How to determine the test budget without magic universal numbers. This guide sets out practical checks, examples and the limits of the recommendations.
Media spend is not the entire launch budget
The minimum amount for advertising does not depend on the universal tariff, but on demand, the cost of the targeted action and the risk that the business can accept. At Klientum, the calculator starts with a $300 advertising budget. This is the lower limit of the tool, not proof that $300 is enough for any niche. For a local service and an expensive B2B product, the same budget will yield different amounts of data.
Separate media spend, agency management and business preparation. Preparation may include a website, filming, a product feed or lead integrations. Check your plan's scope so you neither pay twice for the same work nor overlook work that needs separate approval.
Start with an acceptable customer acquisition cost
For the first purchase, the benchmark is not revenue, but the amount that remains after variable costs and that can be directed to attraction. If you're targeting repeat purchases, use actual retention data, not an optimistic forecast. Determine whether your CAC limit includes only platform costs or material handling and production as well.
A conditional example, not a case result: a business can spend $40 on attracting a client, and 10% of qualified enquiries go into sales. The price guide for such an enquiry is $4. If the conversion is 5%, the benchmark becomes $2. Such a calculation helps to form the problem, but does not predict the auction price.
Keep the test focused
For a small budget, narrow the test: one service or category, one realistic geography, a clear offer and an agreed target action. Don't split $300 between dozens of audiences, multiple platforms, and all products. First, determine the question that the test should answer: is there a demand, does the offer work, or can the desired quality of contact be obtained?
- Agree on the maximum costs of the test and the conditions for stopping it.
- Check that forms and calls are reaching the business.
- Consider the time from enquiry to sale: a quick conclusion can miss deals that have not yet been completed.
- Schedule a fallback for one subsequent hypothesis test, not endless restarts.
An agency-fee example with a separate media budget
For a Standard paid social plan and an $800 media budget: $160 fixed fee + 25% of $800 ($200) = $360 in agency fees. The indicative total is $1,160 per month, excluding separately agreed work. The percentage applies to the full budget, not only the amount above a threshold, so the result can change in steps at bracket boundaries.
Google Ads uses a separate fee schedule; do not apply paid social rates to it. The calculator accepts exact budgets, including amounts over $15,000. For these budgets, it shows an upper estimate using a commission of up to 10%; the final fee is agreed individually. See the current scope and calculation in our paid social plans and Google Ads packages.
When should the budget be revised?
If the test gives not enough targeted calls to conclude, it is not necessary to add money immediately. Check that the task is not too broad, that the page meets the request, that there are no losses in measurement and sales. The budget is increased when there is a clear limitation of scope and acceptable economy, not to compensate for a lack of understanding of the result.
In the case of children's playgrounds , cost per customer appears alongside cost per lead. Connecting enquiries to sales shows what the business receives for its spend. Do not apply that case's price to another market.
What to prepare for the conversation
- Product, geography and priority service.
- Margin or acceptable acquisition cost.
- Previous spend, enquiries and sales, if available.
- Available test budget separately from management.
- Contact person responsible for enquiry processing.
After that you can discuss launch of Google Ads or targeted advertising without a fictional universal minimum.
Before the next step
Record your starting data and decision criteria. Calculation examples are illustrative unless explicitly linked to a case study. Results from one project do not guarantee results for another business.
