What Simma is
Vision. A permanent operating system for the business, with an AI team working on top of it. The one system runs its website, email, customers, bookings, products and money, all Simma's own. Other AI platforms plug into the tools a business already pays for (Google, HubSpot, Xero) and sit on top; Simma is the tools, so its AI team works on one record of the business, not ten.
Who it's for
Plan. First, instructors and studios, where the product fits best today; then other service and commerce businesses. The same platform serves franchise networks (head office sets the products, price bands and brand; each location runs its own business inside them; see Franchise networks). Franchise networks are the next route, planned from year two: talks with one operator are paused for now, until the core product is complete, and reopen once a network can be shown working. Larger businesses come later: within five years the ambition is dedicated teams for mid-market and enterprise customers. That is an ambition, not a plan with a budget: those teams' costs are not in the current model.
Where it stands today
Measured, status.html re-measure, 2026-09-24.
- In production since 2026-09-03; built solo by the founder.
- 0 paying customers and £0 revenue — pre-launch. Paid onboarding starts when the founder goes full-time at the first close.
- Two design partners: a UK dance school, whose website, timetable and bookings were built in Simma; and a UK media business the founder co-founded, which is shaping the publishing and multi-site side. Both shape what we build next. Neither is named.
- The full, dated list of what works and what doesn't is on the status page.
The company
Measured. Simma Labs Limited, United Kingdom. Founder: Paul O'Connor.
The ask
Plan. £250k SEIS Angel round: a priced round of ordinary shares, all under SEIS, which is the SEIS company limit. Six months to prove the Local Partner model (30 paying customers by month six), raising the pre-seed from month 5 on the evidence. £3m pre-money; minimum investment £10k (lower by agreement). The first of four steps: £250k Angel round → ~£1.5m Pre-seed → ~£5m Seed → Series A, each unlocked by evidence, not dates.
Four layers, one record
Vision.
- See it — six lenses across the whole business: Home (the business feed), Compass, Inbox, Calendar, Insight, Cashflow.
- Run it — thirteen areas: Marketing, Website, Sales, Customers, Support · Services, Products, Work, Money · Library, Team, My Business, Settings.
- Meet it — where customers and staff touch the business: the website, the customer portal, Simmi (the business's own AI for its customers), Simma Station (in-person).
- Found — Google Business Profile, SEO and AI search, kept up to date from the same record.
The AI team
Vision. Every account gets an AI team, on every plan including Free; how much it can do grows with the plan. Six AI teammates — Maya (marketing), Sam (sales), Charlie (customers), Molly (money), Oli (operations), Peter (products and pricing) — coordinated by Sage, the chief of staff. They draft; the owner approves. Nothing the AI team starts reaches a customer or moves money without a person's yes. When a customer asks (a booking, a question), Simmi answers within the timetable, prices and policies the business has set, and the customer chooses and pays. Measured. Agent identities, permitted tools, per-step attribution and an audit log are built; the named business agents are the next build.
The Cascade
Measured. One platform, a version for every business: eight layers (platform, industry, specialism, niche, franchise, country, business, person) decide features, menus, words, emails and currency. The industry, specialism and niche layers are built, and industries are added as configuration, starting with instructors; 23 countries are mapped; 180+ email templates are published across the layers. That is configuration, not markets served: today's website templates are for dance schools, and several layers are still waiting for content (see the status page).
Franchise networks
Measured (read from the code, 11 October 2026). Head office keeps one product catalogue and sets a price band for each product; each location prices inside that band, by a percentage or a fixed price, and anything outside it is refused. A brand site has a location finder that leads to each location's own site. Every paid booking or order records the franchisor's royalty against that payment, and a cancellation reverses it. That is a record of what's owed; the money itself is not split when the customer pays. Not yet shown: the whole journey working end to end for one network (head office to location to customer to payment to refund), and a central catalogue of services as well as products. Talks with a multi-brand franchise operator are paused until the core product is complete; nothing is contracted. More: Franchise networks.
Simma builds Simma
Measured (2 October 2026). Simma's own development runs through Simma's task system: 2,532 work items, 1,288 commits attached to 243 items, and 203 coding-agent check-ins since 17 September, 197 of them checked against the commits. 14 went outside the agreed area, each covered by a ticket or an approved extension. Agents asked 77 questions instead of guessing.
Honest position
Measured, 2026-10-01.
- Paying customers: 0. Revenue: £0. Design partners: 2 (not named): Two design partners: a UK dance school, whose website, timetable and bookings were built in Simma; and a UK media business the founder co-founded, which is shaping the publishing and multi-site side. Both shape what we build next. Neither is named.
