay above the daily average exactly where the curve normally drops most: late afternoon and evening. Mid-morning — when scores naturally peak anyway — the gap nearly disappears. We can't control for every variable, but the pattern is consistent across all four dip windows.
247 CONKA-tagged tests across the dip windows · directional consistency across all four windows · users may differ in unmeasured ways
Methodology · APP-01
Per-user deviation from personal mean. Each test compared to that user's daily average across all hours. 7,593 tests across 712 users between November 2023 and May 2026. Timestamps treated as UK local time (timezone not stored).
Mental fatigue
When you feel foggy, the test scores agree.
When you feel foggy, the data agrees. Fatigued days cost about 1.8 points off your personal best.
Strong260 users · 1,248 fatigued tests · 18 months-1.8 pts
On days you feel foggy
points lost vs. your best days
See the full dataHide the full data
+24ms slower reaction
About 6% slower than your fresh-day baseline. Noticeable on anything that needs a quick decision.
-2.7 pts when 'not feeling best'
Your gut sense of readiness reliably tracks how you actually score.
AVERAGE COST
-1.8 pts
On average, fatigued days cost users 1.8 points off their personal best — a measurable dip you can feel before the test starts.
n=260 users · 1,248 fatigued tests
REACTION
+24ms
Reaction time slows by about 6% when fatigued — not huge, but enough to notice on anything that requires a quick response.
n=123 users
SELF-MATCH
-2.7 pts
On days users said they weren't at their best, scores were 2.7 points below their personal peak — the data agreed with them.
n=78 users · 395 tests
The data backs up what you already sense: feeling foggy actually costs you. Fatigued days show a real and consistent dip in both score and reaction time — and the worse the fatigue, the bigger the drop.^^
What we see with CONKA
Faster reaction times on CONKA days.
When fatigued-day tests are split by whether CONKA was logged, the CONKA days show 41ms faster reaction times on average.^^ At 15 users this is an early signal, not a controlled trial — but the direction is consistent.
n=15 users with both conditions · directional, not statistically conclusive · accuracy delta is mixed at this sample size
The ingredient research
App data on CONKA's effect for fatigue is directional. Independent peer-reviewed studies on the relevant ingredients show:
- Acetyl-L-Carnitine (ALCAR)PMID 18937015 ↗
In one study, participants taking acetyl-L-carnitine showed improvements in psychomotor speed and attention, and reductions in mental fatigue (Malaguarnera et al. 2008).¶
Randomised controlled clinical trial · 125 subjects · 90 days
- Rhodiola roseaPMID 19016404 ↗
In one study, participants with fatigue syndrome taking Rhodiola rosea showed an anti-fatigue effect and improvements in mental performance (Olsson et al. 2009).¶
Phase III randomised, double-blind, placebo-controlled · 60 individuals with fatigue syndrome · 28 days
Methodology · APP-01
Per-user delta method. Each user compared against their own fresh-day baseline. 260 users meeting the both-conditions threshold. Data from November 2024 to May 2026.
Stress
Stress costs more than most people realise.
Moderate stress costs more than a heavy night's drinking. About 5 points off your calm-day baseline.
ModerateEarly signal12 users · 44 stress-day tests · 891 stress entries-5.4 pts
On moderately stressed days
points lost vs. your calm days
See the full dataHide the full data
+41ms slower under stress
Roughly 10% slower reaction time, plus about one extra error per session.
53% of all sessions logged under mild stress
Mild stress isn't an edge case. It's the default state most users test in.
MODERATE
-5.4 pts
On moderate-stress days, scores drop over 5 points from each person's calm-day baseline — one of the largest effects across the entire dataset.
n=18 users · 58 tests
REACTION
+41ms
Reaction time slows by 10% under stress — enough to feel on anything that needs a quick decision.
n=12 users
BACKGROUND COST
53%
More than half of all test sessions are taken under mild stress. It's not an occasional event — it's the default state.
n=891 stress entries
Stress is the single largest performance signal in the app data. Mild stress appears in more than half of all sessions — it's the background state, not an edge case. Under moderate stress, scores fall over 5 points and reaction times slow by 10%. The effect under moderate stress is comparable to a heavy night's drinking.
