Please view on a larger screen
The investor portal contains detailed research and visuals that require a desktop or laptop to view properly.
Return to websiteInvestor Portal
The Initial Problem
Opportunity
Over 600 million people worldwide work in shifts, and the sleep technology market continues to grow rapidly. Somna targets the intersection of shift worker wellness and wearable technology, a segment largely overlooked by existing sleep trackers.
Assumption Mapping
Mapping assumptions for our initial concept to de-risk and prioritize our validation efforts. The initial concept is described above: looking at bringing Fatigue Risk Management Systems to medical workers in a personalised approach through Fatigue Coaching.
Highest risk assumptions
Grouping key assumptions from the top right of the matrix.
Doctors would pay for passive features
- Users will pay for these features
- People would pay to solve this problem
- People will implement suggested habit changes
- Users during heavy fatigue have capacity to change habits
- There is enough schedule flexibility for convenient changes
- Software-based features deliver real value
The benefit we bring to users is quantifiable
- Our suggestions significantly impact fatigue short and long term
- There is a long-term benefit of waking in light sleep
- We can collect compelling, high-quality data
Data from existing wearables is not sufficient
- A distinct wearable is necessary
- Our USP is unique enough to not be absorbed by competitors
- The UX is distinct enough vs competitors
- Current solutions give users data overload and are intrusive
Risk 1: Medical Workers would pay for passive benefits
Research question
Whether medical workers see value in passive sleep tracking through existing wearables, and would pay for features requiring zero manual input.
Hypothesis
Medical workers are over-stimulated and will pay for features that require zero manual input to reduce decision fatigue.
Success criteria
At least 60% of interviewed users express willingness to pay for a passive-only sleep tracking product.
Methodology
We conducted a series of user interviews with medical workers across different shift patterns and specialities. The interviews were structured around three core questions: whether they currently use existing wearables (despite not being able to wear them during shifts), whether those wearables fully serve their needs compared to colleagues who don't use them, and whether non-users would want one if given the option.
The goal was to understand whether passive tracking — requiring no manual input from an already over-stimulated user — held enough perceived value to drive a purchase decision.
Findings
- 30%
of interviewed medical workers currently own a wearable device, primarily motivated by the data it provides — step counts, heart rate, and sleep summaries.
- 100%
of interviewees said they would be interested in understanding how any claims a device makes, or interventions it suggests, are derived from their own data. Transparency into the reasoning behind recommendations was a universal expectation.
- 60%
said they would only somewhat trust the results of a wearable-based fatigue tool — reflecting scepticism toward coaching-style recommendations that don't account for the realities of their working conditions.
The hypothesis was invalidated. Medical workers would not pay for a coaching-based fatigue tool — it represented a weak value proposition. Interviewees were clear that any sort of suggested intervention (e.g. "take a nap", "avoid caffeine after 2pm") would be more annoying than helpful, because they simply need to get on with their jobs. They don't have the luxury of adjusting their behaviour between shifts in the way a coaching model assumes.
The success criteria was not met: the majority of users expressed no willingness to pay for passive tracking or coaching-based fatigue management.
How this informed next steps
This result redirected us toward more tangible features and interventions. Even if fatigue coaching could theoretically help medical workers, there's no point building it if they won't seek it out and wouldn't listen to it. We needed to build a system they would actively respect and approach with curiosity — not one that lectures them about habits they can't change.
This test led us to explore medical workers' user journeys and pain points in much greater depth — specifically following the "mum rule" of interviewing without leading context — to identify moments where there are genuine opportunities to intervene. Rather than assuming when and how to help, we needed to observe where the real friction points occur and design interventions around those moments.
Pivot
This result caused a major pivot away from coaching-based fatigue management. The original value proposition — passive sleep tracking with coaching recommendations — was invalidated. The company redirected toward identifying tangible intervention points through deep user journey mapping, fundamentally reshaping the product from a fatigue coach into a sleep transition tool built around moments where medical workers actually have agency.
Research question
Is it feasible to intervene during shifts, and if so, where in the shift worker's routine is there a meaningful opportunity?
Hypothesis
There is a specific, recurring point in the shift cycle where users have enough agency for an intervention to be effective.
Success criteria
Users consistently identify a common window where they actively attempt to manage their sleep, and describe strategies that could be improved by an external tool.
Methodology
We conducted a series of interviews with people of various interests and backgrounds — not just medical workers, but anyone who had experience with rotating or irregular schedules. We deliberately broadened the participant pool to avoid anchoring on a single profession's coping strategies.
