Search Ads on a Small Budget: When They Pay Off and How to Measure It

Why do small budgets burn so quickly in search ads?

You set up an account, choose a few search terms, set 500 euros a month and let the system do its thing. After two months, 1,000 euros are gone, there were 300 clicks and four inquiries, three of which did not fit. The verdict: search ads do not work for us.

Search ads, often called Google Ads after the largest provider, are paid listings above or next to the regular search results. You usually pay for each click. That makes them fair, because you only pay when someone comes. But it also makes them dangerous, because every click from someone who does not want to buy costs the same as a click from a customer.

With a large budget, such mistakes can be offset for a while. With a small budget they cannot. At 1,000 euros a month, every single search term decides whether the money reaches prospects or the curious, students and competitors. However, precisely this limitation forces decisions that would benefit large accounts too. Put simply, the question is: when do search ads pay off on a small budget, and how do you set them up so that the money reaches prospects? To answer it, we first calculate backward what a click may cost, then choose suitable search terms, ask whether a click has any additional effect at all, and go through the landing page, measurement, and the question of when to manage it yourself, hand it over, or stop.

On a small budget, the winner is not the one who bids the most but the one who knows most precisely which click they do not want.

The backward calculation: what may a click cost?

Before you spend a single euro, you calculate backwards. I call it the backward calculation: from the order back to the click. It shows whether search ads can work for your service at all. You need four numbers that you know from your business or can estimate.

1. What is left from an order?

   The contribution margin is an order’s revenue minus direct costs such as materials and subcontracted work. From that, you decide what share you are willing to spend on winning a new customer.

2. How many inquiries does an order take?

   From your inquiry list: how many inquiries lead to one order on average? This gives you what an inquiry may cost at most.

3. How many visitors does an inquiry take?

   The inquiry rate of your landing page, meaning inquiries divided by visitors. Without your own figure, start with a cautious assumption.

4. What may a click cost?

   The affordable cost per inquiry times the inquiry rate gives the maximum price per click. If typical click prices for your terms are higher, the calculation does not work.

In the example, a click may cost up to 9 euros. If clicks for the most important search terms cost 3 to 5 euros, there is room for mistakes and learning costs. If they cost 15 euros, you need to work on the inquiry rate, the close rate or the choice of terms before ads pay off. How much a customer is worth over the whole relationship, and whether you can therefore spend more on the first order, is described in the article Calculating customer value.

Which search terms pay off on a small budget?

On a small budget, you only buy searches with buying intent behind them. The article SEO for small and midsize businesses distinguishes four types of search queries. For ads on a small budget, almost only one is an option: service plus location, or service plus a concrete occasion.

Research also supports the idea that specific terms work better. Huang and colleagues (2023) analyzed data from a large Chinese e-commerce platform. Click and purchase rates fell as ad position worsened. Specific search terms, however, weakened this dependence on position. Someone who knows exactly what they are looking for also clicks on the third ad if it fits. For small budgets this is decisive, because you can rarely pay for first place.

• Set the match type. Use phrase match or exact match instead of broad match, where the platform itself decides what is similar.
• Maintain exclusions. Terms such as “training”, “salary”, “free”, “template” or “job opening” are excluded. The list grows every week.
• Limit the region. Ads appear only where you actually work.
• Limit the times. If inquiries come mainly by phone, ads run only when someone answers the phone.

Does a click really bring something that would not have happened without the ad?

Too few advertisers ask this question. Blake, Nosko and Tadelis (2015) studied it at eBay in a series of large field experiments. They deliberately switched ads off and compared sales. The result was sobering: because people with buying intent click more often anyway, simple analyses considerably overestimate the benefit of search ads.

