Job tracking
Create a job in seconds — client, description, price, date. Status moves through scheduled, in progress, done and paid with one tap.
Micro-SaaS · Idea → Paying users
KaziDesk — a small product that pays for itself.
A lightweight job and client desk for small service businesses — plumbers, electricians, cleaners, fundis. Built from scratch in eight weeks, shipped to real teams, and now paying its own way. This is what micro-SaaS looks like when you actually finish it.
8 weeksIdea to first paying customer
KES 84kMonthly recurring revenue
23Paying service teams
Today's jobs
+ New jobMRR
KES 84k
▲ 18%
Active jobs
27
▲ 6
Collected
61k this wk
▲ 12%
Open jobs
4 of 27
KES 4,500
In progress
KES 12,000
Scheduled
KES 3,200
Paid
KES 18,500
Awaiting approval
The opportunity
Small service teams run on notebooks, WhatsApp and memory.
Across Nairobi and beyond, thousands of small service businesses — plumbers, electricians, cleaners, fundis — take jobs over the phone, track them in a notebook, chase payment over WhatsApp, and hope nothing slips. Existing tools are either too expensive, too complex, or built for markets that don't match how they actually work.
What they had
- Jobs written in a notebook, easily lost or double-booked
- Invoices made by hand, payment chased verbally
- No idea which jobs actually made money
What KaziDesk does
- Every job logged, scheduled and assigned in one place
- Invoices sent over WhatsApp, paid via M-Pesa in one tap
- A weekly report showing profit per job, per client
The build
Eight weeks from notebook sketch to first invoice.
A real product build isn't a code sprint — it's interviews, an MVP you can demo, a payment rail that works, and the discipline to ship before it's pretty. Here's how it went.
WEEK 1
Problem interviews
Twelve conversations with small service teams across Nairobi. Asked the same three questions every time: how do you take jobs, how do you track them, how do you get paid? The same pain surfaced repeatedly — payment chasing.
DiscoveryWEEK 2–3
MVP build
Jobs, clients, and a simple invoice generator — nothing more. Deliberately excluded scheduling, teams, reporting and anything that would slow the first release. The goal was one usable workflow, not a feature list.
BuildWEEK 4
First three pilot teams
Onboarded three service businesses from the interview group. Sat with each one for a week, watched them use it, and rewrote the invoicing screen twice based on what confused them.
PilotWEEK 5–6
M-Pesa + WhatsApp integration
Added the two things the pilots wouldn't use the product without: STK Push payments on invoices, and WhatsApp delivery so clients receive the invoice where they already are. Reconciliation built in from day one.
IntegrationsWEEK 7
Pricing & billing
Added a free tier, a paid Pro tier, and self-serve subscription billing. The free tier does real work so teams can evaluate without a sales call — most upgrade within the first month once job volume grows.
MonetisationWEEK 8
First paying customer
A pilot team upgraded to Pro without being asked. That was the signal to stop building and start growing. From this point on, every feature gets justified by whether it moves retention or upgrades.
First revenueWhat's in it
Small enough to learn in an afternoon.
Every screen was stripped back until a fundi with a phone could use it without a manual. Deliberate simplicity is the product.
Client records
Every job ties back to a client, so you can see history, repeat work and outstanding balances without digging through a notebook.
One-tap invoicing
Generate a branded invoice from any job, send it over WhatsApp, and get paid through M-Pesa without leaving the app.
M-Pesa payments
Built on the Daraja API with STK Push, automatic reconciliation, and a daily settlement view — so payments trace straight to jobs.
WhatsApp delivery
Invoices and reminders go out through WhatsApp, where clients actually read them — not buried in email inboxes nobody checks.
Weekly profit report
A short Monday summary showing revenue, jobs completed and profit per client — the first real numbers most of these teams have ever had.
Pricing model
Priced for the people it's built for.
A free tier that does real work, a Pro tier priced at roughly the cost of one small job per month, and a Team tier for businesses with staff and multiple sites. No sales calls, no annual lock-in.
Solo
For one person running their own jobs.
Free
forever
- 1 user
- Up to 10 jobs per month
- Basic invoicing
- WhatsApp sending
Pro
For established teams doing steady work.
KES 2,500/mo
billed monthly · cancel anytime
- Up to 5 users
- Unlimited jobs & clients
- M-Pesa payments on invoices
- Weekly profit report
- Priority email support
Team
For businesses with staff, sites and roles.
KES 6,500/mo
billed monthly · cancel anytime
- Unlimited users
- Roles & permissions
- Multi-site support
- API access
- Priority support & onboarding
Traction
Real users, real revenue, real retention.
Not vanity metrics. The number that matters most for a micro-SaaS is whether people keep paying — and whether the product earns more than it costs to run.
KES 84kMonthly recurring revenue
23Paying service teams
4.2%Monthly churn — improving each quarter
8 wksFrom first interview to first payment
“Before this, I didn't actually know which jobs were making me money. Now I see it every Monday morning.”
— J. Mwangi, Pro subscriber · NairobiStack
Built with.
Next.js
TypeScript
Node.js
PostgreSQL
Prisma
Tailwind CSS
M-Pesa Daraja
WhatsApp Business API
Stripe
Vercel
AWS S3
Got an idea for a small product that could pay for itself?
I build micro-SaaS from first interview to first paying customer — narrow scope, real users, real revenue. If you've been sitting on an idea, let's pressure-test it.