The interviews focused on two areas: how participants learned to manage rotas when they first entered the workforce, and what their key learning points were — specifically, how they arrived at their current sleep management strategies through trial and error.
Findings
We considered a broad range of potential interventions, including: tracking water consumption, tracking and recommending coffee intake, recommending power naps, alerting when it's not safe to drive home, estimating alertness and suggesting breathing exercises to improve concentration mid-shift, wind-down periods before sleeping, purchase of ambient products (eye masks, sun lamps), and community features allowing users to share fatigue tips.
Intervening during shifts proved infeasible — users have no flexibility while on duty. However, the transition between shift types emerged as a clear intervention point. It differs between each person how they approach the end of a series of night shifts — some force themselves to stay awake all day to reset their body clock, others allow themselves to sleep but only for a few hours. Everyone had developed their own improvised strategy, but nobody was confident it was the right one.
This conversation led users to reflect on how, for their worst schedule rotations, they often have to take more time off between long shift days so they can actually recover — leading us to focus on how we can help them transition their sleep more effectively.
Feature prioritisation survey
We surveyed medical workers on which potential features they considered essential, somewhat useful, or unwanted. We received 61 responses, of which only 41% had irregular shift patterns — meaning the majority of respondents were evaluating these features without direct experience of the core problem.
Software syncs with calendars to automatically track shift schedules and sleep windows
Bioimpedance sensors automatically track hydration and caffeine intake (key fatigue factors)
Silent alarms to avoid disturbing a partner
Short-term safety insights inform users if they are alert enough to drive home safely
Post-shift plotting of stress and fatigue for review
A smart alarm wakes the user during their lightest sleep phase to reduce "sleep inertia"
A custom routine planner assists users in transitioning between night and day shifts
The device detects high stress and initiates gentle vibration-guided breathing exercises
Coaching advice on wind-down periods
Community features allow users to share fatigue management tips
How this informed next steps
Since in-shift intervention was not viable, this pivoted our focus toward features that create interventions outside of work — specifically the transition window between shift rotations. The variance in recovery strategies confirmed a strong opportunity for guided, personalised sleep transitions. If every user is improvising differently, there is clear value in a system that can optimise this process.
Pivot
In-shift intervention was ruled out. This pivoted the product toward features that create interventions outside of work — targeting the transition window between shift rotations rather than the shifts themselves.
Research question
Whether the core problem facing shift workers goes beyond what passive data and insights can address — specifically around transitions between shift types.
Hypothesis
Shift workers face a recurring, high-stakes transition problem that passive data and insights alone cannot address.
Success criteria
Users independently identify shift transitions as a top-3 pain point, and describe consequences that passive tracking cannot mitigate.
Methodology
We conducted 16 in-depth user interviews with medical workers across different specialities and shift patterns. Critically, we gave participants no context about what we were building — no mention of sleep, wearables, or fatigue management. Following the "mum test" principle, we took a step back entirely and asked open-ended questions about their daily lives, what frustrates them most about shift work, and what they spend the most effort trying to solve at home.
The goal was to surface genuine, unprompted pain points and identify recurring patterns — not to validate our existing assumptions. We wanted to see what problems shift workers would describe if nobody told them what to complain about.
Hidden pain points (N=16)
The following pain points emerged organically from the interviews, without prompting. We tracked how many of the 16 participants independently raised each theme.
| The "Hidden" Pain Point | (N=16) | Prevalence |
|---|---|---|
| The "Wired" State (Inability to wind down after high-stress shifts) | 16 | 100% |
| Hydration/Nutrition Neglect | 15 | 94% |
| The "Drunk" Commute (Unsafe driving/cycling post-shift) | 15 | 94% |
| Rota Resentment (Switching day/night too fast) | 15 | 94% |
| The Blue Light/Circadian Gap | 14 | 88% |
| Systemic Gaslighting (Lack of data to prove fatigue to management) | 14 | 88% |
| Silent Alarm Necessity | 13 | 81% |
| The "Specialist" Device Preference | 13 | 81% |
| DIY Recovery Hacks (Use of sleeping tablets or 24hr "reset" stays) | 13 | 81% |
| Workplace Napping Taboo | 12 | 75% |
| Anxiety-Induced Insomnia | 11 | 69% |
| Household Friction (Waking up partners/ruining family routine) | 11 | 69% |
Findings
The interviews revealed a consistent pattern: participants spend enormous effort devising their own solutions to adjust between shifts, and they deeply resent how different rota schedules force completely different approaches at home. Multiple participants described how their rotas change every six months — some rotations are considerate and manageable, while others are brutal due to understaffing, with day-to-night switches happening far too quickly for the body to adapt.