- Production last released 2026-09-03; 2,120 commits await the next release (everything on the development branch since then, documents included).
- Customer websites on simmasite.com are not reachable today — fixed in code on 2026-10-01, waiting on DNS and deployment steps.
- Known gaps are named on the status page, including: Simma Pay's plan prices not yet applied on every payment path; the Google Business Profile connection; customer-portal booking changes.
What is not claimed
No measured customer economics are given: acquisition cost, lifetime value, churn and retention can't be measured without customers. The margin and churn figures in the financial model are assumptions, labelled as plan.
Estimated unit economics (plan, not measured)
From the v10 model (draft prices, 11 October 2026), before VAT, with 60% of customers using Simma Pay. Direct paid (one customer who signs up straight to Lite; before the support team and any Free accounts): about £1,652 a year, contributing about £109 a month after hosting, messaging, AI use, Stripe Connect and payment losses. With 40% leaving each year, lifetime value is about £3,284; at £400 to win (the model's year-3 cost per paying self-serve customer, calibrated from the earlier model's guess, not measured) that is about 8.2x, earned back in about 4 months. At the first customers' mix (55% Lite, 35% Starter, 10% Growth), direct paid is about 19.6x. Fully loaded self-serve funnel (year 3: only 22.4% of sign-ups end up paying; the cost to win adds every Free account's running cost, about £41 each over its life, and lifetime value takes off support at about £14 a paying customer a month): about £548 to win, 12.7x, earned back in about 3 months. The direct-paid figures alone are not the economics of self-serve. Working, and every plan: model.html#unit-economics.
Company gross margin, after serving customers, Stripe Connect's charges, payment losses and the customer success and support team, is 57% in year 1, 70% in year 2 and 72% by year 5 (77% on the earlier basis, with Connect netted off revenue). Every AI action is metered against the business and the person who asked, and each business has an enforced monthly AI budget; allowances that rise with the plan are set before the first paying customer. Free accounts cost money: about £2 a month each while in use, after their payment fees; making Free cheaper to serve is being decided now.
How many small businesses close (measured, ONS)
Of UK businesses started in 2019, 94.6% were trading after one year, 74.7% after two and 38.4% after five. Instructor businesses do better: after five years, 50.5% of other-education and 44.1% of sports businesses were still trading. Source: ONS Business demography, UK: 2024 (VAT or PAYE registered businesses only).
Sizing
Estimate, 2026-10-11, at the v10 draft prices. Priced at what a business would spend with Simma (subscription plus Simma's payment fee), not total software spend.
- TAM ~£405bn a year — every business in the world, if all of them ran on Simma. £118.1bn of it in the UK and US.
- SAM £114.9bn a year — UK and US businesses of every industry with fewer than 250 staff (30.5m businesses).
- Beachhead £50.8bn a year — UK and US service businesses (the wedge) with up to nine staff, solo included (17.9m businesses).
- First customers: 15,560 registered UK instructor businesses and studios (412,515 education and sport businesses with up to nine staff, most self-employed): sports coaching, dance and music, driving schools, fitness.
Obtainable
Plan. 53,513 paying customers by the end of year five, an output of the v10 model (not a target): 26,735 Lite, 20,083 Starter, 6,359 Growth, 263 Pro, 34 Scale, 39 Enterprise; £120.9m revenue in year 5, a run rate of about £165.6m at the end of it. Built bottom-up from what Simma spends to win customers, not as a share of SAM.
Sources
UK Business Population Estimates 2025; US Census NES/SUSB 2022; Eurostat 2023; World Bank 2025 income groups; SME Finance Forum / Statista. Method: v10: solo half Lite (£99 inc. VAT) half Starter (£229 inc. VAT); micro half Starter half Growth (£449 + VAT); small Pro (£799); medium Scale (£1,499); extra members at the model's averages; enterprise a £5,000-a-month contract, no payments counted. Before VAT; outside the UK the UK price before VAT; world: UK + US spend per head scaled to each World Bank income group by income per head (x3.43 overall, unchanged from v9).
Subscription
Plan: draft prices, 11 October 2026, not yet charged. Priced per business, per month. Extra team members £5 a month. 20% off for paying a year up front, extra team members included; members added mid-term prorated to renewal.
