Google Ads or Facebook Ads for a small business? The short answer: Google when people already search for what you sell, and Meta (Facebook and Instagram) when you need to show people something before they want it. A locksmith, an accountant or a dental practice usually gets better enquiries from Google. A pilates studio, a jewellery maker or a new restaurant usually reaches more people for less on Meta.
The bigger question is whether paid ads will pay off at all. This guide covers the difference between the two, 2026 cost benchmarks, which kinds of business win on each, a simple calculation of what you can pay per lead, and a five-question decision tree.
Do Google Ads work for small businesses?
They do when three things are true: people search for your service in your area, a click costs less than a customer is worth to you, and the ad lands on a page that answers the searcher's question. They don't when demand is thin, when your industry's clicks are expensive relative to your margins, or when the campaign targets broad keywords that attract the wrong searches.
The same test applies to Facebook and Instagram. "Is it worth it?" is really "will my cost per paying customer stay below the profit a customer brings?" You can estimate that before spending a dollar, and you can measure it after a month. Both are covered below.
The core difference: catching demand vs creating it
On Google, the ad appears in response to a search. Someone types "emergency plumber near me" and your ad answers them. You don't create the need, you catch it. The number of customers you can win is capped by how many people search, so Google has a natural ceiling.
On Meta, the ad appears in the feed, Stories and Reels of people who weren't looking for you. The system chooses who sees it based on what it knows about them and who responded to similar ads. That lets you create demand: someone who never thought about pilates sees a short video of a class near home and books a trial.
| Google (Search) | Meta (Facebook and Instagram) | |
|---|---|---|
| Who sees the ad | People searching for the service right now | People the system predicts will be interested |
| What you need to prepare | Keywords, text ads, a landing page | Creative: images, short video, copy |
| How the enquiry arrives | A call from the ad, a form on a landing page, a lead form | An instant form, a chat in Messenger, Instagram or WhatsApp, a landing page |
| Lead intent | Usually high | Varies, often "interested" rather than "ready to buy" |
| Where money gets wasted | Keywords that are too broad, irrelevant searches | Weak creative that wears out fast, forms so easy that people don't mean it |
| How fast you see data | Fast, if there is enough search volume | Fast, but finding creative that works takes time |
What each costs in 2026: benchmarks
Benchmarks don't tell you what you will pay, but they show the order of magnitude. Both sets below come from WordStream by LocaliQ and cover U.S. campaigns only.
| Median, all industries | Google Search | Facebook lead campaigns |
|---|---|---|
| Cost per click | $5.42 | $1.80 |
| Conversion rate | 8.18% | 8.54% |
| Cost per lead | $66.69 | $27.39 |
| Data | 13,474 campaigns, April 2025 to March 2026 | 452 lead campaigns, from April 2025 |
Sources: Google Ads benchmarks, published May 2026 and Facebook ads benchmarks, published September 2026, both as reported by PPC Land.
Industry matters more than platform. A few examples from the same reports:
| Industry | Google Search | Facebook lead campaigns |
|---|---|---|
| Attorneys and legal services | $9.87 per click, $131.63 per lead | No benchmark published |
| Dentists | $8.00 per click | $61.56 per lead |
| Home and home improvement | $8.33 per click | $42.95 per lead |
| Restaurants and food | $2.05 per click, $30.57 per lead | No benchmark published |
| Real estate | Not reported here | $13.74 per lead |
If you advertise in the UK, Canada or Australia, these U.S. medians are only a rough guide. Google's free Keyword Planner shows estimated click costs for your keywords and location, which is a better starting point.
Which businesses should start with Google
- Services people book when something goes wrong: plumbing, electrical, locksmiths, HVAC repair, towing, pest control. The search is the moment the need appears.
- Professionals people compare carefully: lawyers by practice area, accountants, mortgage brokers, dentists. They search, compare and call.
- Niche B2B: suppliers, office maintenance, labs. The audience is small, but the person searching is exactly the customer.
- Businesses that show up on the map: clinics, repair shops, local stores. A complete Google Business Profile brings enquiries without paying per click.
