About Irvine, 2026
The best channel in the world, and nobody owns it.
Ask anyone who runs a small business where their best customers come from and you'll get the same answer. Someone sent them. It has never been measured, never been paid for, and never been anybody's product — which is exactly why it still works.
Why this is narrower than it could be
Every product that has tried to industrialise word of mouth has done it by widening the pipe. Collect more reviews. Rank more places. Show the recommendation to more people. Each of those decisions makes the thing measurably bigger and quietly worse, because the value was never in the volume — it was in the fact that you knew the person telling you.
So Favor is built the other way round. One place, one sender, one recipient, one first visit. No feed to fill, nothing to rank, no way for a business to buy its way into your mouth. Almost every feature request we get is a request to widen the pipe, and the answer is usually no.
What we did add is the part that was always missing: a way for the person who sent you somewhere to be thanked for it in something other than gratitude. Businesses have always been happy to pay for a new customer. They just had no way to pay the person who actually delivered one.
What we're actually betting on
That a recommendation with money quietly attached to it, sent to one person, from someone who genuinely goes there, is still a real recommendation — and that the market for those is large, slow and durable rather than fast and disposable.
If we're wrong, we'll be wrong in a specific and visible way: the referrals will stop converting and the businesses will stop paying. That's a better failure mode than most, because we'd know early.
Principles
Four rules we've agreed not to break.
These are the ones that cost us something. Anything that doesn't cost you anything isn't a principle, it's a preference.
One
The recommendation comes first
If a design choice makes the money more visible than the words, we don't ship it. This is why the thank-you is never shown to the person receiving a favor, and why it arrives days later.
Two
No placement for sale
There is no amount a business can pay to appear in someone's recommendations. Building an ad product would be the single fastest way to grow revenue and it's permanently off the table.
Three
Small enough to read
Complaints get read by a person, and cities open only when we can keep doing that. Growth that outruns our ability to answer for the product is growth we turn down.
Four
Your spending isn't the product
We read your card to know where you've genuinely been. That's the only use. It is not sold, not shown to businesses, and not turned into a profile for anybody.
Where it works
One city at a time, in that order.
A favor needs a business on the other end, so a city doesn't open until enough places inside it have. These are targets rather than promises — if a city isn't ready, it waits.
IrvineOpen now
Costa MesaNext
Long BeachLater 2026
PasadenaLater 2026
San Diego2027
Los Angeles2027
San Francisco2027
Somewhere you areTell us
Not in Irvine?
Leave your email anyway. Which city we open next is decided mostly by where the waiting list is heaviest and where businesses have asked to be listed — so joining from a city that isn't open yet is the most useful thing you can do about it.
Somebody you know has the good list.
Favor is invite-only while we grow city by city. Leave your email and we'll find who can vouch for you.