- Free — £0 a month · Starting out or trying Simma · Simma's payment fee 2% + 20p · 1 team member · 100 customer records · 1 location · AI allowance: Taster
- Lite — £99 a month inc. VAT · A solo instructor or therapist · Simma's payment fee 1.5% + 15p · 2 team members · 500 customer records · 1 location · AI allowance: Light
- Starter — £229 a month inc. VAT · A studio with a few staff · Simma's payment fee 1% + 10p · 5 team members · 2,000 customer records · 2 locations · AI allowance: Standard
- Growth — £449 a month + VAT · An established studio, or 2–5 locations · Simma's payment fee 0.75% + 10p · 10 team members · 10,000 customer records · 5 locations · AI allowance: Plus
- Pro — £799 a month + VAT · Multi-site, up to about 10 locations · Simma's payment fee 0.5% + 5p · 25 team members · 50,000 customer records · 10 locations · AI allowance: Pro
- Scale — £1,499 a month + VAT · A chain or regional group, up to about 25 locations · Simma's payment fee 0.4% + 5p · 50 team members · 250,000 customer records · 25 locations · AI allowance: Max
- Enterprise — from £3,499 a month + VAT · Franchise head offices and large groups · Simma's payment fee negotiated (0.4% provisional) · team, customers and locations by contract · AI allowance: Custom
Every plan, including Free, includes the planned AI team (six teammates and Sage; what is live today is on the status page); its allowance rises with the plan. Each plan's users, customer records, locations and AI allowance are set (Paul, 11 October 2026); AI credits top up any plan; who each plan is for is indicative, not a contracted limit. AI credits can be bought on any plan, Free included.
Simma Pay
Plan. Every payment a business takes in Simma runs through Simma Pay, built on Stripe Connect. The Free plan gives away the whole operating system with no monthly subscription: payment fees apply, and optional AI credit top-ups can be bought. Paid plans cost a subscription and add users, customers and AI-team capacity; they also take less per payment. One fee per payment, not itemised: the card's actual cost at what Stripe charges, plus Simma's fee. Only Simma's fee is revenue. Simma's fee falls with the plan: 2% + 20p (Free) down to 0.4% + 5p (Scale), so the total per payment is 3.5% + 40p on Free and 1.9% + 25p on Scale. Total per payment, online, with a standard UK consumer card: the card's actual cost (1.5% + 20p for this card, passed on at what Stripe charges) plus Simma's fee. Premium and international cards cost more; in person, the card cost differs (Stripe: 1.4% + 10p for UK cards). Whether Simma's fee carries VAT is pending accountant review: if it does, it is included on Free, Lite and Starter and added on Growth and up. Payouts weekly on Free, twice a week on Lite and Starter, every working day from Growth; on Free, Simma's fees are at least £2 in any month Simma Pay pays out. Stripe Connect's charges (£2 per active business a month, 0.25% + 10p per payout) are a cost to Simma. VAT treatment of the payment fee is an assumption pending accountant review (the model takes the cautious case). In the model, Simma's payment fees are 34.6% of year-5 revenue. The rates will be tested with the first studios before they are fixed. Plan. Proposed v10 rates, 11 October 2026. Applying them across all payment paths remains a delivery task; the status page records what is implemented.
Franchises
Plan. Custom contract.
Five-year model
Plan — draft 3.1; month by month for 60 months from tier cohorts, Free accounts by age, yearly contracts with deferred revenue, VAT when invoiced, corporation tax by HMRC timing (2026-10-11). As of 2026-10-11.
- Year 1 — revenue £201.1k (subscriptions £130.9k, AI credits £6.3k, Simma Pay fees £63.8k) · EBITDA −£494.7k · 271 customers at year end · 13 people
- Year 2 — revenue £2.7m (subscriptions £1.6m, AI credits £81.3k, Simma Pay fees £921.7k) · EBITDA −£135.0k · 1,670 customers at year end · 38 people
- Year 3 — revenue £13.8m (subscriptions £8.6m, AI credits £425.1k, Simma Pay fees £4.7m) · EBITDA £999.0k · 7,110 customers at year end · 80 people
- Year 4 — revenue £46.7m (subscriptions £29.1m, AI credits £1.4m, Simma Pay fees £16.2m) · EBITDA £9.9m · 22,408 customers at year end · 167 people
- Year 5 — revenue £120.9m (subscriptions £75.3m, AI credits £3.8m, Simma Pay fees £41.8m) · EBITDA £32.4m · 53,513 customers at year end · 307 people
Base plan: positive EBITDA in year 3 (year 2 EBITDA is −£135.0k); EBITDA is not accounting net profit, and the downside cases below do not reach it. Gross margin rises from 69.8% in year 2 to 71.5% in year 5. Cash after amounts owed to HMRC at the year ends: £1.3m, £1.6m, £9m, £20.5m, £53.3m; lowest about £120.0k in month 6, just before the pre-seed lands. Assumptions pending accountant review: VAT on Simma's payment fees (cautious case: they carry VAT), and corporation tax at 25% of EBITDA less losses brought forward, paid by HMRC's normal, large or very large company timing.
