Beyond Google Ads: Where Meta, Microsoft and Amazon Fit in a UK Service Business PPC Budget

Google, Microsoft, Meta and Amazon compared for local service businesses

Most trades and service firms start with Google Ads, and for good reason. Someone types "emergency electrician Worthing" and your ad appears at the moment they need you. The trouble is that a single platform is a single point of failure, and the cost of that platform tends to creep upwards over time. This section explains what each alternative actually does, what it costs in the UK, and which kind of business it suits.

Why relying on Google alone gets expensive

There are two separate problems. The first is price. Cost per click is set by an auction, so when a competitor, a national directory or a franchise starts bidding on your keywords, your price rises even though your work has not changed. Imagine a plumber paying £4.50 a click on a £600 monthly budget. That buys about 133 clicks. If the average click climbs to £6.50 over a year, the same £600 buys about 92 clicks. That is roughly 30% fewer enquiries for the same money, and nothing you did caused it.

The second problem is a ceiling. Search ads can only reach people who are already searching. If 300 people a month in your area search for "boiler service" and you already appear for most of them, doubling your budget will not double your enquiries. You will simply pay more for the same people. Beyond that point, the next pound is better spent on a different type of audience, which is where the other platforms come in.

There is also a practical risk. Google accounts can be suspended over a policy flag, a billing hiccup or a misread landing page, and getting reinstated can take days. If every enquiry you receive comes from one account, that is a week of silence in your diary.

The searching, scrolling or shopping test

Ask yourself what the customer is doing at the moment your ad appears. This quickly narrows down which platform is worth testing.

• Searching: they have a problem and are typing it into a search box. Google and Microsoft Advertising both serve this person. A burst pipe, a locked-out homeowner or someone comparing quotes for a new roof all fall here.
• Scrolling: they are looking at friends, football or recipes and did not know they wanted you. Facebook and Instagram (both run through Meta) reach this person. Patios, extensions, kitchen fitting, window cleaning rounds and anything visual or planned suit this mode.
• Shopping: they are ready to buy a physical product and are comparing prices and reviews. Amazon is built for this person, and only if you sell products that can be posted.

Most service businesses are searching or scrolling businesses. Very few are shopping businesses, which is why Amazon is the last option on the list for the typical tradesperson.

Microsoft Advertising: the closest cousin to Google

Microsoft Advertising runs ads on Bing, and on the search engines that use Bing's results, including Yahoo, DuckDuckGo and Ecosia. In the UK, Bing handles roughly one in ten searches overall, but the share is noticeably higher on desktop computers and on work or Edge browser devices. The people using it skew older, are more likely to be at a desk during working hours, and often have a higher household income than the average social media user. For a business that serves homeowners over 45, landlords, offices or other small businesses, that is a useful audience.

The big attraction is cost. Click prices are often 20% to 30% lower than Google for the same keywords, because fewer advertisers compete. A keyword costing £3.50 on Google might cost £2.50 to £2.80 on Microsoft. You also get less volume, so a Microsoft account is a supplement, not a replacement.

A worked example. Say an electrician puts £300 a month into Microsoft Advertising at an average £2.50 a click. That is 120 clicks. If 9% of visitors phone or fill in a form, that is about 11 enquiries, or £27 each. If roughly one in three enquiries turns into a booked job, that is 4 jobs at about £75 each. With an average job worth £380, it is a good result. If the same figures came out at £150 per job on a £90 job, you would stop.

Setting it up is easier than most people expect. Microsoft Advertising has an import tool that copies your existing Google Ads campaigns, keywords, ads and location settings across in a few clicks. Here is where people go wrong:

• Leaving everything as imported. Check that match types and negative keywords came across properly, because broad matching can pull in irrelevant searches such as "electrician jobs" or "electrician courses".
• Not checking location settings. Choose to target people in your area, not people who merely show interest in it, unless you really do want holidaymakers and people researching from elsewhere.
• Using automated bidding too early. With low volume the automatic strategies have little to learn from. Start with manual bidding or a capped "maximise clicks" setting, and move to automated bidding once you have about 30 conversions.
• Leaving the partner network on without looking. Review the placement and search partner reports in the first month and switch off anything sending traffic that never converts.

