Google Ads for roofers: get the plan right before you pay for clicks

Is paid search the right next step?

Google Search Ads can put your business in front of people searching for roofing services. You pay for clicks, not completed roofing jobs. Whether those clicks become worthwhile work depends on the searches you reach, what you offer, your website and how you handle the leads.

Before opening a new campaign, answer four questions:

• What work do we want more of? Be more specific than “roofing.” Replacement projects, smaller repairs and commercial maintenance can need different messages and different follow-through.
• Where can we do that work reliably? Include travel time and the practical limits of your crew, not just places you'd like to appear.
• Can someone handle the next lead? A busy owner on a roof needs a backup, not another phone that rings unanswered.
• What can we afford to test without relying on it to pay next week's bills?

If you're fully booked, have a broken form or still owe customers estimates, fix that first. Buying more traffic doesn't clear the backlog.

Paid search also isn't the only way to bring in work. Existing customers, referrals and better local search visibility may be a better next move. Choose the channel around your situation, rather than treating ads as a requirement for running a serious business.

First, separate Search Ads from Local Services Ads

These aren't two names for the same campaign.

Google Search Ads are the ads you can show around relevant searches. You choose campaign settings, write ads and usually send people to a website page, with calling options where appropriate. This guide focuses on that route.

Local Services Ads, often shortened to LSAs, are a separate lead-based product. Availability depends on the business category and location, and there are screening and verification requirements. Roofers appear in Google's US eligible-category list, but that doesn't confirm eligibility for your particular business or another country. Check the current eligibility information before planning around them.

Neither product guarantees that a lead will become a job. Don't mix their costs and outcomes in one report and then assume you know which one is working.

Updating an old account? Google's current transition guidance says new call-only ad creation ended in February 2026 and existing call-only ads stop receiving impressions in February 2027. For a new phone-led Search campaign, discuss responsive search ads with call assets, not a new call-only setup. Google's transition guide

Give the campaign one clear job

“We need leads” leaves too much room for interpretation. A more useful brief is:

“We want replacement-roof requests from homeowners in these service areas. We have inspection slots available, and we want people to request an estimate through this page or call during these hours.”

That is an illustrative brief, not a recommendation that every roofer advertise replacements.

A leaking roof is not the same sale as a planned replacement

Start with the customer situation, then decide how the account should be arranged.

Work you want

Those are planning categories, not four campaigns you must launch. A small residential roofer might start with one service in one genuine working area. A larger business may need separate budgets and pages for commercial and residential work.

Separate campaigns when you need different spending control, coverage, schedules or goals. Use relevant ad groups and pages to keep the message clear within that structure. Ask the person building the account why each separation is useful. More boxes in an account do not automatically mean better management.

Commercial work deserves its own conversation. The person searching may manage several properties, need another person's approval or be collecting options for a later project. A request for an assessment isn't the same as a signed contract. Record the next step and decision timing rather than forcing every lead into a same-week sale.

An example you can adapt

Imagine a roofer who wants more planned residential replacements, has inspection appointments available and does not offer out-of-hours emergency attendance. This is a fictional planning example, not a campaign result.

Their brief could be:

• Service: replacement estimates for the roof types we genuinely install.
• Area: the neighbourhoods our team can cover reliably. Confirm the property address when the lead arrives.
• Message: explain the replacement process and invite the homeowner to request an estimate. No emergency or same-day promise.
• Page: replacement project photographs, a clear description of the estimate visit, genuine credentials and a usable mobile form.
• Handoff: the office owns each request, with a named backup if the office is unavailable.
• Review: suitable replacement leads, estimates issued, decisions still open and accepted work, alongside all campaign costs.

Compare that with advertising “all roofing services, anywhere, 24/7” and hoping the right jobs come through. A clear brief gives everyone something concrete to work towards.

Write down what you don't want

This is just as useful as a keyword list:

• Services you don't offer.
• Locations you don't cover.
• Work you cannot schedule right now.
• Recruitment, training or supplier searches that aren't customer leads.
• Requests that your offer genuinely isn't designed to serve.

Keep this as a business brief first. Your ads specialist can turn it into appropriate targeting and exclusions. Don't paste every word into a negative-keyword list without checking the effect.