When the guesses are wrong
Plan (scenarios). Spending is unchanged unless the scenario explicitly specifies a hiring or marketing change.
- Base plan — EBITDA year 2 −£135.0k, year 5 £32.4m; lowest cash after HMRC £120.0k (month 6)
- Free conversion 1% — EBITDA year 2 −£181.9k, year 5 £29.1m; lowest cash after HMRC £119.8k (month 6)
- Free conversion 5% — EBITDA year 2 −£88.1k, year 5 £35.7m; lowest cash after HMRC £120.2k (month 6)
- Payment fees VAT-exempt — EBITDA year 2 £7.0k, year 5 £38.7m; lowest cash after HMRC £120.6k (month 6)
- Customers choose cheaper plans — EBITDA year 2 −£734.0k, year 5 £5.8m; lowest cash after HMRC £110.8k (month 6)
- Simma Pay at half take-up — EBITDA year 2 −£483.3k, year 5 £16.8m; lowest cash after HMRC £118.6k (month 6)
- Free accounts rarely take payments — EBITDA year 2 −£186.4k, year 5 £29.8m; lowest cash after HMRC £119.8k (month 6)
- Free: half the AI allowance — EBITDA year 2 −£123.6k, year 5 £33m; lowest cash after HMRC £120.1k (month 6)
- Free: half the hosting cost — EBITDA year 2 −£89.3k, year 5 £34.8m; lowest cash after HMRC £120.4k (month 6)
- Free: unused accounts cost nothing — EBITDA year 2 −£131.6k, year 5 £32.7m; lowest cash after HMRC £120.0k (month 6)
- Free: all three cuts — EBITDA year 2 −£74.4k, year 5 £35.6m; lowest cash after HMRC £120.5k (month 6)
- Dearer to win (half as much again) — EBITDA year 2 −£576.7k, year 5 £6.5m; lowest cash after HMRC £114.1k (month 6)
- Higher churn — EBITDA year 2 −£216.1k, year 5 £24.6m; lowest cash after HMRC £119.9k (month 6)
- AI used to the full allowance — EBITDA year 2 −£243.4k, year 5 £27.3m; lowest cash after HMRC £119.3k (month 6)
- Pre-seed 3 months late, no cuts — EBITDA year 2 −£135.0k, year 5 £32.4m; lowest cash after HMRC −£9.7k (month 9); runs out in month 9
- Pre-seed late, marketing paused — EBITDA year 2 −£217.5k, year 5 £32.4m; lowest cash after HMRC −£12.4k (month 9); runs out in month 9
- Pre-seed late, hiring frozen — EBITDA year 2 −£135.0k, year 5 £32.4m; lowest cash after HMRC £87.4k (month 9)
- Yearly leavers idle for half their last term — EBITDA year 2 −£154.8k, year 5 £31.3m; lowest cash after HMRC £120.0k (month 6)
- Yearly leavers idle for their whole last term — EBITDA year 2 −£202.3k, year 5 £29.5m; lowest cash after HMRC £119.7k (month 6)
- No seed — EBITDA year 2 −£135.0k, year 5 £32.4m; lowest cash after HMRC £120.0k (month 6)
- No follow-on funding — EBITDA year 2 −£135.0k, year 5 £32.4m; lowest cash after HMRC −£227.4k (month 15); runs out in month 9
- Combined downside — EBITDA year 2 −£1.3m, year 5 −£24.7m; lowest cash after HMRC −£29.1m (month 60); runs out in month 37
- Combined downside, no follow-on funding — EBITDA year 2 −£1.3m, year 5 −£24.7m; lowest cash after HMRC −£35.6m (month 60); runs out in month 9
Rounds
- SEIS Angel round — £250k, now (priced SEIS round, £3m pre-money)
- Pre-seed — £1.5m, indicative; raised from month 5, planned to close in month 7
- Seed — £5m, indicative; start of year 3, for the US
- Series A — later; global scale, not needed by the five-year plan
The workbook, with live formulas, is available on request.
Terms
Plan.
- £250k SEIS Angel round — Priced round of ordinary shares, all under SEIS.
- Pre-money valuation: £3m (post-money £3.25m; the round is 7.69%).
- Minimum investment: £10k, lower by agreement.