You will also need to add the Microsoft UET tag to your website so conversions are recorded. Microsoft Clarity, a free tool from the same company, records anonymous session replays and heatmaps, which is a cheap way to see why visitors leave your landing page without calling.

One extra option that Google does not offer in the same way is LinkedIn profile targeting. You can adjust bids for people in certain industries, job functions or company types. For a commercial cleaner, an IT support firm or an accountant selling to other businesses, that can be worth a look. For a domestic trade it is irrelevant.

Meta (Facebook and Instagram): reaching people before they search

Meta's platforms are the opposite of search. Nobody is looking for you, so the ad has to earn attention in a second or two. That makes the creative (the photo or video and the first line) more important than anything else in the account. It also makes Meta better at generating demand for work that people plan, rather than work that is an emergency.

The UK costs are lower per click than search, but the intent is lower too. As a rough guide, expect a click to cost £0.40 to £1.50 and a lead form submission to cost £8 to £35 for most local services. Higher-value jobs such as landscaping, extensions and bespoke joinery tend to land towards the upper end, because you are competing for attention with every other advertiser in the area. Treat those figures as ballparks for planning, not promises, and compare them with your own results.

A worked example. A landscaper spends £400 in a month promoting finished patio projects to homeowners within 15 miles. At about 90p a click that is around 440 clicks. If 4% of visitors send an enquiry, that is about 18 leads at roughly £22 each. Perhaps 6 of them are serious enough for a site visit, and 2 become jobs worth around £5,500 each. The cost per job is £200, which is excellent against an £11,000 return. The same campaign for a window cleaner whose jobs are worth £15 a visit would be a failure, because the enquiry cost alone would swallow the first few months of income.

Targeting that works for local businesses on Meta

• Location first. You can target by town, by postcode area or by a radius around a pin on the map. Choose "people living in this location" rather than "people living in or recently in" so that holidaymakers and passing commuters are excluded.
• Keep audiences broad. Since Meta removed many of its detailed interest options in 2022, and as its system has improved at finding likely buyers, tight targeting by interest, such as "home improvement", matters less than it used to. A sensible age range and a good location with a strong ad usually outperform a complicated audience.
• Retargeting. Show a reminder ad to people who visited your website or engaged with your page in the last 30 days. These audiences are small and cheap to reach, and are often the best-performing part of a Meta account.
• Lookalikes from customers. If you have a list of at least a few hundred past customers, uploading it can help Meta find similar people, although this is a second-stage tactic, not a starting point.

For creative, use real photos and short phone-shot video (15 seconds is plenty) of your own work and your own team. Before and after images do particularly well. Stock photos of smiling people in hard hats are ignored. Put a clear offer or next step in the ad, such as "Free quotes for patios in Worthing and Littlehampton, booking for spring".

A few traps are worth knowing about:

• Boosting posts. The blue "Boost post" button is the easiest way to spend money on Facebook and the least controlled. Use Ads Manager so you choose the objective, location and placement properly.
• Instant Forms attract casual leads. Meta's built-in lead forms are quick to fill in, so you get more leads that are less committed. Choose the "higher intent" form option, add a question such as "What do you need doing?" and a postcode field, and phone back within 10 minutes. Leads go cold quickly.
• The learning phase. Meta's system wants about 50 conversions per week per ad set to optimise properly, which a small business will not reach. Run one ad set with a few ads, optimise for leads rather than purchases, and avoid editing the budget every day.
• Tracking setup. Install the Meta Pixel and, ideally, the Conversions API too, because browser privacy settings block a growing share of pixel data.

