How Much Should a Small Business Spend on Performance Marketing?

We know what you are probably wondering: how much are your competitors spending on marketing? The company that keeps appearing above you on Google, the business whose ads follow you around Instagram, the local competitor that seems to be everywhere. Is it a few hundred a month? Several thousand? Are they buying their way to all those inquiries, or do they know something you do not?

It is a reasonable question. Advertising feels less risky when you know what other businesses are doing. Unfortunately, knowing a competitor's budget would still leave the most important questions unanswered. You would not know their margins, how many inquiries become customers, whether those customers return, or whether the owner is quietly losing money on every campaign.

A useful advertising budget starts with the economics of your business. Competitor research helps you understand the market around it. Reverse that order, and you can end up funding someone else's assumptions.

In this article, you'll learn how to estimate a sensible starting budget, what cost per lead tells you, how to research competition without pretending to know private spending figures, and when increasing spend is justified. We will also walk through Google Trends, Keyword Planner, and competitor research tools, because the answer should leave you with something more useful than “it depends.”

Why there is no universal small business advertising budget

A bookkeeping firm selling ongoing services does not have the same economics as a restaurant selling a single meal. A specialist construction company may be happy to pay much more for an inquiry than a business selling a low-margin household product. A new company without reviews or a clear offer faces a different problem from an established company with spare capacity and a proven sales process.

Even within one industry, two businesses can afford very different acquisition costs. One has a high renewal rate. Another earns most of its money on the first transaction. One can answer an inquiry within minutes. Another responds three days later. One has a recognizable niche and clear pricing. Another sends every visitor to a general homepage.

That is why a recommendation such as “spend ten percent of revenue” can serve as a rough planning conversation but cannot tell you whether a campaign will work. It also becomes awkward for a business with little revenue: ten percent of almost nothing may buy too little information to evaluate anything.

Start with four questions instead. What is a new customer worth in contribution after the cost of delivering the service? How much of that contribution can you afford to spend acquiring the customer? How many customers can you serve? How much uncertainty can the business finance while testing?

Those questions connect advertising to the business you actually operate.

Understand what CPL measures before you use it

Cost per lead, usually shortened to CPL, is advertising spend divided by the number of leads attributed to that spend. Spend £1,000 and receive 20 inquiries, and your media CPL is £50. That calculation is simple. Deciding whether £50 is good requires more work.

First, define a lead. A person who clicks a phone number is not necessarily a completed call. A form submission can be spam, a job application, or an inquiry outside your service area. An instant form on social media may capture someone who has not read your pricing. If you count all of those events together, the average can look reassuring while the sales team sees very little useful demand.

Keep at least three stages visible: an initial inquiry, a qualified opportunity, and a paying customer. A qualified opportunity should meet criteria relevant to your business, such as service, location, timing, and realistic budget. You do not need a complicated CRM to begin. A consistently maintained spreadsheet is more useful than an elaborate system nobody updates.

Be consistent about the period and costs in each calculation. If your sales process takes six weeks, dividing this month's spend by this month's closed sales may mix unrelated groups. Track leads by the month they arrived, then update their eventual outcomes. That gives you a much clearer picture of what the original campaign produced.

Supply, demand, and competition affect prices—but they are not the whole equation

Advertising costs are partly a market story. How many people are looking for a service? How many advertisers want to reach them? How valuable is that customer to those advertisers? How narrow is the location, audience, or specialist requirement?

On Google Search, advertisers compete in auctions. Google explains that bids are only one part of the process; ad and landing-page quality, context, thresholds, and the expected effect of assets also matter.[1] A business cannot assume that the highest bidder wins every useful position or that paying more fixes a weak destination page.

Specialization adds another layer. A narrow service might attract fewer searches, but those searches can express a much clearer need. Its clicks could cost more because competitors value them highly, or less because fewer advertisers target that niche. The better question is whether those visitors turn into profitable customers.

For a simplified campaign, CPL can be expressed as cost per click divided by the click-to-lead conversion rate. At a £4 cost per click and a 5% conversion rate, the implied CPL is £80. At the same click price and a 10% conversion rate, it is £40. These are illustrative assumptions, not expected results. They show why your offer, website, and measurement can matter as much as the auction price.

There is also a tradeoff hidden inside conversion rates. Removing all qualification from a form can create more leads and a lower reported CPL. It can simultaneously make the sales process worse. You want the right amount of friction: enough information to establish a useful next step, without turning a straightforward inquiry into an application to buy your service.

What current industry benchmarks can—and cannot—tell you

Benchmarks are useful for checking whether your assumptions are wildly disconnected from observed advertising costs. They are much less useful as promises.

LocaliQ's 2026 search advertising report draws on its customers' Google Ads and Microsoft Ads campaigns. The published figures include the following category averages.[2] This is a vendor sample, not a representative forecast for every UK or US small business, and search advertising figures should not be treated as Meta advertising benchmarks.

The differences illustrate why generic budget advice is weak. They do not mean that your legal practice should expect precisely $131.63 per lead or that a restaurant can buy a profitable customer for $30.57. Lead definitions, campaign mix, locations, and the businesses represented all influence an average.