- SEIS: the whole £250k under SEIS (the company limit). Advance assurance applied for before first close, and before anything is signed.
Use of funds (six months, then runway to raise)
Plan (draft, Paul 2026-10-03). People 64.5% · marketing and community 14.6% · product and infrastructure 12.9% · legal, admin and other 8.0%. £157.9k spent over six months; about £92.1k kept to raise the pre-seed, roughly 3 months at the month-six run rate of about £33k a month. Customer revenue in the period is treated as upside. Hires: the founder full-time from month 1, building with AI agents; employed Simma Local Partners from month 3 (#1) and month 4 (#2), in different parts of the UK, each trained on Simma by the founder, aiming for about 10 customers each in their first three to four months. The pre-seed is raised from month 5 and planned to close in month 7; a third Partner, the Head of Local Partners, customer success and two engineers (one owning infrastructure and releases) start in month 8. Month by month, in the v10 model: cash after amounts owed to HMRC is lowest at about £120.0k in month 6, so the pre-seed's timing is the plan's main risk. If it lands three months late, the hires it pays for wait until it does (see the FAQ).
Milestones
Plan (draft).
- Before first close — IP assigned to Simma Labs Limited; SEIS advance assurance applied for
- Months 1-2 — Founder full-time: completes the launch work; Local Partner #1 recruited
- Months 3-4 — Local Partner #1 starts (month 3), trained on Simma by the founder; Local Partner #2 (month 4), in a different part of the UK; First paying customers through Simma Pay
- Months 5-6 — About 10 customers per Partner in their first three to four months; Start raising the pre-seed (about £1.5m) in month 5, on the evidence; 30 paying customers by month six
- Months 7-9 — Close the pre-seed, planned for month 7; its final size and price are set by the evidence; Month 8: Local Partner #3, the Head of Local Partners, customer success, a third engineer and a DevOps engineer start
Paying customers: 30 by month 6. The five-year plan's year-one figure is 271.
The funding roadmap
Plan (indicative).
- SEIS Angel round — £250k, now. Prove the commercial model, at £3m pre-money.
- Pre-seed — about £1.5m, raised from month 5, planned to close in month 7. Prove the UK machine, to profitable UK growth.
- Seed — about £5m, start of year 3; raised from strength (in the base plan the UK business doesn't depend on it). Take it international, starting with the US.
- Series A — later. Global scale.
Later sizes are indicative and depend on the evidence; only the current round has a price. The Angel round, pre-seed and seed (£6.75m in all) fund the five-year model's cash plan, which needs no Series A. Cash after amounts owed to HMRC is lowest at about £120.0k in month 6, just before the pre-seed lands. In the base plan the business doesn't depend on the seed: without it (the pre-seed still arrives, and costs and countries stay as modelled), about £4m would still be in hand after amounts owed to HMRC at the end of year 3. Whether the US could open without the seed's people is not checked, and it does not cover the combined downside case (see 06 Financials).
Founder
Measured. Paul O'Connor — paul@simma.ai · https://www.linkedin.com/in/o-c/
- Head of Product, Event Platforms, a FTSE 100 global events company (2024-01–now): 50-person team, three squads; £3m+ budget; 2,000+ live sites; 386-site migration onto one platform
- Senior Product Manager, Event Platforms, a FTSE 100 global events company (2021-11–2023-12)
- Senior Product Manager, Vacancy Filler (now HireRoad) (2021-05–2021-11)
- Founder, Workwear City (2020–2021): 22,000+ SKUs; sold via trade sale 2021
- Managing Director, Rail Media (2017–2021): 40-person company; revenue doubled
- Product Director, Rail Media (2012–2017): built Datlyn (CRM, subscriptions, events, sales in one system); RailwayPeople.com to 1m+ monthly users and £100k+ monthly SaaS revenue
- Designer to Product Manager, Rail Media (2002–2012): there from day one: started by the founder and his father as a newspaper; his father ran sales, he ran product and technology, launched the recruitment platform and started the events business (Paul, 2026-09-27); launched RailwayPeople.com; co-designed iStudio Publisher ($100k/month in sales); sold it to its developers for a 50% revenue share (Paul, 2026-09-27)
The people who'd join
Plan. Eight senior people the founder has worked with for years, who know Simma and have asked to join (nothing agreed in writing; none is in the first round's budget; initials until they consent to be named): VT (Head of Engineering), PH (Head of Sales), DI (Customer Success), LO (Head of Marketing), JG (Product), MS (Operations), MS (Creative), ST (Content and media). The first hires, separate from the bench: employed Simma Local Partners #1 from month 3 and #2 from month 4; engineers start in month 8, with the pre-seed. The founder is the technical founder; there is no co-founder and no advisors yet. The franchise conversations are with an industry contact, not an adviser.
