Landing Page vs Homepage for Google Ads

Your ads point at your homepage because that is what exists. Here is what that costs you, what a dedicated page needs, and how to brief a developer.

By Gavin Sevastian · Updated August 11, 2026 · 10 min read

Many contractors run their ads to their homepage. Not because anyone decided to, but because the homepage is what already exists. This page is about what that costs, what a page built for the click looks like instead, and how to get one without being sold a full website rebuild.

Run This Test on Your Own Ad Before You Read Anything Else

Do this first. It takes about two minutes and it will tell you more about your own account than the rest of this page will.

Search the way a customer would, on your phone, not your desktop

Use your phone, not the computer where you check your ads. Small screen, one hand, whatever signal you happen to have standing in a driveway. Open a private or incognito tab so the results are not shaped by everything you have already clicked.

Use Google's Ad Preview and Diagnosis tool to view the result for an exact phrase you are paying for without generating an impression or clicking your own live ad. Then open the ad's final URL in a private browser on your phone.

1. 1

   Preview the phrase, not your name

   Use Google's Ad Preview and Diagnosis tool for a keyword you pay for. This lets you inspect the result without affecting your metrics.

2. 2

   Open the final URL on your phone

   Copy the final URL from the ad and open it in a private browser. Do not click your own live ad to run this test.

3. 3

   Stop and do not scroll

   Freeze when the page loads. Read only what fits on the screen. That first screen is the entire basis on which a stranger decides to stay or go back.

4. 4

   Ask the three questions

   Does this page name the service I just searched for. Does it tell me it covers my town. Can I phone someone from here with one tap.

5. 5

   Count the taps it takes to fix a no

   For every question you answered no, count how many taps it takes you to find the answer. That number is what you are asking a stranger to do for a business he has never heard of.

What to look at on the first screen after the page loads

You know your own business, so you will fill in the gaps without noticing you did it. Try to stop. Look only at what is physically on the screen, before any scrolling, and answer those three questions honestly.

Three yes answers mean the page has cleared its first-screen test. A no means the visitor has to go looking for something you already paid to put in front of him.

Count the clicks it took you to find what you searched for

If you had to tap a menu, then a service, then scroll for a phone number, that is three actions performed by someone who owes you nothing. Plenty of people will do it. The ones who do not are invisible to you. They arrive in your account as a click that cost money and produced no call, which looks identical to every other reason your ads might not be getting calls.

What a Homepage Actually Does to a Paid Click

A homepage is organized around your business, not around his question

A homepage answers the question "who are you and what do you do." A paid search click is a much narrower question: can you fix this specific thing, near me, soon. Those are different questions and they want different answers.

Your homepage is not badly built. It is built around your business: services, about, gallery, service areas, contact. The visitor has to translate his problem into your structure before he gets anywhere useful. That translation is work, and he never agreed to do it.

Every menu item is an exit, and paid traffic arrives undecided

Someone who arrives from a referral already trusts you and uses your menu to find something specific. Someone who arrives from an ad has no relationship with you at all. He is comparing you against the next three results.

Every nav item, footer link and blog teaser is a reasonable place for an undecided person to wander off to. The back button is the shortest route of all. This is not an argument that navigation is bad. It is a recognition that the homepage's definition of success (the visitor gets somewhere) is the landing page's definition of failure.

The click is already spent whether he stays or not

Nobody is being punished here. There is no switch Google flips because you pointed an ad at your homepage. The mechanism is duller and more expensive than a penalty.

You bought the visit at the moment of the click. Everything that happens afterwards only decides whether you got anything back for it. A visitor who navigates around your site for eleven seconds and leaves cost you exactly the same as the one who called.

Message Match: The Search, the Ad, and the Page Should Say the Same Thing

The immediate confirmation a visitor needs

The words he typed, the words in your ad, and the first line on your page should be recognizably about the same need. When they are, he gets immediate confirmation that he is in the right place. When they are not, he has to reconcile the gap himself, and reconciling costs attention the page has not earned yet.

The trade calls this message match. The name is not the useful part. The useful part is that the page repeats his question back to him before it tries to do anything else.

Google describes this in its own help pages, in plain language

Google's own advice is about as plain as advice gets:

Give people what they're looking for. If someone searches for "flannel shirts" and clicks your ad for "flannel shirts," the landing page they go to should feature your flannel inventory.

That is from Google's guidance on improving ad quality. The same idea appears in Google's Ad Rank documentation, which lists among the things it assesses "the expectations users have about your landing page based on the clicked ad creative." That sentence is message match, defined by the company charging you for the click.

The way contractors ruin it after reading an article like this one

The predictable overcorrection is to jam the keyword into the headline, the subhead, three paragraphs and the alt text. Google addresses this directly and says "the word-for-word phrase from a search term doesn't need to be on your landing page."

Answering the question is the job. Repeating it verbatim is not the same thing, and a page written for a keyword instead of a person reads exactly like what it is.

The First Screen: The Service, the Service Area, and a Number He Can Tap

Three things have to be visible before anyone scrolls. Each one is doing a specific job, and it is worth knowing which job, because that is what stops a developer from redesigning them away.

Name the service in his words, not the industry's words

Homeowners do not search for residential exterior envelope solutions. They search for a leaking roof, a garage door that will not close, a furnace making a noise. Your headline should use the words on the searching end, not the words on the invoice.

This is also the one element you should write yourself and hand over finished. A developer can build anything you specify. He cannot guess how your customers describe their problem, and he will default to describing your company instead.

State the service area, because it qualifies and reassures at the same time

Naming the municipalities you cover does two jobs at once. It stops the homeowner two towns over from calling and burning half your afternoon on a quote you were never going to take. It also signals that you are a local business rather than a national lead broker reselling his details, which is a real and growing suspicion.