In the extreme case, ads on the company’s own brand name, they found no measurable short-term benefit at all. Someone searching for “eBay” finds eBay even without an ad. For generic terms, ads had a positive effect on new and infrequent users. But most of the money went to frequent users who would have bought anyway, so the average return was negative.

eBay is one of the best-known brands in the world; your business probably is not. If competitors run ads on your name, an ad of your own on it can make sense. The lesson still holds: do not count how many inquiries came via ads, ask how many of them would not have come without ads. New customers who do not know you yet are the part for which ads are most likely to pay off.

Conversely, ads sometimes work without a click. Sahni and Nair (2020) studied how ads in mobile search results work on the restaurant search platform Zomato with more than 200,000 users. The additional conversions the ads brought did not come about through clicks on the ad. Users saw the ad and later clicked on the same provider’s regular listing. Merely seeing it was enough for most of the effect. For you this means: do not judge ads only by clicks, but by whether more inquiries from new customers come in overall.

Where does the ad lead? Landing page and availability

The click is paid for. What happens next decides whether the money comes back. Two things must be right: the page the visitor lands on, and the response to their inquiry.

The ad never leads to the home page but to a page that picks up exactly the ad’s promise, addresses one audience and offers a single next step. How that page is built is described in the article Building a landing page. A better inquiry rate is the cheapest lever of all: if it rises from 3 to 4.5 percent, cost per inquiry falls by a third without you spending a single euro more.

Then availability. Someone who clicks on an ad is usually comparing several providers. An inquiry from an ad is therefore particularly time-critical. How to set responsibility and response time so that paid inquiries do not sit in the inbox is covered in the article Aligning marketing and sales.

How do you measure whether search ads pay off?

The ad account shows clicks, click prices and, if set up, submitted forms. Whether the money is well spent only becomes clear when it is linked to the inquiry list. The following worked example shows one quarter at 1,000 euros a month.

You need three things for this. First, conversion tracking in the ad account that counts submitted forms and calls, set up with due regard to your visitors’ consent. Second, the source “search ads” on every inquiry, kept until the order or the rejection. Third, patience: with three orders per quarter, chance plays a part. Judge after one quarter at the earliest, better after two. What the whole measurement from click to close looks like is described in the article Measuring marketing success.

A survey of 302 e-commerce professionals by Makrydakis, Spiliotopoulos and Lymperi (2026) fits this sober view. Search engine marketing showed the biggest gap there between effectiveness and efficiency: it delivers results, but the costs from bidding competition eat up much of them again. The respondents named the difficulty of attributing success to the right channels as the biggest obstacle.

Do it yourself or hand it off, and when do you stop?

A small account with ten to twenty search terms, one region and one landing page can be run yourself if you plan 30 minutes a week for it: review search terms, add exclusions, match inquiries. If you cannot or do not want to do that, it is better to hand it off than to let it run unattended.

If you hand it off, make sure the ad account is registered to your business and you have full access, and agree on reports that go all the way to inquiries and orders. Which questions to ask any agency is covered in the article Choosing an agency. One option is quik-services, a provider of websites, SEO and lead generation and a business partner of Owners Council.

Pausing is not a defeat. Often the better next step is to first get the free business profile and your own website in order. How the profile leads to inquiries is described in the article Google Business Profile.

The next step

Before you set up an account or let the existing one keep running, take an hour for the calculation and the basics.

• Do the backward calculation with your numbers: contribution margin, share for new customers, inquiries per order, inquiry rate.
• Note the maximum price per click and compare it with the estimated click prices for your most important terms.
• Choose ten search terms with buying intent and create an exclusion list.
• Limit region and ad times to your actual service area and availability.
• Build or check a landing page for the most important service.
• Add the source “search ads” to the inquiry list and schedule a weekly 30-minute slot.

After one quarter you will know your cost per order. Then you decide with your own numbers rather than gut feeling whether search ads become a permanent channel.

We set out to ask
search ads pay off on a small budget and how to set them up. They pay off when the search terms show buying intent, the landing page matches the search, and someone handles inquiries quickly. The backward calculation shows in advance the maximum a click may cost. Whether it works out is measured by cost per order, not by click price.