The core pain point was rota scheduling itself, and everything else cascaded from it. The "Wired State" — the inability to wind down after high-stress shifts — was raised by every single participant (100%). Rota resentment (94%), unsafe post-shift commuting (94%), and hydration/nutrition neglect (94%) followed closely. Participants described DIY recovery hacks including sleeping tablets, forced 24-hour resets, and improvised blackout setups — all symptoms of a system that gives them no real tools to manage the problem.
The success criteria was met: shift transitions were independently and unanimously identified as a top pain point, and the consequences described — unsafe driving, family disruption, reliance on medication — clearly go beyond what passive data alone can address.
How this informed next steps
This confirmed that the core problem is not a lack of sleep data — it's a lack of tools to adapt to constantly changing rota structures. Our approach became twofold: first, help individuals understand how to best recover and transition their sleep based on their specific rota, giving them personalised guidance instead of the improvised strategies they currently rely on. Second, use the aggregated sleep and fatigue data we collect to eventually help inform better rota design — turning individual recovery data into evidence that can push for systemic change.
The "systemic gaslighting" finding (88%) — that workers lack data to prove fatigue to management — further validated this long-term direction. If we can quantify the impact of bad rotas on sleep quality and recovery, we create a feedback loop that benefits both the individual and the institution.
Research question
Whether a guaranteed wake-up and active transition support resonates more with users than passive sleep tracking alone.
Hypothesis
Users will find a guaranteed wake-up with transition support significantly more compelling than passive sleep tracking alone.
Success criteria
At least 70% of users prefer the active wake-up narrative over passive tracking when presented with both options.
Methodology
We ran a paid A/B test on Meta (Facebook/Instagram) targeting medical workers in the UK. Two ad variants were served to matched audiences over a 14-day period, each driving to a dedicated landing page with a mailing list sign-up as the conversion event.
- Ad A: Passive sleep tracking for doctors — "Understand your fatigue. Personalised coaching to help you recover between shifts."
- Ad B: Guaranteed wake-up + transition support — "Never sleep through an alarm again. Smart wake-ups that help you shift between day and night rotations."
Findings
| Metric | Ad A (Passive) | Ad B (Wake-up) |
|---|---|---|
| Impressions | 12,400 | 12,600 |
| Click-through rate | 1.2% | 3.4% |
| Landing page → sign-up | 8.1% | 18.7% |
| Total sign-ups | 12 | 80 |
| Cost per sign-up | £6.25 | £0.94 |
Ad B outperformed Ad A across every metric. The tangible promise of a guaranteed wake-up and transition support generated nearly 3x the click-through rate, more than double the landing page conversion rate, and 6.6x more sign-ups at roughly one-seventh of the cost per acquisition.
How this informed next steps
This confirmed the guaranteed wake-up as the lead feature differentiating Somna from passive trackers. The passive coaching narrative failed to compel action — people scrolled past it. The tangible, outcome-focused message resonated immediately. This shaped all subsequent positioning and pricing tests, moving the product narrative from "track your sleep" to "shift your sleep."
How we mitigated this risk
Our initial assumption — that shift workers would pay for passive tracking — was wrong. Through iterative user interviews, we discovered the real pain point: the transition between shift types. Users have almost no flexibility between shifts, making passive data alone unhelpful. However, the window between shift rotations — where each person improvises their own recovery — emerged as a clear intervention point. This pivoted our entire product from a sleep tracker to a sleep transition tool, with the guaranteed wake-up as the lead feature. The risk is mitigated because the value proposition is now grounded in a validated user need rather than an assumed one.
Risk 2: The benefit we provide is quantifiable
Even if doctors like the concept, the business fails if we can't prove (a) the product creates real fatigue relief and (b) that relief is measurable in a way that supports short-term engagement and long-term retention. We need evidence for two time-horizons: short-term relief where users feel a meaningful improvement immediately after use (or within 24–72h), and long-term improvement where repeated use improves fatigue-related outcomes across weeks in a measurable way.
Research question
Does the sensor stack required to enable our core features (wake-up, sleep staging) also enable fatigue measurement — or does fatigue tracking require additional hardware or cost?