Larger customers
Vision. Within five years, the ambition is dedicated teams for mid-market and enterprise customers: sales, product, implementation and customer success, with the security and governance those buyers expect. It is an ambition, gated on the earlier milestones; those teams' costs are not in the current model until they are budgeted.
From Series A
Vision. The company organises around Build · Partner · Scale: Build makes the product and the playbook, Partner takes it into communities through employed Local Partners, Scale grows the network and the brand.
Who first
Plan. One-person service businesses: Pilates instructors, Dance teachers, Hairdressers, Driving instructors, Dog groomers, Florists. Next: franchises and multi-location groups, planned from year two, once one network can be shown working end to end. Talks with one franchise operator are paused for now, until the core product is complete.
Customer ramp by channel
Plan (outputs of the v10 model (2026-10-11): self-serve sign-ups from the marketing budget, 20% starting paid and 3% of Free sign-ups converting within six months; Local Partners; central sales for larger accounts; Enterprise contracts from year 3. Franchise excluded from this iteration).
- Year 1 — 271 paying customers at year end; new in the year: 183 self-serve straight to a paid plan, 9 Free accounts upgrading, 97 won by Local Partners
- Year 2 — 1,670 paying customers at year end; new in the year: 991 self-serve straight to a paid plan, 92 Free accounts upgrading, 573 won by Local Partners, 2 larger accounts by central sales
- Year 3 — 7,110 paying customers at year end; new in the year: 4,996 self-serve straight to a paid plan, 474 Free accounts upgrading, 1,280 won by Local Partners, 30 larger accounts by central sales, 6 Enterprise contracts
- Year 4 — 22,408 paying customers at year end; new in the year: 16,195 self-serve straight to a paid plan, 1,572 Free accounts upgrading, 2,150 won by Local Partners, 62 larger accounts by central sales, 12 Enterprise contracts
- Year 5 — 53,513 paying customers at year end; new in the year: 36,006 self-serve straight to a paid plan, 3,663 Free accounts upgrading, 3,838 won by Local Partners, 121 larger accounts by central sales, 24 Enterprise contracts
UK-only in years 1–2, then 35% outside the UK by year 5.
Channels
Direct and self-serve (the product sells itself to a one-person business), product-led referral (every Simma site, portal and receipt carries it), alliances (associations, accountants and industry bodies that refer businesses), and franchise and multi-location groups. Product-led is the engine: most businesses find Simma, start free, set themselves up and run it with the AI team. On top of it, employed Simma Local Partners do local marketing, sales and support: they win the businesses that want a person, onboard them and look after them (Build · Partner · Scale). A Partner's portfolio is the customers they win; self-serve sign-ups in their area stay self-serve, and the welcome email introduces their Local Partner. The first two start in months 3 and 4, in different parts of the UK. A Partner's area holds about 7,000 small businesses; portfolios lose about 35% of customers a year at the starting plan mix (the model sets churn by plan, the same for self-serve), and the Partner earns 3% of the revenue each customer generated in the previous 12 months, paid at every customer anniversary; it exists to reward keeping customers.
Franchise networks, in order
Plan. Show one network working end to end (head office to location to customer to payment to refund) → reopen the paused talks → a pilot only if both sides agree, measuring what head office gains, how many locations take it up and what it costs to support → then more networks. Head office buys; each location still has to adopt, so a signed network is not the same as live locations. Franchise sales are led centrally, not by Local Partners.
Later: larger customers
Vision. Dedicated mid-market and enterprise teams within five years, once the product and the earlier customers justify them. Their costs are not in the current model.
Plan — the founder's view of what could go wrong, and what's being done.
- No customers yet. The month-6 target is 30 paying customers, won by the founder and the first Local Partners.
- Solo founder today. Two employed Local Partners in the first six months and two engineers from month 8, with the pre-seed; a bench of eight senior people who have asked to join (nothing in writing).
- Founder employment and IP. The founder is employed full-time; IP is assigned to Simma Labs Limited before first close, and the employment position confirmed with a lawyer.
- Code ahead of production. 2,120 commits await release; a weekly release cadence resumes with funding.
- Payments revenue depends on fees being applied everywhere. Being fixed now (status page).
- "AI operating systems" from larger companies. They connect to rented tools; Simma owns the record and the public faces.
- Payments and data regulation. Stripe Connect carries card data (Simma never touches it); KYC is audited; GDPR footers by country.