Use a relevant benchmark as one input when creating a cautious, central, and optimistic scenario. Then replace it with your own evidence as soon as you have reliable data. If the published average is far below your result, investigate the reason before assuming your campaign is broken. You might be selling a more specialized service, measuring a stricter outcome, or targeting a more competitive area.

Work backward from what a customer is worth

Consider a hypothetical service business that earns £800 in contribution from an average new customer before acquisition expenses. This is revenue after the variable costs of delivering the work, not simply the invoice amount. The owner decides that a media acquisition cost of up to £200 is acceptable, leaving room for other expenses and profit.

If 20% of inquiries become customers, the implied target media CPL is £40: £200 multiplied by 0.20. If only 10% become customers, the implied target drops to £20. The campaign has not necessarily become more expensive; the sales process now requires more inquiries to produce the same result.

This is a planning model, not a forecast. Forty leads will not reliably produce exactly eight customers, especially in a small sample. The point is to make assumptions visible so you can challenge them.

For recurring services, you may consider contribution across an expected customer relationship. Be careful. A new business cannot spend hypothetical future lifetime value with the same confidence as cash already collected. Use conservative retention assumptions and consider how long acquisition costs take to recover. A campaign can appear profitable on paper and still create a cash-flow problem.

Research demand with Google Trends

Google Trends is useful for understanding patterns in interest. It is not a tool for finding an exact monthly search volume. Its values are normalized relative interest on a scale from zero to 100; low-volume terms may not produce useful data.[3]

Begin with the service a customer would actually name. Open Google Trends, enter the term, and select the country you want to assess. Research the UK and the US separately before drawing conclusions about both. A combined English-language strategy does not mean demand behaves identically in both markets.

Use a longer period to examine seasonality and a shorter period to inspect recent movement. Compare related terms within the same view, keeping the location and time range consistent. Pay attention to whether you are comparing search terms or topics. A topic can represent a broader concept than the exact words you typed.

Suppose you offer garden maintenance. A spring rise might suggest that preparing campaigns and landing pages before the busy period is more useful than waiting until the peak. Suppose a specialist service has almost no usable Trends data. That does not prove nobody needs it. The query may be too narrow for the tool, and Keyword Planner or your own inquiry records may be more informative.

Record what you learned in plain language: seasonal peak, stable interest, unclear data, or different terminology by market. Avoid turning a relative chart into a precise revenue forecast. Trends helps you ask better questions about timing and language; it does not price your campaign.

Use Keyword Planner to test search-market assumptions

In Google Ads, open Keyword Planner through the planning tools and choose the option to discover new keywords. Start with a small set of real services or a relevant website. Set the location and language deliberately. If you serve one region, a whole-country estimate may be misleading.

Review suggested terms before accepting them. A list can contain research queries, job searches, product searches, and services you do not offer. Group terms by the need behind them: emergency help, comparison, price research, or a specific specialist service. These groups may deserve different content and budgets.

Google's historical metrics include estimated average searches and advertiser competition; the figures depend on the selected settings and can include close variants. Advertiser competition is not organic SEO difficulty. Forecasts are estimates affected by assumptions, rather than commitments about your future results.[4]

Download a focused shortlist. Add columns for relevance, likely intent, landing page, service margin, and whether you can fulfill that demand. Twenty carefully chosen terms are more useful than hundreds of loosely related ideas. A high search volume for work you do not want is not a commercial opportunity.

When examining suggested bid ranges, treat them as context for potential click costs. They do not tell you the cost of a lead or customer. You still need a conversion assumption. A £6 click with a strong offer may be more useful than a £1 click from someone looking for a free template.

Look at competitors without inventing their spending

You can learn a great deal from public advertising. Meta's Ad Library lets you inspect ads available in its library. Google's Ads Transparency Center can help you examine advertiser creatives. Look at the offers, language, landing pages, formats, and whether the same message appears repeatedly.

For ordinary commercial campaigns, do not assume those tools disclose a competitor's complete actual budget. Special transparency disclosures for political or other regulated advertising are a different case. The number of visible ads is not a reliable spending calculator: one business may run many low-budget variations, while another spends heavily on a few.

Ahrefs and similar tools can add another view of search visibility and modeled traffic. Ahrefs describes paid traffic cost as an estimate based on observed paid traffic and click prices; organic traffic value is a hypothetical paid equivalent, not money the business necessarily spent.[5] These are directional research tools, not access to a competitor's finance department.

Write a short competitor note rather than collecting screenshots indefinitely. What do they promise? What proof do they show? Where does their page leave a question unanswered? What could your business credibly explain better? That is where research becomes useful marketing work.

Choose a test your budget can actually support

A small budget divided across five platforms, three countries, and twelve services often creates fragments of information. You can see some clicks without learning why anything happened. Start with a clear hypothesis and enough focus to evaluate it.

For example: “People searching for this specific service in this region will request a quote when they see a clear scope, relevant proof, and a simple form.” That hypothesis points toward a search campaign, a specific landing page, and an inquiry outcome. A different hypothesis might involve demonstrating an unfamiliar product through Meta video and measuring qualified interest.