Hypothesis
The sensors needed for our core features (accelerometer, PPG, temperature) produce data that can also be used to derive fatigue-related metrics without additional hardware.
Success criteria
The core sensor stack can produce at least two validated proxies for fatigue (e.g. HRV trends, sleep efficiency, time-to-sleep) without requiring sensors beyond those needed for the wake-up and sleep transition features.
Methodology
To be completed — sensor specification review and cross-referencing feature requirements against fatigue measurement literature.
Findings
To be completed
How this informed next steps
To be completed
Research question
To be completed
Hypothesis
To be completed
Success criteria
To be completed
Methodology
To be completed
Findings
To be completed
How this informed next steps
To be completed
Research question
To be completed
Hypothesis
To be completed
Success criteria
To be completed
Methodology
To be completed
Findings
To be completed
How this informed next steps
To be completed
This is not a standalone test but the outcome informed by the tests above — translating whatever fatigue measurement and relief evidence exists into a compelling narrative for users, and validating that narrative through a pricing test. To be completed.
How we mitigated this risk
To be completed — pending test results on fatigue measurement capabilities and relief metrics.
Risk 3: How good is the data quality from common consumer wearables?
Do existing smartwatches provide sufficient data for accurate smart wakeups? If so, there is a risk that Somna's core functionality could be quickly replicated by existing wearable companies.
Research question
Do existing smartwatches provide sufficient data for accurate smart wakeups? If so, there is a risk that Somna's core functionality could be quickly replicated by existing wearable companies.
Hypothesis
Wristband wearable data is insufficient for reliable smart wakeups. Consumer wearables focus primarily on fitness tracking, and their ability to detect sleep stages is structurally limited by the sensors they use. Accurate detection of light sleep requires higher-quality physiological signals than typical wrist devices can provide.
Conclusion
Independent validation shows that accurate sleep stage detection is possible, but common consumer wearables face fundamental limitations that make them unsuitable for accurate sleep tracking.
Sensor architecture of consumer wearables
To investigate this question, our team analysed the sensor technology used by major wearable devices.
Most consumer wearables measure heart rate using photoplethysmography (PPG). This technique uses light emitted at the skin surface to estimate blood circulation. PPG is widely used because it enables convenient wristwatch-style devices, but it measures an indirect physiological signal.
In contrast, electrocardiography (ECG) measures the electrical activity of the heart directly. ECG signals are significantly stronger and less sensitive to motion artifacts than wrist-based optical signals.
Fitness devices therefore favour PPG sensors for convenience, but this comes at the cost of reduced signal quality.
Our team hypothesised that there may be a substantial gap between the marketing claims of consumer wearables and the true reliability of their sleep tracking data. As an illustrative test of sensor reliability, an Apple Watch can consistently report a heartbeat when attached to an orange rather than human skin. This demonstrates how optical sensors can sometimes interpret non-biological signals as heartbeats.
Independent clinical validation
To independently test our suspicions, our team conducted a thorough review of the literature from well-established sleep clinics. This key multicentre validation study analysed 3890 hours of sleep classification results from common consumer wearables. This study repeatedly demonstrated that PPG readings from existing wearables perform poorly at detecting sleep phases. All the wearables tested had sleep-stage classification accuracy between just 0.50 and 0.65.
Sleep stage misclassification
This confusion matrix shows the high level of disagreement between polysomnography and consumer sleep trackers. Each row in the confusion matrix corresponds to the sleep stage annotated by polysomnography (PSG), while each column corresponds to the sleep stage annotated by the wearable. Crucially, when the user is in deep sleep or REM, all of the wearables had at least a 30% chance of misclassifying these phases as light sleep or waking. This would trigger a wakeup alarm at the worst possible moment.
Key investor insight
Clinical sleep studies demonstrate that sleep stage detection is clearly achievable when high-quality physiological data is available. However, consumer wearables are structurally constrained by the limitations of wrist-based optical sensing. While full polysomnography requires clinical equipment and controlled environments, dedicated sleep hardware that incorporates higher-quality physiological signals could bridge this gap while remaining suitable for everyday use.
| Technique | Signal Source | Accuracy | Environment | Key Insight |
|---|---|---|---|---|
| Polysomnography (PSG) | EEG, ECG, respiratory sensors | Very high | Clinical sleep labs | Gold standard for identifying sleep stages |
| ECG-based monitoring | Heart electrical signals | High | Clinical or wearable | Heart rate variability strongly correlates with sleep stages |
| Wrist PPG (smartwatches) | Optical pulse sensing | Moderate | Consumer wearable | Lower signal quality due to motion and optical noise |
Pivot avoided
Had competitor data proven sufficient for our core outcomes, this would have triggered a pivot to a software-only product — abandoning custom hardware entirely and building on top of existing wearable platforms.
