House credit: how $100 you did not ask for changes what you test

Every visitor to our board starts with $100 of house credit. Not after signing up — there is no account to make. It is there when you arrive, it is spent only by a bid that actually wins, and if you never bid it costs us nothing and you nothing. We run the board, so read this knowing that. The part worth writing about is not the giveaway. It is what a small pot of not-quite-real money does to the experiments people are willing to run.

Free changes the question people ask

Watch what happens to somebody deciding whether to advertise on a channel they have never used. With a card in hand, the question is "is this worth $30?" — and answering it requires a forecast they have no basis for. They have never run this channel. They do not know its click-through rate. Any number they produce is invented, and most people, sensibly, decline to spend against an invented number.

With credit sitting there, the question becomes "is this worth twenty minutes?" That one they can answer. Twenty minutes is a quantity they have priced a thousand times.

This is not a trick about money being fungible — obviously $30 of credit is worth about $30. It is a fact about decision cost. The dominant expense of a small marketing experiment is not the media spend, it is the deliberation: the meeting, the spreadsheet, the internal argument about whether an unproven channel deserves budget. Credit does not reduce the media spend by much. It removes the deliberation entirely, and the deliberation was the expensive part.

What people actually do with it

Three patterns show up, and only one is what we expected.

They test a message rather than buy traffic. The credit is enough for several contested slots, which means somebody can run one pitch, look at the opens and click-throughs on their brand's own page, rewrite the line, and run it again. That is a copy-testing loop, and it costs nothing out of pocket. Buying the same loop on a search platform means clearing a daily-budget minimum and waiting out a learning phase — see the arithmetic in the cost to advertise a startup for what that runs to.

They defend rather than attack. A brand that holds a slot and gets outbid can take it back for a dollar over the standing bid, and inside the credit that costs nothing. We did not design for this and it is the most common use.

They do nothing at all. Most credit is never spent, which is the honest headline. A grant that has to be used to be worth anything is a discount with extra steps; this one is worth something mainly as information — it tells a visitor what the ceiling of their risk is before they have read a single rule.

Why it is not a free trial

Free trials have a shape: they are a loan against a future subscription, they expire, and the expiry is the point. The credit here has none of that machinery. Nothing renews, nothing auto-bills, there is no card on file to charge when a period ends, and the credit does not expire into a decision you have to remember to make.

That is deliberate, and it costs us the conversion mechanic that makes trials work. What we get instead is that the offer survives scrutiny. A founder who has been burned by a "free" tier that turned into a monthly charge can read our rules in ninety seconds and confirm there is no such mechanism. On a board whose entire pitch is that its numbers are checkable, an offer that needed a footnote would undercut the product more than it sold it.

The same reasoning drove the free first claim, which I argued through in why the first slot is free: an empty slot has no standing bid to beat, so any amount takes it, and zero is an amount.

The honest limits

Credit does not make a small board big. It makes a small board cheap to evaluate, which is a different and lesser thing. If you need volume this week, our slots will not supply it, and you should read where to spend $50 on advertising for channels that will.

It also cannot be cashed out, and the board's own rules are what make that safe rather than mean. Being outbid loses you the slot and the money, with no refund — so credit converted into a held slot is genuinely spent, not parked. There is no arbitrage where somebody claims slots and redeems value back out. That closes the loop that would otherwise make an open-handed grant impossible.

And it is not a growth hack we can claim credit for inventing. Ad platforms have handed out starter credits for years; Google Ads runs one, as do Microsoft and Meta. The difference is small and worth stating plainly: those credits usually require a spend commitment to unlock, and ours does not require an account.

What I would take from this if I ran something else

The transferable lesson is not "give away credit". It is: find the deliberation your buyer has to do before they can evaluate you, and delete it.

For us that was a payment decision made against an unknown click-through rate. For a tool it might be an integration that has to be built before anyone can see a result. For an agency it might be a scoping call. In every case the expensive part is the thinking the customer has to do before they get any information, and it is usually cheaper to eat that cost than to keep asking people to pay it.

The measurable version of the claim: the number of people who will spend twenty minutes on an unproven channel is much larger than the number who will spend thirty dollars, and the gap is not explained by the thirty dollars.