Seven years of other people’s deadlines
I started freelancing about ten years ago, long before Harilight had a name. Small sites
first, then web apps, then games, mostly for people I never met: a museum in one country, a
gym in another, a language school somewhere else. Nineteen countries by the time I stopped
counting, from Malta to New Zealand. Forty-seven of those clients left a review and the
average sits at 4.9.
That stretch taught me one thing more thoroughly than any framework did. When a project
goes badly, the client is rarely upset about the work itself. They are upset about not
knowing. Not knowing what it will cost in the end, not knowing when it will be finished,
not knowing whether the silence this week means the developer is busy or gone.
The thing people are buying is not code. It is the certainty that comes with it.
So I built the studio around that one problem
Harilight is three years old. Everything about how it runs is a reaction to the seven years
before it. A fixed price quoted against a written scope, so there is no meter running while
you think. A staging link you can open at two in the morning without asking anyone, so you
never have to send the “how is it going” message. A written update every Friday
even when the week was boring. And the same person from the first call to the handover,
because a handover between two of my own people is exactly where a project goes quiet.
None of that is clever. It is just the list of things I wished for as a freelancer and
never had the standing to insist on.
The games were an accident that stuck
A client once asked for a small browser game to run alongside a campaign. The game did
better than the campaign. Then somebody who had seen it asked for one too, and it kept
going from there. Since then I have built a whack-a-mole for a gym, a French conjugation
game shaped like a rocket launch, a vase-repair game for a museum, a top-down racer, a
burger stacker, and a handful built for no reason other than to keep someone on a landing
page for more than eight seconds.
They all run in a browser tab. No download, no app store, no review queue. That constraint
is the entire point: somebody taps a link in a story and is playing two seconds later, on a
mid-range phone, on a bad connection, standing up, using one thumb. Most of the engineering
goes into those two seconds. The rest is just making the game fun.
Small, remote, and staying that way
Harilight is run from Ariana in Tunisia and most of the work goes to Europe and the United
States. It is deliberately small. Small is the reason the person who scoped your project is
the person who writes the code, and the person who answers when something breaks at an
awkward hour.
More than a hundred projects later, the checklist has not changed since the first week:
- A twenty minute call before anything is quoted
- A written scope with milestones and a price that does not move
- A staging link from week one, and an update every week
- Speed, accessibility and browser testing before launch, not after
- You own the repository, the source files and the handover notes
- Anything that breaks in the first fourteen days gets fixed for free
Tell me what you are building
Send over the idea, however rough. You will get back a scope, a timeline and a fixed price
to approve or ignore, with nothing owed either way.