Research question
Whether a software-only product built on top of existing wearable platforms has commercial viability as a secondary offering alongside dedicated hardware.
Hypothesis
Even though existing wearables can't deliver our full feature set, a reduced-capability software tier has meaningful standalone value and can serve as a viable commercial pathway.
Success criteria
A/B testing shows significant user interest in a software-only value proposition, validating two distinct commercial pathways: wearable + software (full features) and software-only (reduced tier).
Methodology
We updated the value proposition to reflect two product tiers: a wearable + software bundle delivering the full feature set (wake escalation, light-sleep timing, wake confirmed loop), and a software-only version offering reduced capabilities (shift planning, basic sleep guidance, schedule optimisation) for users who already own a consumer wearable. We ran A/B tests comparing user interest and willingness to pay across both propositions.
Findings
The software-only tier retains meaningful value for users who already own a consumer wearable. While it cannot deliver the core hardware-dependent features (wake escalation, wake confirmed loop), it still provides shift-aware sleep planning, schedule optimisation, and basic sleep guidance powered by the data their existing device can provide. A/B testing confirmed significant interest in both tiers, validating two commercial pathways: wearable + software for full capability, and software-only for a lower barrier to entry.
How this informed next steps
Validated a dual commercial model. The software-only tier reduces barrier to entry, broadens the addressable market, and creates an upsell path — users who start with software-only can upgrade to the full wearable + software experience. This also de-risks the business: if hardware faces delays or cost issues, the software-only tier provides revenue and market presence.
Research question
Which device form factors unlock the most core features for our target market? A successful hardware design must collect accurate sleep phase data while providing reliability, comfort, and convenience.
Hypothesis
Placement of the sensor on the chest offers the highest accuracy sensor readings for sleep phase detection and smart wakeups. Other sensor locations may fail to provide a significant differentiation in data quality from common consumer wearables.
Conclusion
A flexible puck form factor with two wearing positions (chest and upper arm) offers clear competitive positioning. Both positions unlock core features that wrist-based devices cannot provide: chest placement for the best data quality and wake experience, upper arm placement for comfort and chest sleepers. A flexible form factor serves a larger number of customer segments.
Electrocardiography (ECG) data quality
ECG measures the electrical activity of the heart, providing significant insight into cardiac function. These readings provide reliable insights into sleep activity.
We conducted an experiment to evaluate the suitability of ECG monitoring for sleep-phase detection across various sensor attachment locations. Implementing this technique involves a trade-off between accuracy and convenience: clinical-level ECG monitoring requires shaving chest hair and placing up to 12 leads (viewports) at various body landmarks. Clinical-level data is infeasible and excessive for everyday sleep monitoring purposes. Instead, our experiment used the single-lead ECG monitor on the Movesense programmable sensor to identify attachment locations that offer a clear view of the QRS complex and RR intervals.
In line with our hypothesis, placing the sensor on the chest provided unparalleled high-fidelity readings of cardiac electrical signals. While not as clear as the chest readings, the shoulder readings also provided sufficient data to calculate heart rate and heart rate variability to a good standard. In contrast, the wrist and other placement locations resulted in significantly more noise than actual signal. This experiment revealed merits in both shoulder and chest-based wearable placement.
| Sensor Location | Signal to Noise Ratio (SNR) | Data Quality | ECG Signal |
|---|---|---|---|
| Wrist | -6.5 dB | Low |  |
| Chest | 0.8 dB | High |  |
| Shoulder | -0.5 dB | Medium |  |
| Ear | -5.2 dB | Low |  |
| Thigh | -7.8 dB | Low |  |
Research question
Placeholder
Hypothesis
Placeholder
Conclusion
Placeholder
How we mitigated this risk
A sleep-specialised wearable is justified — existing wearable data is insufficient for our core feature set (wake escalation, light-sleep timing, wake confirmed loop, medical-worker constraints). Two commercial pathways are validated: wearable + software for full features, and software-only as a reduced tier for existing wearable owners. A flexible puck form factor supports both chest and upper arm placement, unlocking different strengths for different user preferences. The positioning strategy is clear: medical workers as the niche beachhead (can't wear wrist devices during shifts), broadening to data-addicted users (charge their Apple Watch at night, missing sleep data).
