Google Ads on a Small Budget Guide

On a small budget, your real constraint is not money — it is conversions. Google’s automated bidding needs conversion data to work, and a campaign producing fifteen conversions a month cannot feed it. Almost every small-budget failure traces back to that one fact, and almost every fix follows from accepting it.

Do the sum before you spend anything

Before choosing a keyword, work out whether the channel can work at all.

Break-even cost per acquisition = average order value × gross margin

Sell at $60 with 45% margin and you can pay up to $27 for an order before losing money. If your keyword costs $3 a click and your site converts at 2%, each order costs $150. The channel cannot work at those numbers, and no amount of optimisation closes a gap that size. Either the conversion rate has to move a long way, or the order value does, or this is the wrong channel for now.

Our ad budget planner does this in reverse — you give it the revenue you want and it tells you the budget required, and whether the target is reachable at your margin. Break-even ROAS covers the same ground in more detail.

Narrow, then narrow again

A large budget can afford to discover which searches work. A small one cannot, so every dollar has to land on demand you already understand.

• One or two campaigns. Not eight. Splitting a small budget across many campaigns starves every one of them of data.
• Exact and phrase match first. Google’s documentation on keyword match types describes broad match as capturing everything narrower plus more — and on a small budget, “plus more” is where your money disappears. Broad match works well with Smart Bidding and a healthy conversion volume, which is precisely what you do not have yet.
• Buying intent only. “Best running shoes” is research. “Buy Brooks Ghost 16 size 10” is a purchase. Leave research traffic to SEO, where it costs nothing per click.
• Tight geography and hours. If you only ship to three countries, or only answer the phone on weekdays, say so.
• A negative keyword list from day one. “Free”, “jobs”, “cheap”, “DIY”, your competitors’ names if you do not want to bid on them. Review the search terms report weekly for the first month.

Bidding: manual first, automated when you have earned it

Google’s Smart Bidding optimises bids at auction time using signals no human can evaluate. It is genuinely better than manual bidding — once it has enough conversions to learn from.

A workable sequence:

1. Under ~30 conversions a month: manual CPC or maximise clicks with a bid cap. You are buying data, so spend it where you can see what it bought.
2. 30 to 50 conversions a month: maximise conversions. Let the system learn while you still watch it closely.
3. Above 50 a month, consistently: Target ROAS or Target CPA, set from your real break-even rather than an aspiration.

These thresholds are rules of thumb, not settings Google publishes. The principle behind them is not in dispute: automated bidding needs a conversion signal, and a weak signal produces confident nonsense.

Track conversions properly, or none of this works

This is the step people skip and the step everything else depends on. Send real revenue values, not a flat number, through conversion values. Count purchases and qualified enquiries as conversions; do not count page views, time on site, or a click on your phone number that nobody answered. One badly configured conversion action will teach the bidding system to buy the wrong traffic, enthusiastically and at speed.

Four things that waste small budgets

1. The Display Network left on by default. Search campaigns opt into display partners unless you opt out. On a small budget this quietly consumes a third of your spend on placements you would never choose.
2. Bidding on your own brand before you need to. If nobody is bidding against your name, you are paying for clicks you would have received free. Check first.
3. Sending every ad to the homepage. The homepage answers no specific question. A page about the advertised thing converts several times better, which is the cheapest improvement available — see conversion rate optimisation.
4. Changing everything every three days. Automated bidding re-enters a learning period after significant changes. Constant edits mean it never finishes learning and you never see a clean result.

A realistic first 90 days

• Weeks 1–2. Conversion tracking verified end to end. One campaign, exact and phrase match, 10–20 keywords with clear buying intent. Landing page matching the ad. Negatives loaded.
• Weeks 3–6. Search terms report weekly. Add negatives. Pause keywords with spend and no conversions. Do not add campaigns yet.
• Weeks 7–12. Enough data to see cost per acquisition against your break-even. Scale what clears it, cut what does not. Consider automated bidding if conversion volume supports it.

Anyone promising profitability in week one is describing luck, not a process.

Frequently asked questions

What is the minimum sensible budget?

Enough to buy roughly 30 conversions a month, which is cost per acquisition × 30. At a $25 CPA that is about $750 a month. Below that you can still advertise, but treat it as buying information rather than buying growth.

Google Ads or Meta Ads on a small budget?

Google captures demand that already exists; Meta creates it. If people are already searching for what you sell, Google is usually the cheaper first channel. If nobody knows the product exists, Meta with strong creative may be the only option — and then the creative carries the campaign, which is why the first three seconds matters so much there.

Should I use Performance Max?

It needs conversion volume and good creative assets to work, and it reports less about where your money went than a search campaign does. On a small budget, start where you can see what happened.

How much should an agency charge to manage this?

If the fee is a large fraction of the media spend, the maths rarely works: a $500 fee on a $1,000 budget means the campaign must earn back 50% more before anyone profits. Our guide to choosing an agency goes through what a retainer should actually include.

Where to go from here

Do the break-even sum first. If the channel cannot pay at your current conversion rate and margin, fix that before spending — it is cheaper. If it can, start narrow and let the data widen it. That is how we run paid advertising, and where the account already exists, the first thing we do is look at where the money is going in a free audit.