My First SaaS Business

Selling TeamTalk, from nothing to twelve months in

You have a working product, three warm introductions and years of selling behind you, but not software subscriptions. This guide covers the part that is different, and turns the introductions into a company with paying customers. Read it in order once; then live in Chapters 5, 6 and 10.

The whole thing in five lines
  1. What you sell: notices to every worker on site in their own language, with proof of who read them. £180 per site per month, or a company plan from £550.
  2. Who buys it: the person who owns toolbox talks or food-safety notices at companies with big multilingual site teams. Not IT, not marketing.
  3. How it goes: 20-minute discovery call → 30-minute demo on a phone → £2,500 60-day pilot on up to three sites → annual plan → the division → the group.
  4. What it costs you to serve: about £161 a month of platform plus 1.5p–8p per workerS1S6. Gross margin above 94%.
  5. Year one, base case: four paying tenants and one expansion, £43k ARR, about £34k of cash received, against £50k–£65k needed to live on. Conservative is two tenants and £12k. The honest plan is a slow first year that produces the references for a fast second.

How this guide was made, and what it is not

Who wrote it. Simon Minton, with Claude (Anthropic's AI) doing the research legwork and drafting, on 27–28 August 2026. Simon set the direction and reviewed it; the AI fetched pricing pages, Companies House records and company sites, and wrote the first draft of every chapter. Treat it as a well-researched briefing from a colleague, not as gospel.

What is solid. The description of the product (read from the code itself). Competitor prices and features (fetched from each vendor's own pages, dated). Unit costs for SMS, video and translation (vendor price lists, dated). The 100 target companies (each checked against Companies House or the company's own site, with a source link per row). Certification and ICO fees (published price lists).

What is judgement. The price list is a recommendation, argued in Chapter 3, not a market fact. The revenue scenarios are arithmetic on assumptions, clearly labelled. Lead times and reply rates are estimates from general UK B2B practice, not measured for TeamTalk, and the guide says so wherever they appear.

What we could not verify. Live web search ran out part-way through the target research, so 49 headcount cells in the 100-target list are marked UNVERIFIED and rest on private accounts. Multilingual workforce is published evidence for food manufacturing and care (visa sponsor registers); for construction, catering and facilities it is inferred and marked Likely. Twilio's UK channel fee for verification codes is not itemised on its page; the all-in login cost is derived. Sales-cycle lengths and funnel conversion rates are reasoned estimates from published B2B benchmarks, not TeamTalk data, and are labelled as such wherever they appear.

What it is not. Legal advice (the DPA and CDM sections tell you what to ask a solicitor and your HSEQ contacts), a forecast (the scenarios are targets you can aim at), or a substitute for the first three conversations, which will correct it.

What is different about selling SaaS

You already know how to open a conversation, read a room and close. Keep all of that. These seven things are new.

You are used toIn SaaSSo
Winning the orderWinning the renewal. Revenue arrives monthly for years; a customer who leaves after 11 months lost you money.Onboarding and the first 30 days are part of the sale. You go to site.
A price per job, negotiatedA published price list. Buyers check it before they speak to you.Publish it, quote it verbatim, discount only for references.
Proposals and pitchesDemos. The product does the persuading on a phone in their hand.Discovery first, then a 30-minute demo of only what they mentioned.
One decision-makerA user, a budget holder, procurement and a DPO, in that order.Sell to the user, get the budget holder's yes, then run procurement in parallel, never after.
Delivering what was soldLand small, expand. One site becomes a division.Paid pilot, agreed success metric, day-60 review with the expansion ask.
Margins on cost of goodsGross margins above 85%; the cost is your time and the acquisition.Price on value to them, not on cost to you. Never free.
Instinct about how it is goingTen numbers every Friday: activation, confirmed reach, pipeline stages, churn.Chapter 10. The numbers tell you two weeks before your gut does.

The chapters

1 · What you are selling

The one sentence, what the product does in the buyer's words, what it is not, and the stack answer to "where is our data".

2 · Market and competitors

Twenty vendors priced and compared. Blink is the benchmark; WhatsApp is the incumbent; nobody does translated captions or UK residency.

3 · Pricing

Per site or per band, never per head: Site £180, Company £550–£2,600 by workers enrolled, Enterprise from £3,600, a £2,500 refundable pilot. Why, and how to quote it.

4 · Costs and margins

£161 a month fixed, pennies per worker, one product decision (login policy) that decides the cost line, and a calculator.

5 · The 12-month plan

Twenty-two numbered steps, sales cycles by buyer type, the verified selling calendar, three revenue scenarios and the cash they need.

6 · Your three warm leads

Barratt London, The Menu Partners, Fernham Homes: the play for each, what to ask for, in what order.

7 · 100 targets

Verified UK companies that look like your warm leads, tiered, with buyer titles and warm paths. Filter, sort, download.

8 · Sales kit

Discovery script, demo flow, objections with answers, cold and warm emails, follow-ups, the one-pager and the proposal.

9 · Procurement and compliance

Cyber Essentials, DPA, insurance, the supplier questionnaire, the CDM/HSE angle. Lead times and costs.

10 · Metrics and the weekly review

Ten numbers and a Friday hour. Targets for pipeline, product health and money.

11 · Templates

Order form, pilot wording, DPA contents, security summary, onboarding checklist, reference ask, renewal notice.

12 · Sources and assumptions

Every figure in the guide, with its origin, date and whether it is published, derived or an estimate.

How to use this

The five things that decide whether this works

  1. You send the three intro requests this week. Everything else waits on them.
  2. You never run a free pilot. A paid pilot has an owner.
  3. You go to site for onboarding. Activation is the number that predicts renewal, and it is won in person.
  4. You protect two mornings a week for outreach from month 3, whatever else is on fire.
  5. You do the Friday hour. If you skip it twice, you have stopped selling and not noticed.