Risk Assessment
Beyond the three core assumptions tested above, these are the forward-looking risks that must be mitigated as the company scales from validation to launch.
Feature Absorption by Incumbents
HighEstablished wearable brands — Apple, Garmin, Oura, Whoop — already have large user bases, mature supply chains, and dedicated sleep-tracking teams. Without strong IP protection, many of Somna's core features (smart alarm timing, sleep-phase detection, shift-schedule planning) could be replicated as incremental updates to existing products. These companies can ship features faster and distribute them to millions of existing users at near-zero marginal cost, potentially commoditising our differentiators before we reach scale.
Mitigation: Our defensibility lies in specificity, not patents. The shift-worker use case demands a combination of features that no general-purpose wearable is optimised for: chest/arm sensor placement (medical workers can't wear wrist devices on shift), silent haptic escalation wake-ups, dual form factor flexibility, and an algorithm trained exclusively on shift-worker sleep data. We need to build a proprietary dataset and community moat that makes switching costly — personalised models that improve over months of use, employer-facing dashboards, and integrations with rostering systems. Speed to market and depth of niche focus are the primary defences while the IP portfolio matures.
Revenue Sustainability
HighThe business model relies on two revenue streams: hardware sales (£125–£179 per unit) and software subscriptions (£3/month). If hardware adoption is slower than projected — due to price sensitivity, distribution challenges, or competition — the company becomes dependent on the software-only tier. At £3/month, the software subscription alone may not deliver enough perceived value to retain users long-term, especially if they're using it with a general-purpose wearable that provides a degraded experience compared to the Somna Puck. Low retention at the software tier combined with slow hardware uptake creates a scenario where neither revenue stream independently sustains the business.
Mitigation: Validate willingness-to-pay early through pre-order campaigns and pricing A/B tests before committing to manufacturing volumes. Explore tiered software pricing — a higher-priced premium tier bundled with clinical insights, employer reporting, or telehealth integrations could increase ARPU without requiring hardware. Build the software experience so it delivers clear, measurable value even without the Puck (shift schedule optimisation, sleep debt tracking, circadian coaching) while making the hardware upgrade a compelling upsell rather than a requirement. Monitor unit economics closely and set clear kill criteria: if CAC exceeds 3-month LTV at the software tier, pivot distribution strategy before scaling spend.
Hardware Bypass via Supplier Direct Purchase
MediumBecause the current hardware is based on the Movesense sensor platform — which is commercially available — technically savvy users could purchase the same sensor directly from Movesense at a lower price. Since Somna offers a software + existing wearable subscription tier, these users could then pair the cheaper sensor with the Somna app and access the full feature set without ever buying the Somna Puck. This undermines the hardware margin and effectively turns the company's own compatibility into a loophole that cannibalises the premium product.
Mitigation: In the short term, this risk is limited — Movesense sensors are not consumer-friendly (no retail packaging, no onboarding, requires technical setup) and the overlap audience is small. As we transition to custom hardware, the bypass disappears entirely since the Somna Puck will not be available from third parties. In the interim, the software tier can be structured to differentiate: full feature access (wake escalation, confirmed-wake loop, clinical-grade sleep staging) requires a verified Somna device, while the general wearable tier offers a reduced feature set using less accurate data. This creates a clear value gap that makes the Puck worth paying for, without punishing users who genuinely want to try the software first with their existing wearable.
Value Proposition
- Shift workers have no purpose-built fatigue management tools — Somna automates sleep schedule transitions using a dedicated wearable and companion app
- Silent haptic wake-ups timed to light sleep phases replace jarring alarms, without disturbing partners or requiring a phone on the nightstand
- Clinical-grade PPG, temperature, and accelerometry provide real sleep stage detection — not motion guesswork — enabling personalised recovery plans that improve over time
Business Model
Our product model emerged directly from validation findings. The dual wearable approach (chest + arm) was confirmed through user testing, and the tiered pricing structure reflects ongoing pricing sensitivity tests.
Business Model Canvas
Somna monetises through a tiered model: a free companion app drives adoption, a £3/month subscription unlocks personalised sleep-shift plans using existing wearables, and the Somna Puck (£125–£179) adds clinical-grade ECG/PPG sensing for the highest accuracy. The beachhead market is medical shift workers — nurses, doctors, and paramedics — expanding to 600M+ shift workers globally. Key partnerships with hardware manufacturers, sleep clinics, and research institutions de-risk both the product and go-to-market. Long-term, algorithm licensing to third-party wearable OEMs creates an additional revenue stream.