Match the searches, the ad and the page

Think of these as one conversation.

A person searches for help with a leaking roof. The ad describes the repair service you actually provide. The page explains that service, the genuine area you cover and how to ask about availability. The person answering the phone knows what was advertised.

When any part changes the story, the customer has to work out what you really do.

Choose searches that fit the work

Start with services customers would ask for, such as roof leak repair, roof replacement or commercial roof repair. These are examples to investigate, not a ready-made keyword list. Local demand, wording and costs need checking for your market.

Don't confuse the keywords selected in the account with the searches people actually type. Ask to see the search terms report after the campaign begins. It helps you spot searches that don't fit, though it does not expose every individual search. How Google's search terms report works

Even an “exact match” keyword can reach searches with the same meaning or intent. It does not mean Google will show the ad only when someone types those exact words. Google explains close variants.

Match types and bidding settings affect how widely a campaign can reach. There isn't one setting this guide can responsibly prescribe for every account. What matters for the owner is that the person running it can explain the choices and show whether the resulting leads fit the brief.

Use exclusions carefully

If you only want customers, a search for roofing apprenticeships probably isn't useful. But excluding “cost” could remove someone researching a replacement they intend to buy. Excluding “free” could conflict with your own free-estimate offer.

Ask for a short explanation beside each exclusion: what are we avoiding, and could this also block a good lead? Negative keyword match types behave differently from ordinary keyword matching. Have the implementation checked rather than assuming a word blocks every related phrase. Google's negative keyword guidance

Check the location option, not just the map

Choosing a city does not, by itself, tell you everything about location targeting. Google's options can include people interested in an area as well as people located there.

If your goal is to reach people in your working area, ask your ads manager to review the “Presence” option and explain any reason for a different choice. An out-of-town property owner may still be a good customer, so distinguish the person's location from the property's location when checking lead quality.

Google describes location targeting as a best-effort system, not perfect geographical fencing. State your service area clearly and verify the actual job location when someone contacts you. Google's location options

Plan around the weather and your actual diary

A roofing calendar in Calgary is not a roofing calendar in Brisbane. Local weather, daylight, the work you offer and your crew's availability all affect what you can deliver. Don't lift another country's monthly spending plan and assume it fits.

Before a busy period, ask:

• Are we trying to fill spare appointments, or is the diary already full?
• Can we deliver the work safely in the expected conditions?
• Who updates availability if weather changes the schedule?
• If demand suddenly rises, do we have permission and capacity to spend more?
• Would planned assessments for a later date be more useful than promising immediate attendance?

A storm is not an instruction to increase every bid. First check safety, service availability and whether the advertised response is still true. Don't imply that you have inspected a customer's roof, that a roof is damaged or that an insurance claim will be approved when you do not know that.

For phone-led urgent work, match the advertised calling option to real coverage. For planned replacements, an out-of-hours form can still be useful if it clearly says when someone will respond. “The office is closed” and “we cannot collect a request” are different things.

Your ads manager should distinguish ordinary seasonal planning from Google's technical “seasonality adjustment” feature. The latter is an advanced bidding tool for expected conversion-rate changes, not a switch every roofer should turn on each summer. Google's explanation

Send people to a page that helps them decide

You may already have a suitable service page. A separate landing page isn't automatically necessary.

Check that the destination answers:

1. Do you do the roofing work I need?
2. Do you cover my area?
3. Why should I trust you with this job?
4. What happens if I call or request an estimate?

Use genuine project photographs and accurate qualifications or review material you have permission to use. Show a straightforward next step on mobile without making someone hunt through a long introduction.

Keep promises aligned. Don't advertise emergency attendance, same-day visits, a free inspection or a warranty unless the business can honour the exact offer. If an assessment has a charge, make that clear before a customer commits.

Our roofing website checklist goes deeper into the page itself. You don't need to rebuild the whole website just to correct a phone number, unclear service description or awkward mobile form.

Agree the spending boundary before launch

There is no honest national figure that tells every roofer what they should spend. Local competition, the work you're advertising, the available searches and your costs all matter.

