Google Ads for SaaS puts your page in front of people who are already searching for the problem your product handles. A first campaign can answer one question: does a search you can name bring a buyer you can serve, at a cost you can bear? Most guides assume a marketing team, a monthly budget and a CRM. This one is for a founder with a working product, limited cash and no ad history. Use it to decide whether to run a test at all, then to scope one you can afford to lose.
Decide whether search ads fit your product today
Search ads capture demand that already exists. They cannot make people search for a category they have never heard of. The first check is free: type the phrases your buyers use into Google and read the results. If the results describe your kind of product and other companies advertise there, people are looking and someone pays to reach them. If the results are about something else, your phrase means something different to most searchers.
The second check is your own readiness. An ad sends a stranger to a page. If the page cannot explain the offer, or you cannot tell whether the visitor did anything useful, the money buys clicks you cannot interpret.
- Search evidence: you can name three to five phrases a buyer would type, and the current results match your category.
- Buyer language: the phrases come from interviews, support messages or community threads, not from your own product vocabulary.
- Landing page: one page says who the product is for, what it does and what to do next.
- Countable outcome: you can count one useful action, such as a signup, a demo request or a completed first task.
- Affordable loss: you can lose the whole test budget without touching money the product needs.
Start with one search theme and a list of exclusions
A theme is a small group of phrases with the same intent. "Client approval software for agencies" and "agency client approval tool" belong together. "Project management software" is a broader intent, and on a small budget it will absorb your spend before you learn anything. One theme gives you one result to read.
Google separates the keyword you choose from the search term a person types. How closely the two must match depends on the match type, and the default is the loosest one. For a first test, phrase or exact match keeps the searches close to your theme.
Write the exclusions before launch: job seekers, free templates, tutorials, and every meaning of your phrase that is not your product. In Google Ads these become negative keywords.
Exact match shows ads on searches with the same meaning or intent as the keyword. Phrase match covers searches that include the keyword's meaning. Broad match covers related searches, including ones without the keyword's direct meaning. Broad match is the default.
Google Ads Help: keyword matching optionsNegative keywords exclude search terms from a campaign, so the ad does not show for those searches.
Google Ads Help: negative keywordsThe report shows how your ads performed on the actual searches that triggered them. A search term is what the person typed; a keyword is what you chose.
Google Ads Help: search terms reportSet the spending ceiling before you write the ad
Decide the total you are willing to lose, then work backward. A ceiling of $210 over 14 days is an average daily budget of $15. Google may spend up to twice the daily budget on a single day, so one day can cost $30. A campaign left running for a month can reach 30.4 times the daily budget, which is $456. The daily budget does not end the test. An end date and a written ceiling do.
Do not copy cost-per-click averages from an article, including this one. Costs differ by keyword and change over time. Keyword Planner in Google Ads shows search estimates and bid estimates for your own phrases. Treat them as a range to plan around.
- Write the ceiling: the total you can lose.
- Choose the length of the test and set the end date in the campaign.
- Divide the ceiling by the number of days to get the average daily budget.
- Name the one action you will count, and confirm that a test visit is counted.
- Write the stop rules, with dates, before launch.
On a given day a campaign might spend up to twice its average daily budget. Over a month it spends no more than 30.4 times the average daily budget.
Google Ads Help: overdelivery and average daily budgetKeyword Planner shows monthly search estimates and the average cost for an ad to show on a keyword. Google notes that campaign performance depends on a variety of factors.
Google Ads Help: use Keyword PlannerWorked example: a fictional first test
Everything in this example is fictional, including the product, the phrases, the prices and the dates. The figures are planning assumptions. They are not benchmarks, customer results or output from a Soloop task.
