Industries/ Startups/ A narrow first release
MVP development that ships one workflow you can charge for.
We build one complete workflow with proper sign-up, accounts and billing, and put it in front of real customers early. What they do with it decides what gets built next, not the longest feature list in the room.

- Used by
- Your first customers, the founders and whoever runs support
- Takes the place of
- A full-roadmap build, pilot calls and spreadsheets done by hand
- Connects to
- Card payments, email, product analytics and error alerts
- Built with
- Our web apps & saas service
Where it hurts today
A normal week, before it exists
The roadmap was written in a strategy session and it is long: teams, integrations, a mobile app, an admin area, a reporting suite and three kinds of user. The quote to build all of it would take most of the round, and nothing in it has been tried by anyone who would pay. Every month spent building the whole thing is a month in which the company learns very little about what customers want.
Meanwhile the work is done by hand. A co-founder onboards each pilot customer on a video call, runs the core process in a spreadsheet and emails the result. It proves people care, which matters. It also means the team cannot tell which part customers value, nobody can sign up without a calendar invite, and invoices go out from a template whenever somebody remembers.
The usual shortcut is a demo built to impress rather than to be used. It looks right on the founder's laptop, then falls over on the first real account because sign-in, permissions and payments were left for later. Later is when they cost most to add. Investors notice the gap too: a demo is a promise, while a few customers paying every month is evidence.
What it does
A narrow first release, feature by feature
Narrow in scope, and finished in the places that are painful to retrofit.
- 01
One workflow, end to end
We agree the single job a customer would pay for, from the first screen to the result they take away, and build only that. Everything else goes on a written not-yet list with the reason beside each item, so nothing is lost and nothing is quietly half done.
- 02
Self-serve sign-up and onboarding
A customer can find the product, create an account, verify an email and reach the useful moment without a call with a founder. Sample data and a short first-run checklist get them there, and you can see exactly where people stop on the way.
- 03
Multi-tenant accounts from day one
Organisations, invitations and roles are built in from the start, with each customer's data kept apart from every other customer's. Adding this after launch is one of the most painful rewrites a young product can face, so it is part of the foundations even when you only have one pilot.
- 04
Subscription billing, not a payment link
Trials, a small number of plans, card payments, invoices and receipts, sales tax and VAT handled by the payment provider where it offers that, and retries when a card fails. Customers can upgrade, downgrade and cancel themselves, so nobody edits subscriptions by hand at midnight.
- 05
An admin console for the founders
A simple internal screen showing every account, its plan, its last activity and its payment state. Founders can extend a trial, see what a customer sees and reset access, without asking an engineer to run a query against the live database.
- 06
Product analytics and error alerts
Analytics on the handful of steps that matter, a way for customers to send feedback from inside the product, and error alerts that reach a person with enough detail to fix the problem. The weekly review then starts from what happened rather than from opinions.
A day with it
How the day runs instead
A normal week once the first release is live and a handful of customers are using it.
- Monday, 9amFounder
The founder opens the admin console and sees two trials from the weekend. One finished setup and invited a colleague. The other stopped at the import step. The founder sends the second account a short personal email asking what got in the way.
- TuesdayEngineer
A customer reports that an export is missing a column. The error alert already shows the account, the page and the request. The engineer writes a fix with a test, it goes out through the pipeline that afternoon, and the customer gets a reply.
- WednesdayCo-founder
A monthly card payment fails. The product emails the customer a link to update their card and the payment provider retries on its schedule. The co-founder sees the account flagged in the console and decides no personal follow-up is needed yet.
- ThursdayWhole team
The weekly review. Activation, feedback and support emails sit side by side. Three customers asked for the same export format and nobody used the settings page. The team moves one item up the list and leaves two requested features in the not-yet column.
- FridayEngineer
A small change to the onboarding checklist ships behind a feature flag to new accounts only. If next week's numbers look worse, it is switched off without a release. Nobody stays late for it.
What it connects to
Works with what you already run
A first release talks to as few outside systems as it can get away with. These are the four it usually cannot do without.
Card payments and subscriptions
Plans, trials, invoices and failed-payment retries handled by a payment provider such as Stripe or Paddle, whichever suits your markets, with the product reading subscription state from it rather than keeping its own copy.
Transactional email
Verification, password reset, invitations, receipts and trial reminders sent through an email delivery service, with your domain's sending records set up properly so messages are less likely to land in spam.
Product analytics
Events for the few steps that decide activation, sent to an analytics tool such as PostHog or Mixpanel if you prefer one, and limited to what your privacy notice says you collect.
Error and performance monitoring
Errors and slow pages reported to a monitoring service with the account and request attached, and routed to whoever is on duty, so a broken sign-up can be noticed before a customer has to email about it.
The first version
Start small, then grow it
A first version should be useful on day one and small enough to get right. This is where we would usually draw the line, and what tends to follow once people are using it.
In the first version
- One complete workflow a customer can pay for, from sign-up to result
- Accounts, organisations, invitations and two or three roles
- Subscription billing with trials, plans, receipts and self-serve cancellation
- A founder admin console for accounts, trials and access
- Analytics on key steps, in-product feedback and error alerts
Once it is in use
- Integrations with the tools your customers already use, chosen by request
- A public API once a customer or partner actually needs one
- Single sign-on for larger customers whose IT teams ask for it
- Usage-based pricing and annual plans once pricing has settled
Buyer's guide
Before you commission one
Four decisions that shape a first release more than any choice of framework.