Ask for a market-specific forecast, with its assumptions visible. Treat it as a planning estimate, not a promise. Then separate:

• Money paid to Google.
• Setup or management fees.
• Landing-page work, call tracking and other required tools.
• Any applicable taxes and other charges.

Write the currency down. Agree who can change the budget, who checks spend, when they review it and what triggers a pause.

An average daily budget isn't a hard daily cap. Google says most campaigns can spend up to twice their average daily budget on a day, with a monthly limit generally based on 30.4 times that budget. Changes and campaign types can affect how limits work. Ask for the relevant limit in writing, especially for a short test, and don't assume “daily budget × test days” is a guaranteed cap. Google's budget explanation

For a new, time-limited Search campaign, ask about a campaign total budget. Google's current guidance supports a set total over 3–90 days for Search, with no daily spending limit and charges not exceeding that campaign total. It is a different budget type, not a setting you can switch an existing daily-budget campaign into. Check availability in the account and whether it suits the test before creating anything. Management, tools and other charges still sit outside that Google campaign budget. Google's campaign total budget guide

An alert can help someone notice spend; it isn't the same thing as a guaranteed stop. Agree a monitoring routine that fits the possible spend and name the person who can pause the campaign.

Also check the job economics. Sales value isn't profit. Materials, labour, subcontractors, disposal, sales costs, marketing and business overhead still have to be paid. Use your existing job and lead records rather than importing another company's claimed return into your plan.

Prove that a lead reaches a person

Before real traffic arrives, walk through each advertised contact route with your team. Mark test records clearly so they aren't counted as customers.

For a form, check the whole journey: a successful submission, the message arriving in the intended place, a usable lead record and an assigned person who knows what to do next. Test an invalid submission too. It should not appear as a successful lead.

For a phone route, check the number, the device's call action, where the call goes and what happens when nobody answers. Don't buy your own ad clicks as a testing method. Use the landing page directly and your specialist's diagnostic tools.

Then ask what the advertising report counts as a conversion. A button tap, a completed form, a connected call and a suitable customer are different events. Seeing a thank-you message doesn't prove the office received the lead. A phone-link click doesn't prove a conversation happened.

Get your specialist to verify the measurement and avoid duplicate counting across installed tools. Ask which actions the campaign uses to guide bidding, not just which numbers appear in the dashboard. Google's primary and secondary conversion settings have different roles, with an exception for custom goals. Google's conversion-action explanation

A call that lasts a minute is not automatically a roofing lead. An appointment is not automatically accepted work. Choose definitions that reflect your business, then make sure the reporting uses those definitions consistently.

Keeping job outcomes in your CRM or a spreadsheet does not automatically send them back to Google. Manual records are a useful starting point. When the records are reliable, decide whether suitable-lead or accepted-job outcomes should be imported into Google Ads, with working matching, the right conversion settings and appropriate data permissions. That is a separate setup to verify, not something a website tag proves. Google's offline-conversion setup guidance

Keep new-customer campaigns distinguishable from people already searching for your company name. Both can have a role, but mixing them can make it harder to see whether advertising is finding new demand or catching customers who already knew you.

Tracking, call recording and customer-data sharing need appropriate privacy arrangements for your location and tools. Don't switch them on just because a tutorial says to. This guide is a planning aid, not legal advice.

For the human side, use our first-response guide. A clear manual handoff is better than an automation nobody has tested.

Your final checks before spending

Keep each box unchecked until someone can show it works.

• □ The service, area and availability in the ad are accurate.
• □ The destination makes the same offer and works on mobile.
• □ A test form and test call reached the right person, including the backup route.
• □ Reporting counts the intended actions without duplicating test records or leads.
• □ The budget type, dates, other costs and spending authority are agreed.
• □ A named person can pause the campaign if something breaks or capacity changes.

If a necessary check fails, hold the launch and assign the fix. An account being ready to switch on isn't the same as your business being ready to receive its leads.

Review the test without fooling yourself

Choose a review date before launch, but allow enough time for the kind of job you're selling to reach a decision. A replacement estimate still being considered is not automatically a lost job.

At the review, follow the same group of leads from their first contact rather than comparing this month's clicks with unrelated jobs completed this month.

