The $500/Month Ad Experiment (Without Betting the Business)

You can test Google Ads for 60 days with about $500 a month, one city, and written stop rules — and finish knowing exactly what a lead costs your business, without betting the company on it. The goal of the experiment is not bookings first; it is data first. Bookings are the reward for reading the data correctly.

Why $500 a month is the right experiment size

Roughly $500 a month is about $16–17 a day — not enough to fight for saturated keywords in a big city, and exactly enough to learn in a focused market. To put it in context: with the 2026 national average cost per lead around $67 (WordStream, 13,000+ campaigns), $500 a month buys roughly seven leads in an average market. That is a small but real sample: in 60 days you will have seen more than a dozen inquiries and you will know your conversion pattern. You do not need more data than that to make the go/no-go call — and you need far less money to get it.

Step 1: One geography, one offer

The experiment fails the moment it becomes ambitious. Pick one service area you can actually serve well — a set of zips or one town — and one service people search with intent, like "emergency water heater repair" rather than "plumber." Feed the ad to a specific landing page that answers the question and makes the phone number impossible to miss. Narrow is the entire strategy: a tight experiment gives you a clean read, and clean data beats more data every time.

Step 2: Set the dials so nothing surprises you

• Search only. No display, no Performance Max, no "smart" budgets on run one. You want to see who clicks a search ad, period.
• Call tracking plus a form. Use a dedicated phone number and a form on the page, so every inquiry is counted and every cost per lead is real. Phone and form leads also behave differently — track them separately.
• Hard budget cap. Daily budget of roughly $16–17 with a strict monthly ceiling. When it caps, ads stop; you cannot overspend what you cannot spend.
• Tight keywords, exact and phrase match. A handful of high-intent terms plus negatives ("jobs," "salary," "DIY") keeps the clicks honest.

Step 3: The 60-day calendar

Two warnings about the middle weeks. First, do not judge the experiment before day 14 — small budgets need time to accumulate meaning. Second, the temptation to "mess with it daily" is the biggest failure mode; schedule changes, do not react to them.

Step 4: The stop rules, written before you start

Write these down before the first dollar spends — that is what keeps a $500 experiment from becoming a $5,000 regret.

• Cost per lead vs. your break-even. If your average job value and close rate say a lead is worth $165 (our example from the lead-cost guide), then leads under that figure are viable. The rule: if cost per lead stays above break-even for 60 days, stop.
• Zero conversions with real volume. If you spent $300+ with 50+ clicks and no calls or forms, the problem is usually the offer or the page, not "more time." Pause, fix, and treat the next month as a fresh test.
• One clear winner. If one service, one zip, or one ad group produces most of the leads, that is your proof of concept — the scale-up candidate, not the whole business.

Measure like an owner, not an advertiser

Cost per lead is only half the equation; speed of follow-up is the other half, and it is free. The national data on lead response is brutal — businesses that reply in minutes get the conversation, and slow ones lose it: the 24–48 hour rule exists because most owners are much slower than that. For an experiment this small, every lead counts, so answer every single inquiry same-day, and answer calls within minutes. If you cannot, add the AI follow-up system before you spend another dollar — otherwise you are paying for leads you are losing by hand.

What "success" looks like after 60 days

Success is a number: your true cost per lead in your city for that service, measured over 60 days with clean tracking. Whether that number says scale or stop, you won — because you now know what the channel costs you, and most of your competitors have never bothered to find out. If the number says scale, grow it like a business (slowly, one slice at a time), and consider Local Services Ads as the stabilized alternative. If it says stop, you spent $1,000 to avoid wasting $20,000 — one of the best trades in marketing. And if you already get enough referrals to stay busy, read this before you start any of it.

FAQ

Is $500 a month enough to test Google Ads?

For one city and one service, yes — roughly $16–17 a day, which at average cost per lead means about seven leads a month. That is enough to see a pattern in 60 days.

How long should the ad experiment run?

Sixty days: four weeks to learn, four weeks to refine, and a decision at day 60. Shorter than that is noise; longer without a decision is procrastination.

What if the ads get plenty of clicks but no calls?

That is a landing page or offer problem, not a budget problem. Pause, fix the page and the call setup, and restart the clock with the improvements.

Should I use Local Services Ads instead for the test?

If your trade is eligible, LSA is a good second experiment because it charges per lead. But run one experiment at a time — mixed data from two new channels tells you nothing.

When should I increase the budget?

Only after 60 days of data shows cost per lead under your break-even. Then scale in slices — a modest daily-budget increase on the single best-performing ad group — not across everything.

Want help designing the experiment — keywords, tracking, stop rules — before you risk a dollar? Book an AI consultation and we will build the 60-day plan with you.

Ready to turn AI into measurable growth?

Let's discuss how we can build smarter systems and stronger campaigns for your team.