How to run ads for a digital product
Paid traffic is a multiplier. Multiplying nothing is still nothing, and it costs about $20 a day to find that out.
The short version
- Ads scale what works. They do not create it, and they expose what does not.
- Do not run ads until something has sold organically. Ever.
- Two numbers decide everything: cost per lead, and what a lead is worth.
- $20 a day for 14 days is the smallest test that tells you anything.
- Send cold traffic to a free thing, not to a checkout.
When are you ready to run ads?
When the product has already sold to organic traffic, at the price you intend to charge, more than a handful of times. Until then an ad account is an expensive way to discover that the offer was the problem, and it will not tell you that clearly, because a failing campaign looks the same whether the creative, the audience or the product is at fault.
This is the part of paid traffic nobody wants to hear, and it is the whole thing. Ads take an existing conversion rate and buy more visitors at it. If your sales page converts cold traffic at 0.2%, ads buy you 0.2% at scale, and you lose money faster than you were losing it before.
| Situation | Ads will |
|---|---|
| Product sells organically, list buys | Work, and scale it |
| Product sells occasionally, no list | Work if you advertise the lead magnet |
| Nothing has ever sold | Cost money and teach you nothing |
| No sales page, just a Gumroad link | Waste every click |
| You have not read your own analytics | Be premature |
Exhaust free traffic first. Not for reasons of virtue. Because organic performance is the data that makes paid traffic survivable. A hook that works in a reel works in an ad, and you got to test it for nothing.
The two numbers
What a lead costs you, and what a lead is worth to you. If the second is larger than the first, spend more. If it is not, the campaign is not the thing to fix. Everything else in an ad account is a detail underneath those two numbers.
Lead value is not guesswork. Take the last hundred people who joined your list and the revenue they produced in thirty days:
| Weak funnel | Working funnel | |
|---|---|---|
| New subscribers | 100 | 100 |
| Buyers in 30 days | 2 | 4 |
| Product price | $29 | $29 |
| Order bump & upsell | None | $46 |
| Revenue | $58 | $162 |
| Value per lead | $0.58 | $1.62 |
Now the decision is arithmetic rather than nerve. At $0.58 a lead you can afford roughly nothing, because leads rarely cost less than that. At $1.62 you can pay a dollar a lead all day and profit on every one. The difference between those two columns is not traffic: it is a bump and an upsell and a sequence that sells, which is the work to do before you open an ad account.
What to actually advertise
The free thing, not the product. Cold traffic converts on a $29 sales page at a fraction of a percent; the same traffic gives an email address at 20–40%. Buy the address, then sell for free, forever.
- Ad → lead magnet → sequence → product. The standard route, and the one that works at small budgets.
- Ad → low-ticket product → upsell. Viable once you know your numbers cold. A $9 entry product that pays for the ad makes everything afterwards free.
- Ad → sales page directly. Only for retargeting people who already visited. Never for cold traffic on a small budget.
- Ad → your profile. Not a strategy. You are paying for followers, which is the least valuable thing in the funnel.
Retargeting is the exception to everything here. People who visited your sales page and did not buy are the cheapest, warmest audience you will ever pay for. If you run one campaign, run that one, and it needs almost no budget to be worth it.
Creative: what works without a face
Show the thing. The highest-performing ads for digital products are usually a screen recording of the product being used, or a plain image of what lands in the buyer's folder. Nobody needs to see you; they need to see what they get.
| Creative | Why it works | Effort |
|---|---|---|
| Screen recording of the product | Removes all doubt about what it is | 15 min |
| Text-on-screen video | Reads on mute, cheap to test in bulk | 15 min |
| Your best organic post, boosted | Already proven on a real audience | None |
| Flat image of the deliverables | Makes a digital thing feel like an object | 20 min |
| Voiceover walkthrough | Builds trust without identifying anyone | 30 min |
- Test five creatives, not five audiences. On modern ad platforms the creative is the targeting: the system finds the audience from who responds.
- Change one thing between variants. Five different hooks over the same footage teaches you which hook works. Five entirely different videos teach you nothing.
- Start with your best organic post. It already cleared the only test that matters, and it costs nothing to reuse.
- Refresh every two to three weeks. Small audiences fatigue fast, and rising cost per lead on a flat campaign is almost always creative fatigue.
The smallest test worth running
Twenty dollars a day, one campaign, one destination, five creatives, fourteen days untouched. That is roughly $280 to learn whether paid traffic works for your offer, and the discipline is not spending it, it is leaving it alone while it spends.
- Do not turn it off on day three. The first few days are the platform's learning phase and the numbers are meaningless.
- Do not edit a running campaign. Most changes restart the learning and you have paid for it twice.
- Do not run six campaigns on $20. Split six ways, none of them gathers enough data to optimise anything.
- Do check the cost per lead daily and write it down. The trend over a fortnight is the result; any single day is weather.
At the end of the fortnight there are three outcomes. Leads cost less than they are worth, in which case increase the budget by 20% every few days and keep going. Leads cost slightly more than they are worth, in which case fix the funnel (the offer, the bump, the sequence) rather than the ads. Or leads cost far more than they are worth, in which case go back to organic, because paid traffic has just told you something true about the offer for $280.
That third outcome is the most valuable one, and it is why the size of the test matters more than its success. Ads are not a growth channel for most small digital products; they are a very fast, very honest test of whether the rest of the machine works. Build the machine first, on free traffic, with a launch that proves people will pay, and the ad account becomes a decision about scale rather than a hope.
Frequently asked questions
How much should I start with?
Twenty dollars a day for at least fourteen days, on one campaign. Smaller than that and the platform never gathers enough data to optimise; shorter than that and you are reading noise. If $280 is money you cannot lose, you are not ready to run ads yet.
Do I need to show my face in the ad creative?
No. The best-performing ads for digital products are usually screen recordings of the product, text-on-screen videos, or a static image of what the buyer receives. A face helps in some categories and is never the deciding variable.
Should I send ad traffic to a sales page or a lead magnet?
A lead magnet, almost always, at small budgets. Cold traffic converts poorly on a direct sale, and an email address gives you unlimited free attempts afterwards instead of one paid one.
What is a good cost per lead?
For a low-ticket digital product, roughly $0.50 to $3 depending on the country and niche. What matters more is the ratio: if a lead costs $2 and your list buys at 3% on a $29 product, each lead is worth about $0.87 and you are losing money, which is a fixable offer problem, not an ad problem.
Which platform should I start on?
Whichever one your organic content already works on. The creative transfers, the audience is proven, and you are testing one new variable instead of three.
The system behind this, written down
Everything above is the map. The Income Loop is the work inside it: modules 0–6 from the problem you solve to the offer that pays for it, plus ten traffic paths: the deeper post banks, the content sales systems and the full software build sequence, in one place.
Get The Income LoopNot ready to pay for anything? The Basic Income Loop is free and includes a complete seven-day starter for Threads, Instagram or software, enough to find out which one suits you before spending anything.