Google also has a pay-per-lead format for some local services. Local Services Ads charge only when a customer contacts you from the ad, and cover categories such as plumbers, electricians, HVAC technicians, dentists and lawyers. They run in the U.S., Canada, the UK and several European countries, for selected categories, so check eligibility for your trade and area.
Which businesses should start with Meta
- Products or services that look good: beauty, nails, hair, food, fashion, interior design. The photo or video does the selling.
- Services people don't know they need yet: workshops, classes, personal training, new treatments.
- Events and launches: a new location, a holiday offer, enrolment for a new term.
- Businesses that close in a conversation: Meta's messaging ads open a chat in Messenger, Instagram Direct or WhatsApp straight from the ad. In markets where customers already message businesses, that is a natural route. More on this in WhatsApp marketing for small business.
In Ads Manager you choose one of six campaign objectives: awareness, traffic, engagement, leads, app promotion and sales. For a business that wants enquiries, the right one is almost always leads. Creative is the main lever on Meta, and short, phone-shot or creator-style videos tend to suit it. AI UGC ads explains one cheap way to test several versions.
Cost per lead: why you can't compare them directly
The common mistake is "a Google lead cost $80, a Facebook lead cost $30, so Facebook wins". A Facebook instant-form lead can take two taps, with the details already filled in. A Google lead is often someone who searched, read your page and called. Their close rates can be very different.
The number to compare is cost per paying customer:
| Channel A | Channel B | |
|---|---|---|
| Cost per lead | $30 | $80 |
| Share who become customers | 5% | 25% |
| Cost per customer | $600 | $320 |
The figures are for illustration only. In your business the result could be the other way round, so log the source of every enquiry and whether it became a customer. After a month or two you have your own answer.
When neither will pay off
Before choosing a platform, do one calculation: how much can you pay per lead?
Profit per customer × close rate = your ceiling per lead
For example, if an average customer leaves $400 profit and you close one in four enquiries, your ceiling is $100 per lead. Above that, every new customer loses money unless they come back and buy again.
Paid ads won't pay off on Google or Meta when:
- Profit per sale is low and customers don't return. A $40 product with $12 profit is hard to fund from paid ads.
- There is nowhere to send people. Without a landing page, a complete profile or at least a page with prices and photos, you pay for clicks that go nowhere.
- Nobody replies quickly. A lead who hears back tomorrow has already talked to a competitor. Fix that first, for example with auto-replies and a clear handover to a person.
- The offer isn't clear. If you can't say in one sentence why someone should choose you, the ad won't say it for you.
- There is no patience for learning. The first month is usually a learning month. A budget that lasts one week tells you nothing.
Combining both on a small budget
With a small budget, one channel that works beats two that are half-starved. An order that suits most businesses:
- Free foundation: Google Business Profile, fast replies on the messaging channels customers use, a page with a clear offer.
- First channel: chosen with the sections above. It gets most of the budget for a month or two.
- Measurement: cost per customer, not just per lead.
- Second channel: once the first is stable. Often this is a small Meta retargeting campaign for people who visited your site from Google but didn't get in touch.
- Scale: increase spend in the channel that brings customers below your ceiling.
How much to set aside in total, and how to split it between ads, creative and tools, is in how much a small business should spend on marketing.
A decision tree in five questions
- Do people in your area search Google for your service? Check in Keyword Planner. If yes, go to question 2. If not, start on Meta.
- Does the customer need it now or soon? If yes, start on Google. If it's "someday", Meta is probably more efficient.
- Do you have good photos or videos of your work? If yes, Meta becomes more attractive. If not, Google lets you start with text alone.
- Does the estimated cost per click on Google let you stay under your ceiling per lead? If not, use Meta or free channels.
- Is someone available to answer enquiries quickly? If not, don't start either until that is solved.
If you'd rather have someone run these numbers with you, PowerAds is an AI marketing agent that works in WhatsApp and a web chat and runs campaigns on both Meta and Google Ads. You describe the business and the budget, and it shows a plan with cost and timing before anything goes live. Every charge appears as an approval card and runs only after you approve it, and the ad budget is paid to the platforms separately. What an AI marketing agent is explains how that kind of tool works and where you stay in control.