- Franchise networks may be slower than they look. Head office can buy, but each location has to adopt Simma and move its payments; royalties are only recorded on payments taken in Simma. Specialist franchise software, and Square's franchise product, already serve parts of this. Talks with one operator are paused, and may not resume.
- The AI team acting beyond what's approved. Everything the AI team starts goes through a person's yes, and every step is recorded against who approved it; the named teammates are still being built (status page).
- Larger customers need more than breadth. Security, governance, integrations and implementation teams; that is a five-year ambition, not something the platform can claim today.
- SEIS and IP. Relief depends on the company and each investor meeting the conditions, and on HMRC accepting the compliance statement; advance assurance is applied for before first close and is not yet granted. IP is assigned before first close.
Why £250k? It funds the proof, not the scale: six months to show that employed Local Partners can win, launch and keep customers, and about £92.1k kept back so the pre-seed is raised on that evidence rather than against a deadline. It is also the whole SEIS company limit, so the whole round can be SEIS shares. Whether each investor gets the relief depends on the conditions being met, for the company and for them, and on HMRC accepting the company's compliance statement.
How much of the round pays the founder? A market salary for a Head of Product, set out line by line in the private model. No individual salaries are published.
Why does revenue jump so much from year 1 to year 2? Year 1 is about proving the model: 30 paying customers by month six, and 271 by the end of the year in the plan; the pre-seed (raised from month 5) funds the channels. Customers go from 271 to 1,670, Simma Pay take-up builds through year one, and the yearly contracts and Free conversions start to count.
How do Simma Pay's rates compare with Square or SumUp? They're higher on Free by design. On Free, the whole operating system costs nothing: website, bookings, customers, invoicing and payments. Simma earns when the business gets paid (and from any AI credits it chooses to buy): on Free the total per payment is 3.5% + 40p (the card's actual cost plus Simma's 2% + 20p), and on Free, Simma's fees are at least £2 in any month Simma Pay pays out. At higher volumes a separately bought stack costs less in fees alone, and Simma's case is what the fee buys: one record, the AI team and nothing to join up. Paid plans add people, customers and AI capacity, and lower Simma's fee, down to 0.4% + 5p (Scale) (1.9% + 25p in total). Total per payment, online, with a standard UK consumer card: the card's actual cost (1.5% + 20p for this card, passed on at what Stripe charges) plus Simma's fee. Premium and international cards cost more; in person, the card cost differs (Stripe: 1.4% + 10p for UK cards). Whether Simma's fee carries VAT is pending accountant review: if it does, it is included on Free, Lite and Starter and added on Growth and up. The business sees one fee per payment, not itemised. Draft prices; they will be tested with the first studios.
How will 53,513 customers be won? They are an output of the model, not a target: self-serve sign-ups from the marketing budget (one in five starting on a paid plan, and 3% of Free sign-ups upgrading within six months), employed Local Partners winning and keeping customers locally, central sales for larger accounts, and Enterprise contracts from year three — see Go to market. Franchise networks are planned from year two but are not in the model.
Why a priced round and not a SeedFAST or a SAFE? An advance subscription agreement would convert at its six-month long-stop before the next round anyway, and investors only get SEIS relief once shares are issued. A priced round issues them at once. A US SAFE would not qualify for SEIS at all.
What's live today? The status page answers this, dated and line by line. It is re-measured before each conversation.
When do the eight people join? Not in the first round. They join as the next rounds fund them; the first round hires Local Partners.
Is the founder full-time? From first close. Today the founder leads product at a FTSE 100 company by day; the raise is what makes Simma full-time.
Why aren't the design partners paying yet? Because Simma hasn't launched. Simma has not opened to customers or a waitlist: businesses will not run their bookings and takings on it while the founder works full-time elsewhere (an all-day workshop or a three-day trip abroad would leave them unsupported). The raise makes the founder full-time. The dance-school partner pays from launch, and two more businesses have said they will start with it.
Can the round really hire people, or is it one salary? It hires. Two employed Local Partners start in months 3 and 4 alongside the founder, and two engineers with the pre-seed, and people are 64.5% of the six-month spend. A Local Partner costs about £45k a year fully loaded at 20 hours a week.
Is Simma an AI product? No. Simma is the operating system a small business runs on: its website, bookings, customers, email and payments on one record. That works today without any AI (Status: the platform is live; the named AI teammates are next), and its revenue comes from subscriptions and Simma Pay, not from selling AI. The AI team works on top of that record, which is why it can do real work with the owner's approval. What a competitor can't copy in six months is the record the business runs on, the workflow inside it, and the Local Partners who put it there.