Key Partners
- Hardware manufacturers (sensor suppliers, PCB assembly)
- Healthcare providers (sleep clinics, physicians, for clinical validation & referrals)
- Research institutions (university sleep labs, co-developed studies)
- Distribution partners (pharmacies, wellness retail, e-commerce)
- Regulatory consultants (TGA, FDA, CE/MDR certification)
Key Activities
- Wearable hardware + companion app development
- Sleep algorithm training and personalisation
- Clinical validation studies with real shift workers
- Paid acquisition testing (Google/Meta A/B campaigns)
Value Propositions
- Only purpose-built sleep wearable for shift workers
- Silent haptic wake-ups timed to light sleep phases
- Personalised recovery plans and sleep debt tracking
- Dual form factors (chest for accuracy, arm for comfort)
- Hardware + software moat: clinical-grade sensors paired with adaptive algorithms
Customer Relationships
- Self-service app signup (free tier, zero friction)
- Research transparency: findings published openly
- Tiered engagement: free app → subscription → hardware
- Mailing list for updates and research participation
Customer Segments
- Beachhead: medical shift workers (nurses, doctors, paramedics)
- Broader shift workers (600M+ globally, including factory, logistics, hospitality, security)
- Frequent travellers (business, flight crews, digital nomads)
Key Resources
- Proprietary sleep-shift algorithm and adaptive models
- Accurate sensor stack (PPG, ECG, temperature, accelerometry)
- Growing proprietary dataset of shift worker sleep data
- 6-person team across research, engineering, product, design, and finance
Channels
- Direct website and targeted landing pages
- Google/Meta paid acquisition (A/B tested)
- Mailing list and waitlist pre-orders
Cost Structure
- Hardware manufacturing (sensors, PCB, casing, dual SKUs)
- R&D (algorithm development, clinical validation, prototyping)
- Regulatory compliance (medical device classification, data privacy)
- Cloud infrastructure (backend, data processing)
- Marketing and customer acquisition (paid ads, content)
Revenue Streams
- App + existing wearable subscription (£3/month recurring)
- Somna Puck hardware sales (£125 pre-order / £179 launch, two SKUs)
- Algorithm licensing to third-party wearable manufacturers (future)
Key Partners
- Hardware manufacturers (sensor suppliers, PCB assembly)
- Healthcare providers (clinical validation & referrals)
- Research institutions (university sleep labs)
- Distribution partners (pharmacies, wellness retail, e-commerce)
- Regulatory consultants (TGA, FDA, CE/MDR)
Key Activities
- Wearable hardware + companion app development
- Sleep algorithm training and personalisation
- Clinical validation studies with real shift workers
- Paid acquisition testing (Google/Meta A/B campaigns)
Value Propositions
- Only purpose-built sleep wearable for shift workers
- Silent haptic wake-ups timed to light sleep phases
- Personalised recovery plans and sleep debt tracking
- Dual form factors (chest for accuracy, arm for comfort)
- Hardware + software moat: ECG sensors paired with adaptive algorithms
Key Resources
- Proprietary sleep-shift algorithm and adaptive models
- Accurate sensor stack (PPG, ECG, temperature, accelerometry)
- Growing proprietary dataset of shift worker sleep data
- 6-person team across research, engineering, product, design, and finance
Customer Relationships
- Self-service app signup (free tier, zero friction)
- Research transparency: findings published openly
- Tiered engagement: free app → subscription → hardware
- Mailing list for updates and research participation
Channels
- Direct website and targeted landing pages
- Google/Meta paid acquisition (A/B tested)
- Mailing list and waitlist pre-orders
Customer Segments
- Beachhead: medical shift workers (nurses, doctors, paramedics)
- Broader shift workers (600M+ globally, including factory, logistics, hospitality, security)
- Frequent travellers (business, flight crews, digital nomads)
Cost Structure
- Hardware manufacturing (sensors, PCB, casing, dual SKUs)
- R&D (algorithm development, clinical validation, prototyping)
- Regulatory compliance (medical device classification, data privacy)
- Cloud infrastructure (backend, data processing)
- Marketing and customer acquisition (paid ads, content)
Revenue Streams
- App + existing wearable subscription (£3/month recurring)
- Somna Puck hardware sales (£125 pre-order / £179 launch, two SKUs)
- Algorithm licensing to third-party wearable manufacturers (future)
Stakeholder Map
An overview of everyone who stands to benefit from Somna, from individual shift workers to the institutions that employ them.