The ingredient research
Only 3 users have stressed test days both with and without CONKA. Per-user app data is below the threshold for a CONKA-effect observation. Independent peer-reviewed studies on the stress-targeting ingredients in CONKA show:
- Ashwagandha (Withania somnifera)PMID 23439798 ↗
In one study, participants with chronic stress taking KSM-66 ashwagandha showed a reduction in perceived stress and serum cortisol levels (Chandrasekhar et al. 2012).¶
Prospective, randomised, double-blind, placebo-controlled · 64 subjects with chronic stress · 60 days
- Lemon Balm (Melissa officinalis)PMID 16444660 ↗
In one study, participants taking Lemon Balm showed reduced effects of laboratory-induced stress on anxiety ratings (Kennedy et al. 2006).¶
Randomised, double-blind, placebo-controlled, crossover · 24 healthy volunteers · Single-dose sessions
- Rhodiola roseaPMID 19016404 ↗
In one study, participants taking Rhodiola rosea showed a reduced cortisol response to awakening stress and an anti-fatigue effect (Olsson et al. 2009).¶
Phase III randomised, double-blind, placebo-controlled · 60 individuals · 28 days
Methodology · APP-01
Per-user delta. Stressed days (moderate stress or higher, normalised score ≤ 0.5) compared to each user's no-stress baseline. 12 users met the both-conditions threshold. Data from December 2025 to May 2026.
Alcohol
What a hangover actually does to your brain.
Under 6 drinks: no clear signal. Six or more: nearly 5 points off your sober baseline the next morning.
ModerateEarly signal27 users · 113 hangover tests · 638 entries-4.9 pts
The morning after 6+ drinks
points lost vs. your sober days
See the full dataHide the full data
+29ms slower the morning after
About 7% slower reaction time. The next-day fog is real and measurable.
-16% reported readiness on hangover days
Users feel less prepared, and the test scores agree.
6+ DRINKS
-4.9 pts
The morning after 6 or more drinks, scores drop nearly 5 points from each person's sober baseline — one of the largest single-cause drops in the dataset.
n=11 users · 24 tests
REACTION
+29ms
Hangover mornings slow reaction time by about 7% — enough to notice on anything that needs quick thinking.
n=27 users · 113 hangover tests
READINESS
-16%
People report feeling 16% less ready on hangover mornings. In this dataset, the test scores agree.
n=8 users
Light drinking (under 6 drinks) doesn't produce a consistent signal in this data. The effect appears clearly at 6 or more drinks: scores drop nearly 5 points, accuracy falls, and reaction time slows by 7%. If you've ever felt the next-day fog was real, this data suggests you were right.^^
What we see with CONKA
Faster on CONKA hangover days.
When hangover-day tests are split by whether CONKA was logged, the CONKA days show 56ms faster reaction times.^^ At 11 users this is directional — worth noting, not concluding from.
n=11 users with both conditions · directional, not statistically conclusive · sample may have followed lighter drinking nights
Methodology · APP-01
Per-user delta. Hangover days (1+ drinks the previous night) compared to each user's sober baseline. 27 users met the both-conditions threshold for the core analysis. Data from December 2025 to May 2026.
Coffee vs CONKA
Coffee gets the credit. The data gives it to CONKA.
Coffee on its own barely moved our users' cognitive scores. Every gain in the data tracked with CONKA, not caffeine.
Moderate501 tests with both logged · 19 months+4.0 pts
Adding CONKA to coffee
average total score, 0-100
See the full dataHide the full data
Coffee alone: ~0 score change
Measured against their own scores, a coffee looked almost identical to drinking nothing at all.
Adding CONKA to coffee: +4 points
Among users with both habits logged, nearly two in three improved once CONKA was added on top.
COFFEE ALONE
+0.2
Measured against each user's own baseline, a coffee moved total score by almost nothing, and reaction time actually slipped about 13ms slower.
n=104-222 users · within-person
CONKA EFFECT
+2.1
On the same within-person method, CONKA was associated with a +2.1 point lift in total score, with 60% of users improving.^^
n=47 users · within-person
CONKA ON COFFEE
+4.0
Among users with both a coffee-only and a coffee-plus-CONKA history, adding CONKA was associated with +4 points, 64% improving, and fewer errors.^^
n=22 users · within-person
Caffeine is the habit everyone credits for their focus. In our app data it barely showed up. Measured against each user's own baseline, coffee on its own left scores flat and was linked to slightly slower, less controlled responding. The lifts in score, control and error rate all tracked with CONKA, and when CONKA was added on top of coffee, the same people improved.^^
What we see with CONKA
The gains track with CONKA, not caffeine.