Estimate what the proposed spend could buy under several CPL assumptions. At a hypothetical £50 CPL, £300 buys six leads; at £100 it buys three. Neither sample can support confident claims about a stable closing rate. A limited test may still reveal broken tracking, irrelevant queries, or weak interest, but its conclusions should remain proportionate to its size.

Agree on a spending boundary and review criteria before launch. Define what would cause an immediate pause, such as incorrect geography or a broken form, and what requires more observation, such as a long sales cycle. There is no universal number of conversions that makes every small business experiment conclusive.

Include the costs outside the advertising platform

Your media budget pays for distribution. It does not automatically include the landing page, creative production, campaign management, measurement setup, or time spent handling inquiries. Keep those costs visible, particularly when comparing agencies.

A business spending £1,000 on media and £700 on supporting work is committing £1,700, even if the dashboard only shows the first figure. That wider amount can be reasonable if the work creates useful assets and qualified customers. It becomes misleading when someone presents media-only results as the complete cost of growth.

Some expenses are initial foundations. A functioning landing page may support several months of campaigns. Other expenses recur. A sensible plan distinguishes those categories rather than spreading every cost across a single attractive CPL figure. Our guide to small business marketing agency costs explains how to compare that wider scope.

Operational capacity belongs in the budget conversation too. If your team can only fulfill five additional projects this month, buying enough demand for twenty may create delayed responses and poor customer experiences. Fix capacity, qualification, or pricing before assuming that more leads are always desirable.

Increase spending when the evidence supports it

Scaling is not simply changing a number after a good week. Check whether customer quality is holding, whether sales outcomes are mature enough to evaluate, and whether the team can deliver the extra work. Look for a pattern across an appropriate period, including the lag between inquiry and payment.

When you do increase spend, change it in a way that lets you understand the result. Record the date, reason, and expected effect. Avoid changing the budget, audience, landing page, offer, and tracking simultaneously unless a serious problem requires it. Otherwise, even a better result will be difficult to explain.

Watch the next increment of spend. The first customers can be easier to reach than the next group, so an average acquisition cost that worked at one budget may not hold at a larger one. The goal is profitable additional business, not preserving a beautiful dashboard average at any cost.

If results deteriorate, return to the funnel. Did clicks become more expensive? Did fewer visitors inquire? Did qualification decline? Did the sales response slow down? Each problem calls for a different response. “The ads stopped working” is a starting observation, not a diagnosis.

Read one campaign through the whole funnel

Imagine a hypothetical local service business spending £1,200 on a test. It records 30 inquiries, giving a media CPL of £40. At first glance that meets the owner's target. The sales record tells the rest of the story: 18 inquiries are suitable, nine become conversations, and three become customers. The media acquisition cost is £400 per customer, and the cost per qualified inquiry is approximately £66.67.

Now imagine another test spending the same amount and generating only 20 inquiries. Its reported CPL is worse at £60. But 16 inquiries are suitable and four become customers. Media acquisition cost is £300. If customer contribution and delivery costs are comparable, the second test produces more customers at a lower acquisition cost despite the less attractive headline CPL.

Neither example establishes which platform or creative format is best. They show why you need consistent records beyond the advertising dashboard. They also reveal where to investigate. In the first case, why are twelve inquiries unsuitable? In both cases, how quickly are conversations started? Are the customers buying the same service at the same margin?

Use those questions to plan the next change. Tightening the offer may reduce total lead volume while improving the business result. Improving response time may create value without buying another click. Budget planning becomes much more useful when it can recognize those possibilities.

A budget you can explain is a budget you can improve

The right starting figure is one you can connect to a clear offer, plausible demand, customer contribution, a defined experiment, and a loss the business can tolerate. It should leave enough room to fulfill the work and learn from the results.

Research your competitors, but do not borrow their economics. Read benchmarks, but do not mistake an average for a promise. Use CPL, but follow those leads all the way through to customers and contribution.

At Startup Harbour, that means looking at the website, offer, measurement, creative, and follow-up together. Human judgment sets the commercial direction; tools help the team research, produce, and monitor the work. If those foundations are unclear, buying more traffic usually makes the uncertainty more expensive.

Before your next campaign, write down one page: the service you want to sell, the customer you want, the contribution available, your target acquisition cost, the assumptions behind your lead target, and the point at which you will review the evidence. That page will teach you more about a sensible advertising budget than guessing what the company down the road spends.

Sources and further reading

1. Google Ads: How the ad auction works.
2. LocaliQ: 2026 Search Advertising Benchmarks. Vendor campaign sample; figures are context, not a forecast.
3. Google Trends: FAQ about Google Trends data.
4. Google Ads: Keyword Planner metrics and forecasts and discovering new keywords.
5. Ahrefs: Traffic value and paid traffic cost.
6. Meta Ad Library and Google advertising transparency guidance.

All worked business examples in this article are hypothetical. They illustrate a planning method and are not reported client results or promised advertising outcomes.