Amazon Ads: a niche option for product sellers

Amazon advertising is for selling products on Amazon. There is no way to advertise a plumbing call-out, a roofing quote or a wedding photography package. It becomes relevant when a service business also makes or sells something that can be posted: a pest control firm with its own proofing kits, a cleaning company with a branded product range, a locksmith selling lock sets, or a joinery that makes small wooden items.

The main ad type is Sponsored Products, where your listing appears in search results and on competing product pages. UK click costs vary widely by category, from about 30p to well over £1.50. You pay Amazon on top of the ad spend, too. A Professional seller account costs around £25 plus VAT a month, and most categories carry a referral fee of around 15% of the sale price, plus fulfilment fees if Amazon posts the goods for you. Check the current rates before you work out your margins.

The number that matters is ACOS (advertising cost of sale), which is ad spend divided by ad sales. Your break-even ACOS is your profit margin before advertising. For example, a £30 product leaves £10.50 after Amazon fees, postage and the cost of the item, which is a 35% margin. If a click costs £1 and 1 in 10 clicks buys, you spend £10 to make a £30 sale. That is an ACOS of 33%, only just inside break-even. It looks fine on a dashboard and loses money once you count your time. Amazon works when margins are healthy and the listing already has plenty of good reviews. It rarely works on a brand-new listing with none.

The platforms side by side

• Google Ads: highest volume and highest cost per click. Best for urgent, high-intent work. Start here.
• Microsoft Advertising: lower volume, often 20% to 30% cheaper clicks, an older and desktop-heavy audience, and simple import from Google. Best first test for most trades.
• Meta (Facebook and Instagram): cheap clicks, lower intent, strong for visual and planned work. Needs good photos and quick follow-up.
• Amazon Ads: only for businesses selling physical products on Amazon. Margins and reviews decide everything.

What a diversified budget can look like

There is no standard split, but these examples show the thinking. They assume you already have Google working.

• Under £500 a month: stay on Google only. Splitting a small budget across platforms leaves each one with too little data to learn from. Spend your effort on the landing page and call handling instead.
• Around £1,000 a month: keep about £700 on Google and put £200 to £300 into one test platform, usually Microsoft for emergency and repair work, or Meta for planned, visual projects.
• Around £2,000 a month: roughly £1,200 on Google, £300 on Microsoft, £400 on Meta (including £100 for retargeting), and any remaining money held back for the month's best performer.

Budgets should follow your season, too. A boiler engineer might push Google and Microsoft hard from October to February, then switch to boiler service reminders on Meta in summer. A landscaper will do the opposite, spending on Meta in February and March to fill the spring diary.

Tracking across several platforms

Every platform is happy to claim credit for the same enquiry, so if you add the dashboards together you will usually count more leads than you actually received. A little discipline avoids that.

• Install each platform's tag (Google tag, Microsoft UET, Meta Pixel) through Google Tag Manager so you control them in one place.
• Add UTM tags to every ad link, for example utm_source=facebook&utm_medium=paid_social&utm_campaign=patio-spring, so Google Analytics 4 shows where each visit came from.
• Use call tracking numbers for each platform. Tools such as CallRail and WhatConverts are widely used, and they record which ad produced each phone call. For a trade, phone calls are often most of the enquiries.
• Ask "How did you hear about us?" on every call and form, and write down the answer. What customers say frequently differs from what the dashboard says, and both are useful.
• Keep a simple spreadsheet with the date, source, type of enquiry, whether you quoted, whether you won it and the job value. After three months it will tell you more than any platform report.

UK rules that catch advertisers out

The Advertising Standards Authority enforces the CAP Code, and it applies to paid ads on every platform. Claims such as "Gas Safe registered", "NICEIC approved" or "fully insured" must be true and provable. Words such as "cheapest", "best" or "guaranteed" need evidence behind them. If you advertise "from £99", that price should be one you genuinely charge for a real job. Testimonials must be genuine.