Across every way we sliced the data, each best result contained CONKA, and coffee on its own sat level with drinking nothing. The cleanest signal is the interaction: among 22 users with both a coffee-only and a coffee-plus-CONKA history, adding CONKA was associated with +4 points and nearly two in three improving. The faster raw reaction times in the CONKA groups are real in the averages, but within-person that speed gain concentrates in a minority, so we lead with score, not milliseconds.
Observational and self-reported · CONKA-only cell is small (n=12) · raw group means and within-person deltas disagree on reaction speed, and the within-person number governs
Methodology · APP-05
Two methods. Raw group means average each metric inside the four consumption groups, shown in the chart. Per-user deltas compare each user against their own intake history and govern where the two disagree. Any caffeine logged counts as coffee; any CONKA dose counts as CONKA. 501 tests with both factors logged, practice tests excluded. Data from November 2024 to June 2026.
We compare you
to you.
Every finding above comes from inside the CONKA app, on our own users, each measured against their own normal. Here is how.
- Step 1
CONKA FlowMorning 
CONKA ClearAfternoon Take CONKA
Flow in the morning, Clear in the afternoon, or both.
Step 2Test done
Test in the app
A 90-second cognitive test, FDA cleared and built on Cambridge research.
Step 3Against your baseline
See your change
Every test plots against your own baseline, so the data shows what moved you.
- Sensitivity detecting cognitive changeADePT Study, PMC10533908
- 93%
- Test-retest reliabilityADePT Study, PMC10533908
- 87.5%
- NHS Trusts in clinical validationHRA ISRCTN95636074
- 14
- FDA cleared as a medical deviceCognetivity Neurosciences
- 510(k)
How we compare you to you
The simple way (misleading)
Average every drinker against every non-drinker. Whoever naturally scores higher wins, regardless of the drinking.
What we do (per-user delta)
Build each user's own baseline from their clean-state tests. Measure their impaired-state tests against that baseline. Average the personal changes.
The question stops being “are drinkers worse than non-drinkers” and becomes “how much does this person change relative to their own normal.” That is the only honest way to compare an impaired state against a baseline using observational app data.
What this method can and can't do
- Observational, not a controlled clinical trial.
- Each user must have tests in both conditions to be included, which is why some sample sizes are smaller than the topline user count.
- Where the per-condition sample is thin, we mark it Early signal or skip the claim entirely.
- Wellness factors (alcohol, fatigue, stress, readiness) are self-reported in the app at the moment of testing.
About the test and its validation
The CONKA app uses a clinically validated cognitive assessment developed by Cognetivity Neurosciences from Cambridge University research. The test is FDA cleared as a medical device with 93% sensitivity for detecting cognitive change and 87.5% test-retest reliability, validated across NHS Memory Clinics (ADePT Study, PMC10533908; HRA ISRCTN95636074). Test scores reflect individual cognitive test performance and do not constitute health claims about CONKA products. Many factors, including lifestyle changes, practice effects, and natural variation, may contribute to changes in test scores.
Now see what moves your score.
Everything above is our users' data. The free CONKA app runs the same 90-second test on you, measured against your own baseline from day one.
Free on iOS and Android. No subscription needed.
15+
trials run with professional sports organisations, on the same test.
- Premiership Rugby
- Pro 14 Rugby
- GB Sevens (Men)
- GB Sevens (Women)
- Premier League
- EFL Championship
- Bundesliga
- NFL
- Boxing
- UCI Cycling
- University Sport
All club data is held privately under NDA.
Enquire about a trialHow we look at the numbers.
Every analysis on this page uses a per-user delta method. We compute each user's personal baseline from their own clean-state tests, then compare their impaired-state tests against that baseline. This removes the confound of natural ability differences between users.
Wellness factors (alcohol, fatigue, stress, readiness) are self-reported in the CONKA app on an opt-in basis at the moment of testing. Cognitive scores come from the same test session.
^^ Cognitive test details and validation are documented above in "How we know".
¶ Ingredient-level peer-reviewed studies. Findings as published; not extrapolated to product-level effect.
Food supplements are not a substitute for a varied and balanced diet and a healthy lifestyle.
Kurup, R. (2026). CONKA Real-World Evidence Report: cognitive performance patterns from 712 app users (APP-01 to APP-05). CONKA.
Reviewed