Bring four things:

• Actual spend and the other agreed campaign costs.
• The searches, locations and services the campaign reached, where reporting is available.
• Lead outcomes: genuine or duplicate, suitable or unsuitable, contacted, appointment arranged, estimate sent, accepted, lost or still open.
• Reasons and next steps, including anything the business has not completed yet.

A rise in calls may be useful. It is not enough to tell you that the campaign is profitable.

Use the cost and outcome fields in your existing worksheet or the launch card. For the same lead group, divide all agreed acquisition costs by the number of accepted jobs to get acquisition cost per accepted job. Keep cancellations, completed work and still-open estimates visible. With no accepted jobs, report the spend and zero wins; don't manufacture a useful average. Compare that cost with the amount your business can afford to spend acquiring that kind of job after allowing for delivery costs, overhead and profit, not with the entire selling price. If outcomes are still developing, label the comparison provisional.

What should make us pause?

Pause and investigate a broken destination, missing leads, uncontrolled spend, misleading ads or a capacity problem. These are practical failures, not something you need to tolerate while waiting for a perfect sample size.

What should make us improve the test?

If the leads are for the wrong service or area, check the searches, targeting and message. If suitable leads arrive but no one follows up, fix the handoff. If people repeatedly ask about an offer you don't provide, correct the ad or page before buying more traffic.

Change the problem you can actually demonstrate. Record the change and date so the next review has some context. Where practical, avoid changing the offer, destination and targeting at once: it becomes harder to learn which change mattered.

Don't assume every advice article's instruction to “raise bids at these hours” works with your campaign. Manual bid adjustments and automated Smart Bidding do not operate the same way. Ask your manager to check the actual strategy before changing controls. Google's bid-adjustment guidance

What should make us continue or expand?

Continue within the agreed limit when the basics work and the evidence is still developing. More spending is a separate decision, not the default response to an inconclusive result.

Consider expanding only when the actual job outcomes, costs, available capacity and handling process justify it. If too few jobs have reached an outcome, say “we don't know yet.” You can stop a test without pretending it proved that Google Ads never works.

Common questions before you start

How much should a roofing Google Ads test cost?

There isn't one useful figure for every roofer. Start with the service and area, get a current forecast, include management and other costs, and set an amount the business can afford to test. If the available budget cannot support the proposed scope, narrow the scope or choose a different channel. Don't split it across every service just to make the account look complete.

Do I need a new website?

Not necessarily. A relevant existing service page can work as the destination if the content, proof, mobile experience and contact route are sound. Fix the specific gaps first. A new page can help when the current one cannot explain the offer clearly, but it isn't a substitute for answering the leads.

Should I also run Local Services Ads or other campaign types?

Check eligibility and make that a separate decision. This guide deliberately focuses on a Search Ads test. Display, remarketing and other campaigns can have different roles, costs and measurement requirements. You don't have to launch them all to get started.

What if the report looks good but we aren't winning work?

Follow the leads rather than arguing over the headline numbers. Were they real, suitable and in your area? Did someone respond? Did an estimate go out? Is the decision still open? Was the work accepted, completed or cancelled? That trail tells you far more than a rising click count.

Should I worry about Quality Score?

It can help your ads manager diagnose parts of the account, but it isn't a business outcome. Google says the displayed Quality Score is not an auction input. Ask what needs improving for customers, and how the change will be judged, rather than treating a better score as a promise of cheaper jobs. Google's Quality Score guidance

Your roofing Search Ads brief

Copy these prompts into the document you already use with your team. You do not need a new CRM to answer them.

• The roofing service we want to promote:
• The genuine working area and important exclusions:
• The jobs we can take on, and when:
• The offer, plus promises we must not make:
• The page and phone/form routes we will use:
• The person handling leads, with a backup:
• The agreed spend, currency, other costs and budget-change authority:
• The checks we need before launch, with an owner for each:
• The outcomes we will record and the review date:
• The circumstances that make us pause:

Keep this next to the campaign report. It gives you something useful to compare the numbers against.

Want a second pair of eyes on the plan?

Bring the service you want to promote, your website and any existing campaign results. We can talk through where the plan is clear, what needs checking and whether paid search is the right next move.

Talk through your roofing ads

If you already know you want help running the campaigns, see our Google Ads service.