Healthcare
Transport
Service
Emergency Services
Institutional
Partners
Distribution
Market Analysis
Top-down analysis using the global healthcare shift-worker population multiplied by our direct competitors' weighted average ARPU.
$3.7 billion
Global healthcare workers working shift work, multiplied by our direct competitors' weighted average ARPU.
$336 million
Healthcare workers in English-speaking countries working shift work, multiplied by weighted average ARPU.
$8.4 million
2.5% of healthcare shift workers in English-speaking countries, multiplied by weighted average ARPU.
Bottom-up analysis using WHO data identifying 7.8M English-speaking clinicians (1.7M doctors + 6.1M nurses & midwives), combined with realistic digital reach estimates across creator channels, paid ads, and healthcare online communities.
$553.8 million
7.8M English-speaking clinicians (1.7M doctors + 6.1M nurses & midwives) multiplied by competitors' weighted average ARPU.
$154 million
~2.17M clinicians reachable via creator channels, paid ads, and healthcare online communities, multiplied by weighted average ARPU.
$3.9 million
2.5% of healthcare shift workers in English-speaking countries reached via marketing campaigns, multiplied by weighted average ARPU.
Traction
Traction Channel Framework
Our go-to-market traction channels, based on Gabriel Weinberg's framework. Click any dot to reveal the channel name.
Live Registered Interest
Live registered interest from the pricing page. These numbers update automatically.
0
Total Subscribers
0
App + Wearable
0
App + Somna Puck
Acquisition Pathways
Breakdown of where subscribers are coming from. Existing subscribers without tracked sources show as "Unknown".
No source data available yet. Source tracking starts with new signups.
Business Development
Go-to-market strategy, acquisition channels, and growth plans, informed by the advertising and positioning questions identified in the assumption-mapping process.
Acquisition Channels
Marketing Channels
- → Social media & ads
- → Blogs
- → Influencer marketing
- → Email campaigns
From now
Partnerships — for trials and community building
- → Unions
- → New doctors
- → Movesense — medical-grade sensor platform for hardware R&D and validation
From now
Events
- → Speaker events
- → Pop-ups at universities or hospitals
From full launch (~18 months)
Rollout Timeline
Now
Software + Integration Launch
—
Full Launch
6 months · 200 people
Software + Integration Beta Testing
6 months
Collect app engagement data to drive wearable R&D
6 months · ? people
Hardware Beta Testing
Hardware R&D
Now
6 months · 200 people
Software + Integration Beta Testing
Software + Integration Launch
6 months
Collect app engagement data to drive wearable R&D
Hardware Beta Testing
6 months · ? people
Full Launch
Hardware R&D runs in parallel throughout
Intellectual Property
IP considerations played a direct role in shaping the product strategy. The decision to offer the wearable and software as separable products was driven not only by market demand but by the need to protect and maximise the value of distinct IP assets: the algorithm and data models on one side, and the hardware design on the other.
How IP shaped the business model
Software IP
The sleep-shift algorithm, sleep phase detection, and personalised scheduling logic represent core software IP. By offering the app as a standalone product, compatible with existing wearables, this IP can generate revenue independently of hardware, with lower unit costs and faster iteration cycles.
Hardware IP
The wearable design, optimised for night-only use with dual form factors (chest and arm), represents separate protectable IP. Selling hardware bundled with software creates a premium tier, while the hardware design itself is defensible against competitors who rely on general-purpose devices.
Strategic separation
Keeping software and hardware as separable offerings protects against single points of failure. If hardware manufacturing faces delays, the software business continues. It also opens licensing opportunities for the algorithm to third-party wearable manufacturers.
Partnerships
Strategic partnerships that de-risk hardware development and accelerate time-to-market.
Movesense
We partnered with Movesense to test the feasibility of our core sensor requirements. Movesense provides a medical-grade, open-source sensor platform with ECG, accelerometer, and gyroscope capabilities in a compact form factor, making it an ideal testbed for our sleep-phase detection algorithm and haptic wake-up timing.
This partnership allowed us to validate sensor data quality, battery life under continuous overnight use, and the viability of chest-strap and arm-strap form factors, without committing to a full custom hardware programme.
Contributions
GenAI was used to assist with the coding of this website.