- 01
No-code, off-the-shelf or custom
If your product is mainly a form, a workflow and a payment, a no-code builder or a well-configured SaaS tool may reach paying customers sooner, and we will say so. Custom code earns its place when the workflow itself is the advantage, when you need proper multi-tenant accounts and permissions, or when a tool's limits are already shaping what you can sell.
- 02
What to prepare first
A written description of the first customer and the job they would hire the product for, the steps they follow today, and the spreadsheets, emails or documents they use. Notes from pilot conversations are valuable. Bring a clickable prototype if you have one, and rough pricing ideas too, because how you charge shapes the data model.
- 03
Who should own it
One founder should own the product decisions and the not-yet list, and be available for a short weekly review. A technical co-founder or adviser, if you have one, should review architecture choices. Whoever talks to customers should own the feedback loop, because their notes are what decide the second release.
- 04
How you will know it is working
Watch whether new accounts reach the useful moment without help, whether they come back the following week, whether trials turn into paid plans and whether customers ask for the next thing rather than for the basics to be fixed. Track these from the first week, because producing that evidence is the whole point of a first release.
Terms worth knowing
The words you will meet
- Multi-tenancy
- One application serving many customer organisations, with each organisation's data kept separate. Done properly from the start, it lets you add customers without new servers or copies of the code. Done badly, one customer can see another's records, which is the kind of mistake that ends early enterprise conversations.
- Merchant of record
- A payment company that legally resells your product, so it collects and pays sales tax, VAT and GST in each country and issues the invoices. Paddle works this way. With a standard payment processor account, your company is the seller, and the tax registrations and filings are yours to manage.
- Passkeys
- A sign-in method that replaces the password with a key held on the customer's phone or laptop and confirmed with a fingerprint, face or PIN. Passkeys are built on the FIDO standards and resist phishing, because each one only works on the site it was created for.
Handled with care
The details that matter here
The parts of a first release that are invisible in a demo and decide whether the second year is calm.
Scope written down, both ways
We write a short scope listing what is in the first release and what is deliberately out, with the reason for each. When a new idea arrives mid-build, we show what it would displace, so the trade is a decision the founders make rather than a slow drift in the plan.
Customer data kept apart
Each customer's records are separated in the data layer, not just hidden in the interface, and tested so one account cannot read another's. We design around UK GDPR and EU GDPR with whoever owns data protection for you, keep a record of what is stored, and make deleting an account actually delete it.
Sign-in that is safe and usable
Email verification, multi-factor sign-in and passkeys where your customers' devices support them. WCAG 2.2 asks that signing in does not rely on a memory or puzzle test without an alternative, so we let password managers fill fields and avoid fiddly challenges that lock real customers out.
Code and accounts in your name
The repository, hosting, domain, payment provider and email service are set up under your company from the first commit, and we work inside them as invited members. A conventional, documented stack means your first in-house engineer can take over without starting again.
Guides
Worth reading first
Questions
What buyers ask first
What should an MVP include?
One complete workflow a real customer would pay for, plus the foundations that are painful to add later: sign-up, accounts and roles, billing, and enough analytics to see what people do. Everything else waits. We start with who exactly the customer is and what they do today instead, then cut whatever does not serve that single result, and write the cuts down so they are decisions rather than omissions.
What if our pilot customers all want different things?
They usually do, and it is useful information. We look for the overlap: the step every pilot repeats, even when each describes it differently. That becomes the first release. Requests only one customer makes go on the not-yet list with that customer's name beside them. If the overlap is too thin to build around, we will say so, because that is a positioning question before it is a software one.
Will we have to rebuild our MVP when we scale?
We design so you should not have to. The foundations, meaning accounts, permissions, billing and deployment, are built to last, and everything above them is expected to change as you learn. We keep one well-organised system rather than many moving parts, and load-test to find the first real bottleneck. What we will not do is build for a scale you have no evidence of yet.
Should we use Stripe or a merchant of record for SaaS billing?
It depends mainly on where you sell and how much tax administration you want to carry. A merchant of record such as Paddle takes on sales tax and VAT as the legal seller, which suits small teams selling into many countries, at the cost of some control over checkout and receipts. A payment processor account leaves you as the seller. We set out the trade-off for your markets and build on whichever you choose.
Can our own developer work alongside you?
Yes, and we would encourage it. Your developer can review our pull requests, take tickets from the same board and join the weekly review, so the knowledge stays inside your company when we step back. If you are planning to hire, we can help write the role brief, take part in technical interviews and hand over the codebase with a walkthrough.
Not sure what to build first?
Bring the roadmap, the spreadsheet or the pilot customer emails. We will tell you what the first release should contain and what can wait, and we reply within 48 hours.
More for startups
What else we build for startups
Pitch-ready brand identity
A mark, colours, type and a pitch-deck template sized for where you are now, with room to extend later. Investors, first hires and early customers judge credibility on thin evidence, so give them a company that looks as considered as the product.
Finishing a stalled build
An honest assessment of the half-finished product you were left with, then the parts that matter stabilised, tested and completed. Rebuilding by reflex is expensive, and often the existing code needs finishing rather than replacing. We will tell you which.
AI inside your product
Smart search, recommendations, summaries or document reading built into your own product, each scoped to do one job well, with the cost per customer measured before the feature goes anywhere near your pricing page.
Metrics your board trusts
Signups, activation, retention, revenue and churn pulled straight from your product and billing system into one live view, so investor updates and board packs stop costing a founder a weekend of spreadsheet work.
Phone apps and store launch
When your product needs an app, we build for iPhone and Android together and handle the store side in parallel: developer accounts, testing tracks, privacy declarations and reviewer notes, so submission never holds up a launch date you have announced.
Everything we build for startups
The same idea, in other sectors


