Mavis.

Data room › Section 3 of 8

Business model

High-margin hardware with a subscription attached. One device, one monthly plan, an iPhone-style hardware refresh every five years or so. Assistive devices designed solely for use by a disabled person are zero-rated for VAT in the UK, so the price is what the user actually pays.

£1,230Recommended retail price per unit, zero-rated for VAT. The low-vision wearables it competes with sit between £2,600 and £5,150.
£30Per month including VAT (£25 net, £300 ARR per subscriber), covering live emergency support, GPS and AI updates, new features and warranty
84%Hardware gross margin on a £200 unit cost built line by line from 146 supplier-priced items; 57–63% blended across hardware and subscription in the P&L

Routes to the customer

Disabled Students’ Allowance (live: sole wearable AI product on the SLC catalogue) and Access to Work (live: supply agreement with Microlink) pay for the device on the user’s behalf. Charity referrals through Visionary’s 110+ local charities and a 20%-commission Ambassador programme. Retail distribution with VisionAid, Sight and Sound and Optelec in negotiation. NHS low-vision pilots through NHS Innovation. Direct sales through the website and events for everyone else.

Why the niche is defensible

Consumer smart glasses are a phone accessory: no phone, no AI. Their makers rule out hazard detection as a safety matter and face recognition as a privacy matter, and accessibility exists, in the American Foundation for the Blind’s words, “purely by chance”. To them blind people are an edge case; to Mavis they are the whole business. Mavis is standalone, built for UK funding routes, patented, and able to ship user feedback as updates in hours rather than months.

Path to profitability

The model needs about £365,000 to trade through 2027 and roughly £480,000 to the cumulative low point at the end of 2028, before grants and R&D tax credits; this round covers that with headroom. EBITDA turns positive in 2028, operating cash flow in 2029. Six-year forecast: £788k revenue in 2027, £2.6M in 2028, £6.6M in 2029, £13.7M in 2030 and £22.4M in 2031 at a 61% blended gross margin and £5.55M EBITDA (25%). Forecasts, not guarantees.

The financial model

The six-year model — unit economics costed from supplier invoices, break-even analysis, cash flow, headcount plan and exit scenarios — is in the Documents section for shareholders, or available on request to qualifying investors.

What each unit earns

Direct unit cost of £154.49 — main board £26.42, Smart Case £45.59, camera, battery and PCB fabrication £20.39, LED and front boards £9.28, assembly labour and tooling £12.81, mechanical, packaging and logistics £40 — or about £200 once R&D is amortised, corroborated by an outsourced PCBA quotation of £99.29 per unit at 200 units (April 2026). At £1,230 that is an 84% hardware gross margin before commission and a contribution of £631–£692 per unit after a 20–25% blended Ambassador and channel commission, a 3% refund and warranty provision and year-one subscription. Unit cost is set to fall by around half with volume.

Subscription economics

£30 a month including VAT is £25 net, £300 of ARR per subscriber. It costs £5.49 to deliver — £1.49 for the SIM and £4.00 for AI inference — a 78% gross margin. The plan assumes 80% of units take the subscription and 5% annual churn, giving 21,769 subscribers and about £5.2M of recurring revenue by 2031. Subscription grows from 6% of revenue in 2027 to 23% in 2031; leasing could increase it further.

The market

The UK alone takes us to exit; everything else is upside

340,000 people are registered blind or partially sighted in the UK (RNIB), 2 million live with sight loss, 295 million worldwide. Penetration varies sharply by age group, so the plan is weighted, not a flat percentage.

Students and working age — 90,000

The smaller share of the registered population, but the highest digital adoption and two funded routes already live (DSA, Access to Work). Planned penetration ~3–5%.

Elderly, 65+ — 240,000

The majority of registered sight loss, with lower device adoption and no funded route today; the NHS pilot and charity channels address this group. Planned penetration ~0.5–1%.

Children and young students — 10,000

The smallest segment but the most tech-savvy, and Mavis is the only assistive glasses made in small sizes for children. Planned penetration ~0.5–1%.

30,000Cumulative units by 2031 in the plan: 8.8% of the registered UK population, ~1.5% of everyone with sight loss in the UK, 0.1% of Europe
£418MConservative UK TAM per replacement cycle; SAM £209–277M; SOM £738K in 2027 (600 units) and £2.21M in 2028 (1,800 units)
£7.9BTAM across the first six European markets (~6.4M reachable people); SAM £680–890M; sales from 2028 with no new certification needed
£1.35–8.8BNorth America TAM (~1.1M legally blind, ~7.15M with vision loss); a mid-range priority that needs FCC certification and a US presence, funded from proven European traction
Where we go first
RegionBlind / partially sightedLow visionPriority
United Kingdom340K registered2M with sight lossLive now
Europe — DE, FR, NL, SE, DK, FI first7M23M2028
North America2.1M12MAfter Europe; FCC required
India9M70MLater
MENA3.1M54MLater

The assistive-technology market as a whole was £14.2B in 2023, growing at 8.5% a year to a projected £23.7B by 2030. Europe and other markets are not a near-term priority, though we will cautiously pursue distribution agreements; a large and growing global market also underpins the exit to a strategic buyer or a private-equity aggregator.

Why the others failed

Product design and the pace of AI must be under your own control

eSight, OrCam and IrisVision each failed or nearly failed for a version of the same reason: they depended on someone or something outside their control.

eSight

Seven years to first product, outsourced manufacturing, about $90M spent, never independently profitable. Sold to Gentex in 2024.

OrCam

Smartphone AI ate its core function; revenue concentrated in one fragile region; a cap-table fight froze recovery. Around $90M raised.

IrisVision

Built entirely on Samsung’s Gear VR hardware, which was discontinued in 2020, taking the product roadmap with it. Around $12M raised.

Mavis is fully independent

Software, firmware, hardware, moulding, design and manufacturing are all in-house — nothing rented from a platform that can be discontinued. A certified, revenue-generating product on roughly £450,000 of capital, a fraction of what the comparators spent to reach the same point, with a team of five rather than dozens.

The IP floor

Taking a cost approach, a prudent 30% of R&D and capital invested gives the Mavis IP portfolio an unproven value of around £14M, above the £2M enterprise value of this round. The UK patent runs to the mid-2040s, so the point of maximum value to a strategic acquirer is 2028–29: sales scaling, EBITDA turning positive, and 17–18 years of patent life remaining. Comparators and figures are best-available public estimates, not audited.

Competition

A comprehensive solution at a fraction of the price

Zero-rated for VAT, so £1,230 is what the user actually pays.

FeatureMavisIrisVisioneSightNuEyesRay-Ban MetaEnVision
Navigation & obstacle detection✓✓✗✓✗✗
Text to speech✓✓✓✓✓✓
Object recognition & description✓✓✓✓✓✓
Emergency audio / video call✓✗✗✓✗✗
Custom applications / modules✓✗✗✗✓✗
Surface analysis & measurements✓✗✓✗✗✗
Made for blind & visually impaired✓✓✓✓✗✓
Unit price (RRP)£1,230£2,600£5,150£3,995£379£528