Hermetica Superfood Co.
We surveyed 261 Hush customers between July 2024 and June 2026 and asked one question: what was your sleep score before, and what is it now? The average answer: 43 before, 80 after — a 37-point gain, with 92% reporting improvement. The full anonymized dataset is public, and this article walks through exactly how we collected it, what it shows, and where its limits are.
Why we published our raw data
Every supplement brand says their product works. Almost none show their homework. When we started collecting post-purchase survey responses for Hush Sleep Gummies in mid-2024, the plan was simple: ask customers for a number, not a feeling. A sleep score — whether it comes from an Apple Watch, a Garmin, an Oura ring, or an honest self-rating — is something you can put on a chart and argue with.
Two years later we had 307 responses, and the results were strong enough that hiding them behind marketing copy felt like a waste. So we cleaned the data, stripped every piece of personal information, and published the raw file. Anyone — a customer, a competitor, a journalist, a skeptical Redditor — can download it and check our math.
How the survey worked
The survey ran continuously from July 2024 through June 2026 as a post-purchase email to Hush customers. It asked for a sleep score before starting Hush, a current score, and an optional comment. Respondents used whatever measurement they had: 41% cited a wearable by name — Apple Watch, Oura, Whoop, Fitbit, Garmin — and the rest gave a self-rating.
Wearables report sleep scores on a 0–100 scale, but self-ratings often came in as "4 out of 10." We normalized all 1–10 self-ratings to the 0–100 scale (a 4/10 becomes 40) so every response is comparable. Each row in the public dataset is tagged with its source — `wearable_0_100` or `self_rated_1_10_x10` — so you can analyze the two groups separately if you prefer.
Of the 307 total responses, 261 provided both a before and an after score — those pairs are what every statistic in this article is built on. The other 46 gave only a current score and are excluded from the averages but included in the raw file. We also excluded 18 raw submissions that were unusable (blank scores, duplicates, or test entries) before the file was finalized.
The headline result
The average respondent started at 43 out of 100 — objectively poor sleep — and reported 80 after Hush. The distribution tells the story better than the average does:
Before Hush, the largest group of respondents sat in the 40–49 band, and nobody scored above 69. After, the chart flips: 233 of 261 respondents reported a score of 70 or higher. This is not a subtle shift in the average driven by a few outliers — the entire distribution moved.
Three numbers worth pulling out:
- 239 of 261 (92%) improved. The median gain was 40 points.
- 8 reported no change.
- 14 (5.4%) reported a lower score. Real products have non-responders, and we left every one of them in the data.
Stable across 24 months
A common pattern in supplement survey data is a launch spike: early adopters are enthusiasts, so early numbers look inflated and then decay. That is not what this dataset shows:
The monthly average after-score stayed between 70 and 89 for the entire two-year window, never approaching the pre-Hush baseline of 43. Month-to-month wobble is real (small monthly samples), but there is no downward drift.
What respondents actually said
The open-comment field was optional, and the most convincing comments are the ones that name a device, because a wearable score is not a mood:
Two more, verbatim from the dataset: "great sleep. garmin confirms." (31 → 91) and "my fitbit scores are finally good." Comments like these are why we report the wearable-cited percentage — 41% of respondents anchored their numbers to a device they don't control.
Where the data is weak — read this before citing it
Honest data journalism names its own limits, so here are ours:
- Self-reported. Even wearable scores are transcribed by the customer, not pulled from an API. People can misremember or round up.
- No control group. This is a customer survey, not a placebo-controlled trial. Some of this improvement is regression to the mean — people buy sleep products when their sleep is at its worst.
- Response bias. Customers who had a good experience are more likely to answer a survey than those who quietly churned.
- One product, one brand. We only have data on Hush, a 0.35 mg melatonin microdose formula with organic reishi, passionflower, L-theanine, and GABA. These numbers say nothing about 10 mg megadose gummies — if anything, the clinical literature suggests lower melatonin doses often outperform higher ones for sleep quality without the morning grogginess.
What we can say: 261 real customers, tracked over two years, reported an average 37-point sleep-score improvement, the distribution shifted wholesale, and the effect did not decay. That is a stronger evidence base than almost any sleep supplement on the market publishes — because almost none of them publish anything.
Download the dataset
The full anonymized file is public and free: hush-sleep-survey-2024-2026-anonymized-v2.csv. Each row is one response: month, score source, before, after, delta, and whether a wearable was cited. No names, no emails, no order data.
| What | Value |
|---|---|
| Collection window | July 2024 – June 2026 (24 months) |
| Total responses | 307 (261 paired before/after) |
| Wearable-cited | 41% of paired responses |
| Average before / after | 43 / 80 |
| Median gain | +40 points |
| Improved / unchanged / declined | 239 / 8 / 14 |
If you cite these numbers, link to the full Sleep Score Report or this article as the source, and note that scores are self-reported. We update the dataset as new responses come in.
Common Questions
Is this survey data real?
Why did sleep scores improve so much?
What is a sleep score?
Did anyone's sleep get worse?
Where can I download the raw data?
Pairs Well With This Article
Six functional mushrooms with fully-disclosed dosing to support focus, stamina, and everyday balance.
Purified Himalayan shilajit (500 mg) with fulvic minerals and nine disclosed botanicals for natural energy and endurance. Third-party lab tested with a published COA.
Marine collagen with antioxidants to support glowing skin, hair, nails, and joint comfort.
Relevant Articles
Continue Reading
Lion's Mane Mushroom and Cognitive Benefits: What Six Clinical Trials Actually Show
Hermetica Superfoods Reviews: Is It Legit?
L-Theanine vs GABA: Which Should You Take? | Hermetica
Why People Choose Hermetica