The arithmetic
- One campaign at the US minimum: about $750 a month.
- At $4.50–6.50 a click: roughly 115–170 clicks a month.
- Over 90 days: about 350–500 clicks on one buyer question.
If your average first-year deal is $15,000 and one deal in a quarter pays back the test, the numbers can work.
How to keep it small and useful
- One buyer question, the one closest to a decision.
- One country, desktop only.
- One landing page that answers that question completely.
- Customers excluded with a hashed list.
- Company identification on, so every visit has a chance of becoming a name.
Where small companies get more for less
AI visibility often gives a small company more than ads: a clear product facts page, an honest comparison page and inclusion in two or three independent articles can get you named in answers seen by every user. How to get recommended.
When to wait
- If your category isn't eligible for ads yet.
- If you can't follow up on companies within a day.
- If one customer is worth less than the cost of a quarter's test.
A one-question test, end to end
| Item | Example (fictional) |
|---|---|
| Question | "Field service software for a 40-person HVAC company" |
| Campaign | US, desktop, one ad group, clicks |
| Budget | $25 a day |
| Page | A page for HVAC companies of 20–100 people |
| Tracking | Demo request as a lead event; company identification on |
| Success at day 90 | 10 fitting companies, 3 meetings, 1 deal |
Questions
Can we run AI ads ourselves?
Yes. Ads Manager is self-serve. Budget a few hours a week for writing, checking and reporting.
What's cheaper, ads or visibility?
Visibility costs time; ads cost money. Small teams with more time than money should start with visibility.
How do we know if a small test worked?
Look at companies and meetings, not clicks. One fitting company that books a meeting can justify the quarter.
Start here: AI ads for B2B. More articles in AI advertising strategy.