ReviewQueue is an invented tool for small agencies. It lists deliverables waiting for client approval, shares review links and records each decision. In this scenario the founder has notes from ten conversations in which agency leads describe "chasing client approvals". A trial signup exists, and the founder can see when a new account shares its first review link.
| Dimension | Brief | Reason, or what would change it |
|---|---|---|
| Search theme | Client approval software for agencies. Three phrase-match keywords in one ad group. | The wording comes from the interview notes. If the results for these phrases show proofing tools for designers, the theme is wrong. |
| Exclusions | Negative keywords: free, template, jobs, salary, course, loan, and the names of unrelated approval products. | Each one is a meaning of "approval" the product does not serve. More come from the search terms report. |
| Landing page | One page for agency operations leads. The trial is the only next step. | The homepage speaks to several audiences, so a click from this ad could not be interpreted there. |
| Counted action | Trial signup and, separately, first review link shared. | A signup shows interest. A shared link shows use. Payment is a later, separate fact. |
| Budget and dates | $15 average daily budget from October 5 to October 18, 2026. Ceiling: $210. | One day may cost up to $30. The end date and the ceiling end the test. |
| Mid-test check | October 11: read the search terms report, add exclusions and compare spend with the ceiling. | If most search terms are off-theme, pause and fix the keywords before spending the second week. |
| Stop rules | Stop at $210 or on October 18, whichever comes first. Stop early if the page breaks or off-theme searches stay high after exclusions. | The rules are written before launch, so one good or bad day does not move them. |
| Decision after | On-theme searches and at least one shared review link: plan a second test. On-theme searches and no signups: revise the page first. | Two weeks at this size cannot establish a cost per customer. It shows whether the path from search to page works at all. |
Copy the first-campaign brief
Select and copy this brief into your own document; no account is required. Fill the plan column first, then write what would make you change it.
| Dimension | Plan for this test | Evidence or reversal condition |
|---|---|---|
| Buyer / search theme | [Role; three to five phrases with one intent] | [Where the wording came from; what current results show] |
| Match type / exclusions | [Phrase or exact; negative keywords] | [Meanings of the phrase you do not serve] |
| Ad promise | [One sentence the product already supports] | [Capability you inspected; benefit still unproven] |
| Landing page | [URL; the one next step] | [Who else the page addresses; what could confuse this visitor] |
| Counted action | [Signup, demo request or first use; how it is recorded] | [Confirmation that a test visit was counted] |
| Budget / dates | [Ceiling, average daily budget, start and end date] | [Highest single-day cost; who checks spend and when] |
| Stop rules | [Ceiling, end date, early-stop conditions] | [What would justify stopping sooner] |
| Decision after | [What you will do in each outcome] | [What this test cannot tell you] |
Read the results one stage at a time
When the test ends, look at each stage separately. Each gap points to a different fix, and a single total hides them.
- Few impressions: few people search the theme as you wrote it, or your bids were too low to show. Check Keyword Planner before widening the theme.
- Impressions but few clicks: the ad does not match what the searcher wanted. Compare the ad text with the search terms.
- Clicks but no counted action: the searchers were the wrong people, the page was unclear, or the next step asked too much. The search terms show which searchers you paid for.
- Counted actions: record the count and what each account did next. A handful of signups cannot give a reliable conversion rate or cost per customer.
Bring the brief to Soloop
You can run this brief yourself in Google Ads. To continue in Soloop, copy it into the conversation: My product is [URL]. Buyer [role], search theme [phrases], counted action [event], ceiling [amount], dates [start and end]. Prepare a Google Ads campaign for my review.
The Soloop manual says Soloop can prepare and launch Google Ads campaigns once you complete the campaign setup and the required payment steps. In the current setup, Soloop drafts a Search campaign: keywords with exact, phrase or broad match, and the ad text. You edit the draft, set one total budget and a start and end time, and approve the campaign before paying. It starts after payment is verified and stops at the end time. Google still makes the final ad approval decision. The campaign budget is separate from Soloop credits.
Some parts of this guide stay with you. The ad points to your project's deployed page. The workspace reports estimated cost, impressions and clicks; it does not show the search terms report, accept negative keywords or count signups today. Use phrase or exact match to keep searches close to your theme, and count the action in your own product. If you need the search terms report, run the test in your own Google Ads account.
Soloop does not promise clicks, signups or revenue from a campaign, and the stop rules remain yours. Follow the setup shown in your workspace, because these details can change. The Explore Soloop link opens the public entry, so bring the brief yourself.