Privacy law matters because ad tags place cookies. Under UK GDPR and PECR, tracking tags should only fire after a visitor has agreed through a cookie banner. Google and Microsoft both offer consent modes that adjust measurement when someone says no, and Meta has similar settings. Also check whether the budget figures you see include VAT. On most platforms it is added on top of your spend, so a £10 a day budget will cost slightly more than £300 a month by the time the invoice arrives.

A 90 day plan for testing a second platform without wasting money

Treat the second platform as an experiment with a start date, a cap and a finish line. Without those three things, it quietly becomes a monthly cost nobody questions.

Days 1 to 14: get the foundations right

Before you spend anything, check that a phone call, a form fill and a quote request each get recorded as a conversion. Then look at the landing page: one service, one clear offer, your phone number at the top, and a form that works on a phone. Pick your second platform using the searching, scrolling or shopping test. Pick one only. Microsoft suits searchers, Meta suits scrollers, Amazon suits product sellers.

Test the tracking yourself. Click your own ad from a phone, ring the tracked number and submit the form, then confirm each action appears in the platform. Check your service area settings too, and write a starting list of negative keywords such as "jobs", "training", "free", "DIY" and "course" for search platforms. Agree beforehand who answers the phone and how fast. An enquiry that waits two hours for a reply is often already booked with someone else.

Days 15 to 60: a capped test

Set a fixed daily budget you can live with, for example £10 to £15 a day, and run one or two campaigns at most. Every week, check cost per lead and open the actual enquiries. A cheap lead who wants a job 40 miles away, or just wants a free price, is not a cheap lead. Resist tweaking daily, because small numbers move around a lot.

A short weekly review is enough. Spend 20 minutes on the same four questions each time: how much did we spend, how many enquiries did we get, how many were real jobs in our area, and what search terms or ads produced the worst ones. On a search platform, add the poor search terms as negative keywords. On Meta, pause the weakest ad only when it has had a fair run, which is generally at least a few hundred pounds of combined spend across the account or a couple of thousand impressions on that ad. As a rule of thumb, avoid judging a platform until it has produced at least 15 to 20 enquiries, because anything less is mostly luck.

Days 61 to 90: decide on cost per job

Now work out what each booked job cost you, not what each click cost. Then pick one: scale it up gently (by around 20% at a time), adjust the targeting or offer and run another month, or stop. Write down what you learned, including which ads, areas and enquiry types worked. That note is worth more than the ad spend.

Here is how the decision might look. You spent £700 over the test. It generated 24 enquiries, so £29 per enquiry. Of those, 16 were in your area, 8 became quotes and 3 turned into jobs worth £420 each, giving £1,260 in revenue. Cost per job was £233. If your margin on a £420 job is around £170, you lost money on the first jobs, but if those customers rebook or recommend you, the picture changes. Judge it on what a customer is worth to you over a year, not only on the first invoice. If the same test produced one job, or none, adjust or stop.

Common mistakes that waste a test

• Running two or three platforms at once, which splits the budget and muddies the results.
• Judging by click-through rate or likes instead of enquiries and jobs.
• Sending traffic to the homepage rather than a page that matches the ad.
• Changing the ad, budget and audience in the same week, so you never learn which change made a difference.
• Stopping a Meta campaign after a week because the first leads were poor, when the problem was a vague form or slow call-backs.

Warning signs that justify stopping early

• No tracked conversions after a few hundred clicks, which usually points to a tracking or landing page fault.
• Leads that keep coming from outside your service area, or from people who never wanted your service.

Pause, fix the cause, and only restart if it is something you can actually correct.

When to get help

If you are heading past about £1,000 a month in ad spend, or you cannot find time for that weekly review, a UK agency can set up and manage the accounts. Ask for transparent reporting, access to your own ad accounts, and no long lock-in contracts. A good agency will be happy to show you cost per lead and cost per job, and will not mind you leaving if it is not working.