· 6 min read
Should you run ads before the product exists?
When advertising before launch buys real information and when it only buys a list, what a waitlist signup proves, and the honest cost of finding out early.
Advertising before launch is either the cheapest research a founder can buy or an elaborate way to build a list of people who will not remember you in three months, and which one it turns out to be is decided before any money is spent — by what you ask the traffic to do. (Disclosure: published by BidSurvivor, an advertising board, so weigh the recommendation to advertise accordingly. Half of this article argues against spending anything.)
What a pre-launch ad can actually prove
There are exactly three things, and it is worth being precise because founders routinely believe they have learned a fourth.
That the words make sense to a stranger. Somebody outside your head reads a line and understands what is being offered. This is the lowest bar and the most commonly failed one.
That the offer is interesting enough to interrupt. They stop, read, and act. This is genuinely useful information and it is available for very little money.
That the interest survives a small cost. An email address is a cost. A phone number is a larger one. A pre-order is a serious one. Each step up filters harder, and the filter is the point.
What a pre-launch ad cannot prove is that anybody will pay, use, or return. Those depend on the product, and there isn't one. Treating waitlist volume as demand evidence is the standard mistake, and it is the reason so many well-signed-up launches land flat.
The signup-strength ladder
Not all conversions are the same evidence, and the difference is larger than the difference between channels.
| What you ask for | What it proves | Roughly how many say yes |
|---|---|---|
| Nothing (click only) | The line is interesting | Everyone who clicked |
| Email for updates | Mild curiosity, at zero cost | Many |
| Email plus one qualifying question | Curiosity plus a minute of effort | Fewer |
| A named launch-day discount | Intent to look again | Fewer still |
| A deposit or pre-order | Intent to buy | Very few, and worth all the rest combined |
The useful move is to ask for the strongest thing you can without lying. A deposit for a product with no delivery date is not a test, it is a liability. But a qualifying question costs you almost nothing and separates the people who read from the people who clicked.
Ten signups who each answered "what are you using today?" tell you more than four hundred who typed an address into a box. If you take nothing else from this article, take that. It is also the reason the classic advice to get out of the building has outlasted every growth tactic invented since: the answers live with people, not in a conversion rate.
Where to spend a pre-launch budget, in order
A free slot first. Ours: BidSurvivor sells two twelve-hour slots a day, and the first brand into an empty one pays nothing — no account, no card. Every visitor also starts with $100 of house credit, so contested slots cost nothing out of pocket at current prices. Card opens and click-throughs are published per slot on the front page and each brand's page.
For a pre-launch test that is the right shape of instrument: cold strangers, no goodwill, published counts, zero cost. The honest limitation is volume — the board is young and its numbers are small, and small numbers can tell you whether a line works but not whether a channel scales. That is a real constraint and it is stated on the front page rather than buried; the full rules, including the no-refund rule when somebody outbids you, are in how BidSurvivor works.
Communities second. The subreddit or forum where your future users already complain about the problem. Free, high-intent, and the only channel that answers back — a comment saying "we solved this with a spreadsheet and it's fine" is worth more than any click rate.
Paid search last, and only for a category that exists. If people already search for the thing, a small exact-match campaign tells you the demand is real and roughly what it costs to reach. If nobody searches for it — which is the case for genuinely new categories — paid search will tell you nothing except that your keywords have no volume, which you could have learned free from a keyword tool. Google Ads alternatives for small budgets covers what to do instead.
The case against doing any of this
Here is the argument I find hardest to dismiss, and it deserves to be made properly rather than as a token counterpoint.
Pre-launch advertising measures the appeal of a description. Products are not descriptions. A great line in front of a mediocre product produces a waitlist that converts at nothing, and the founder spends three months building against a signal that was never about the product at all. Meanwhile the same fortnight spent talking to twelve people in the trade would have produced specifics — what they use now, what it costs, what breaks — that no click rate contains.
The stronger version: if you can reach twelve real prospects by asking, do that, and skip the ads entirely until you have something to show them. Advertising is for when you cannot reach people one at a time. Most pre-launch founders can, and reach for ads because ads feel like progress and conversations feel like admin.
The case for advertising anyway is narrow but real: it is the only method that tests cold demand. Twelve friendly conversations tell you nothing about whether a stranger with no relationship to you would stop. If your business depends on strangers stopping — and almost every self-serve product does — then at some point you need that number, and getting it for zero on an empty slot is cheaper than getting it after launch.
A plan that costs nothing to start
- Write the line. One sentence naming the buyer and the problem, tested against five people before it goes anywhere near a budget.
- Build one page. It answers the line in its first sentence and asks for exactly one thing.
- Ask a qualifying question on the form. One field, optional, open-ended. This is where the actual information arrives.
- Claim a free slot and read what cold strangers do with it.
- Post once in one community and read what they say back.
- Then decide whether to build, and whether paid traffic is worth anything to you yet.
Steps one to five cost an evening and nothing else. If the answer at step six is no, that is the cheapest no you will ever get.