Is there demand? For the tools, yes: small businesses already pay for a website, bookings, email, payments and accounts, usually as separate products, and AI tools are part of how most of them work. For Simma itself the evidence is early (two design partners, two more who've said they'll start at launch, franchise talks paused until the core product is complete, no paying customers yet). That is what the Angel round is for: it is spent on customers, not engineering, to reach 30 paying customers by month six and learn what each costs to win.
Why £3m? It's the going rate for a first round, not a stretch. Over the past year, companies raising their first round on SeedLegals were valued at £2.89m on average (SeedLegals Termometer, seedlegals.com/termometer). UK companies at the earliest stage, which the British Business Bank calls the seed stage (a company stage, not a round name: the typical round there was £0.6m), had a median pre-money valuation of £3.2m in 2025 (Small Business Equity Tracker 2026, pp26 and 32-33). £3m sits at the first-round average and below that median, with no AI premium, and the round is sized to SEIS's £250k limit, so it buys 7.69%. UK pre-seed data is thin; treat any single figure with care.
What if the pre-seed is late? It is planned to close in month 7. If it lands three months late and nothing changes, cash after amounts owed to HMRC goes about £9.7k below zero in month 9. So the plan is to hold the hires it pays for (the engineers, the Head of Local Partners, customer success and Local Partner #3) until it lands, and keep marketing going: cash then stays above about £87.4k (the model's scenarios). Pausing marketing doesn't help in the model, because early customers pay back their cost within weeks. Raising starts in month 5 to leave that room.
What will investors get? A short update every month (the numbers, what changed, what's broken, what's next and one ask), sent with Simma's own email newsletter tool, because Simma is run on Simma; SEIS3 certificates as soon as HMRC approves the company's compliance statement, which can be filed once it has traded for four months; and an annual meeting.
How do investors get a return? Most likely through a sale of the company: the businesses consolidating this market are vertical software and payments companies, and an operating system with its own payments and customers is the kind of asset they buy. Shares can also be sold in later rounds. Neither is promised, and SEIS shares must be held for three years for the relief.
What if something happens to the founder? Today that is a real risk. It stays a real risk until the pre-seed: two engineers join in month 8, one owning infrastructure and releases, so from then three people know the platform. A bench of eight senior people have asked to join (nothing agreed in writing). The architecture decisions and the development history are written down and kept with the code, and Simma's own development runs through Simma's task system, so work can be picked up by someone else.
What happens when a business outgrows part of Simma? Simma starts with the connected work service and commerce businesses need most, and plans raise users and limits as a business grows. Some businesses will still want a specialist tool for one job; as Simma serves larger organisations, working alongside those tools will matter more. That is the direction, not a claim about everything the product supports today.
Is Simma only for small businesses? It starts there: instructors and studios, then other service and commerce businesses. The same platform runs franchise networks, and within five years the ambition is dedicated teams for mid-market and enterprise customers. That is an ambition, not a claim that Simma is ready for large companies today.
What does Simma do for a franchise network? Head office keeps one product catalogue and sets a price band for each product; each location prices inside it. A brand site's location finder leads to each location's own site. Every paid booking or order records the franchisor's royalty, and a cancellation reverses it. Built in the code; not yet shown end to end for one network. See Franchise networks.
Do you have a franchise customer? No. Talks with a multi-brand franchise operator are paused until the core product is complete, and nothing is contracted.
Does Simma collect the royalties? It records them: each paid booking or order notes what the franchisor is owed, so nobody has to work it out at month end. The money is not split when the customer pays, and royalties are the franchisor's money, not Simma's revenue.
How big is the franchise market? The British Franchise Association counts 1,009 UK franchise systems and just under 50,500 units. The IFA projects 845,000 US franchise establishments in 2026. Those are the size of the industry, not Simma's market, and are not added to Simma's market figures.
Does the AI team already do everything shown on the site? No. The record, approvals, permissions and audit trail it runs on are live; the named teammates are the next build. Demos that show where it's going are labelled. The status page says what works today.
Shared privately with investors who are engaging:
- The financial model — the five-year workbook (funding rows being rebuilt for the £250k round) and the six-month budget.
- Cap table — the founder holds 100% today.
- IP readiness — what is being assigned to the company, and how.
- The full evidence ledger — a file, query or probe behind every line of the status page.
- The Build · Partner · Scale operating model — how the company and the Local Partner network grow.
- Architecture notes and decision records.
Ask Paul: paul@simma.ai
