There is no magic number for a Google Ads budget on a new product, but there is a defensible way to arrive at one. Size the budget from three inputs: the actual cost per click in your keyword set, the conversion rate of the page you send traffic to, and the gross margin you earn per unit. Everything else is guessing. And be clear about what the first month of spend buys: it buys data, not profit. Founders who expect a positive return in week two usually shut the account off before it has learned anything, which is the most expensive outcome available.

Google sells you an auction, not a package

You are not buying a media plan. You are bidding against everyone else who wants the same search, on every single query. That means your cost per click is set by your competitors, not by Google's price list. In quiet niches a click can run about a dollar or two. In categories crowded with well-funded advertisers — insurance-adjacent, medical, B2B software, anything with a high customer lifetime value — clicks routinely cost tens of dollars.

So the first step happens before any budget conversation: pull real cost-per-click estimates for the ten to twenty searches your buyer would actually type. Keyword Planner inside a Google Ads account gives you a range for free. If the number that comes back is $12 a click and your product sells for $39, you have learned something important without spending a cent.

Does anyone search for your product yet?

Search advertising captures existing demand. It works beautifully when people are already typing a description of the problem your product solves. It works poorly when your invention is so new that no one has a word for it — there is no query to bid on, and the traffic you buy on adjacent generic terms converts badly.

The practical test: are you competing for a known category ("cordless tire inflator") or creating one? If you are creating one, bid on the problem phrasing rather than the product phrasing, keep the search budget small, and put the rest into channels that generate demand instead of harvesting it — email, creators, and communities. DIY market research on no budget covers cheap ways to find out which language your buyers use before you pay for it.

The minimum budget that actually teaches you something

A campaign needs volume before its data means anything. A workable floor:

  • Clicks per day. Somewhere around fifteen to thirty clicks a day is a reasonable learning rate. Below that, weeks pass before you can tell a good keyword from a bad one.
  • Daily budget. Multiply that click target by your researched cost per click. At $2 a click you are looking at a daily spend in the low tens of dollars. At $10 a click, the same learning rate costs several hundred a day — which is often the signal to change channels rather than the signal to spend more.
  • Runway. Commit to four to eight continuous weeks up front. A campaign killed after five days produced noise, and the money is simply gone.

Write that total down as a single number before you launch. If you cannot fund it, do not start a paid search campaign yet.

Work backward from unit margin

The professional method inverts the question. Instead of asking what to spend, ask what a customer is worth. Suppose your product earns roughly $60 of gross margin per unit and you are willing to spend a third of that to acquire a sale — about $20 per order. If your product page converts two percent of visitors, each sale takes about fifty clicks, so your maximum sustainable cost per click is around forty cents.

Run that arithmetic honestly and one of three things becomes obvious: the channel works, your price is too low, or your page converts too poorly. All three are useful. Two of them are fixed off Google entirely — see how to price a product for the margin side, and using a landing page to test demand for the conversion side. Doubling a two percent conversion rate has exactly the same effect on your economics as halving your click cost, and it is usually easier.

Where new-product budgets leak

  • Broad matching with no negative keywords. Google will happily show you on "free," "DIY," "repair," and your competitor's model numbers. Those clicks cost full price. Build a negative list in week one and revisit it weekly.
  • Too many campaigns. A small budget split across eight campaigns starves all of them. One tightly themed campaign learns; eight thin ones do not.
  • Expensive traffic hitting a weak page. No clear offer, no shipping cost, no photos of the actual product, no reason to trust an unknown brand. Every dollar of that traffic is wasted.
  • No conversion tracking. If purchases, form fills, and email signups are not tracked back to the keyword, you are optimizing on clicks — and clicks are the one metric that never pays a bill.
  • Sloppy geography and device settings. Defaults often include locations you cannot ship to and search partners you never intended to buy.

Google is one line in a launch budget

Paid search sits alongside photography, video, packaging, samples, and fees. Seeing it in isolation is how founders end up with an ad budget and nothing worth clicking on. The full picture is in what it costs to market a new product.

Two adjacent channels usually deserve funding before search does, because they compound rather than reset each month: building an email list before launch gives you an audience you own outright, and micro-influencer marketing generates the in-context demonstration a text ad cannot. More launch and sales guides are collected on our selling your invention page.

Where Projects House fits

We are an engineering firm, not an ad agency — but we build the things a campaign needs to work: a product that photographs and demonstrates well, renders and CAD-based visuals before units exist, honest specifications you can advertise without overstating, and a design frozen enough that you are not paying for traffic to a product still changing shape.

Get the product ready before you buy the clicks

If you are about to start spending on ads and are not sure the product, the claims, or the visuals are ready to convert that traffic, get in touch through the contact form and we will tell you what to fix first.