List the places, do not gesture at them. "Serving the GTA" is weaker than naming the towns, and if you work across a spread of municipalities in the Greater Toronto Area the list itself is reassuring.

The phone number: a real tap-to-call link, not a picture of a number

Specific and checkable: a real tap-to-call link, visible without scrolling, repeated at the foot of the page, and never baked into an image. Text inside an image cannot be tapped, cannot be copied, and cannot be read aloud by a phone's accessibility settings.

For a local trade, use a local number rather than a toll-free one. Local reads local. If you want to know which ads and which pages produce which calls, that is a separate build item worth understanding before you brief anyone, and call tracking for contractors is where that decision gets made.

The Form: Ask for Less Than You Want to Know

Every field you add costs you completed forms

The form exists for the people who will not phone a stranger. Evenings, lunch breaks, anyone who wants a rough price before a conversation. Every field you add to it costs you some of them.

There is no universal percentage cost for adding a field. The practical trade-off is simpler: every question adds effort, while a phone number is often necessary if the promised next step is a callback. Keep the fields that make the handoff possible and remove the ones that merely satisfy curiosity.

The qualifying can happen on the callback, with two exceptions

The instinct is to add fields so the leads arrive pre-qualified. At contractor volumes that trade is usually bad. Roof age, property type, budget range and preferred timing are all things you can establish in ninety seconds on the phone, from someone who has already put their hand up.

Two exceptions are legitimate and worth arguing for. If you have to travel to quote, ask for a postal code or a town, because a lead outside your radius wastes a real afternoon. And if a single answer decides whether you can help at all, ask that one question. Everything else waits for the call.

Button copy, and telling him what happens after he hits it

Use button copy that says what happens next: get my quote, book my estimate, send my photos. "Submit" describes the form, not the outcome the visitor wants.

Then put one line underneath it saying who calls, from roughly what number, and how soon. That line is not a nicety. Not knowing what happens next is one of the main reasons people abandon a form they have already half filled in.

Trust Signals That Hold Up in Ontario

The real fear is letting a stranger into the house

Trust can stop a homeowner from calling before price is ever discussed. He is about to hand a person he found on the internet his address, a window of time when he will be home, and access to the thing that is broken. Every trust element on the page helps answer that concern.

Which is why badges do nothing. "Quality workmanship" and "satisfaction guaranteed" are claims anyone can type into a hero section, including the outfit that will not show up. What settles it is something he can go and check.

Put on the page what Ontario tells homeowners to go looking for

There is a shortcut here that almost nobody takes. The province publishes advice telling homeowners how to hire a contractor. Whatever it tells them to demand is exactly what your page should already be showing them, before they have to ask.

One correction worth making, because contractor pages often use vague credential language. Ontario's page on hiring a roofer states that roofing contractors are not licensed by the provincial government. A roofer should name the actual business licence, insurance, training, or credential behind a claim instead of relying on a blanket "licensed and insured" line.

The checkable version is stronger: current WSIB status where applicable, the specific training or certification held, references on request, and a written quote.

Real photos of real work beat everything else on this list

Stock photography reads as a template even to people who could not tell you why. The smiling technician with the clipboard is on four hundred other contractor sites and everyone has seen him.

You are an independent competing against franchise marketing budgets. The real photo of your real crew on a real job is the one asset the franchise cannot buy at scale and cannot fake. Use it, even if the lighting is bad.

What the Page Does to What You Pay Per Click

Landing page experience is one of three things Google grades

Google gives every keyword a Quality Score built from three components: expected clickthrough rate, ad relevance, and landing page experience. That third one Google defines as how relevant and useful your landing page is to people who click your ad.

Landing-page experience is one of the three diagnostic components, but Google does not describe them as equally weighted. Point four ad groups at one homepage and the same broad destination has to satisfy four potentially different searches.

You are graded against the other contractors bidding the same search

This is the part that answers "but our website is fine." Each component is rated above average, average or below average, and Google is specific that the rating comes from a comparison with other advertisers whose ads showed for the exact same search over the last 90 days.

Fine in isolation is not the standard being applied to you. Google compares the experience with other advertisers on the same searches. A homepage can still earn a good rating, but a page dedicated to the searched service has an easier message-matching job.

What Google actually says about price, and what it does not say

Google publishes no percentage discount for good quality. Its wording is deliberately unquantified: higher quality ads generally lead to better performance and typically cost less per click. Treat any specific percentage as a third-party estimate unless its methodology and scope are shown.

The one concrete thing Google does publish lands harder than any of the invented numbers. If your ads are low quality, you may find that your actual CPC is close to your maximum CPC even when there is low competition for the search terms that triggered your ad.

The practical point is that ad and landing-page quality can affect eligibility, position, and CPC. Google does not give you a formula that turns a page rewrite into a promised discount. What the outcome adds up to depends on your trade and your city, which is a separate question worth understanding before you set a number: what Google Ads actually cost for contractors.

Landing Page, Homepage, Service Page: What Is Actually Different

The phrase landing page gets used for three different things, which is part of why contractors end up buying the wrong one.

The difference is navigation, and what each page is trying to succeed at

Ask what each page would count as a win. A homepage wins when the visitor gets somewhere. A service page wins when the visitor understands the service. A page built for paid traffic wins when the visitor calls or fills the form, and nothing else counts as a win at all.

Those three definitions are not compatible with each other. That is the whole reason one page cannot do all three jobs properly, and it is more useful than any list of design differences.

Your existing service page is closer than your homepage, and often close enough to start

If your site already has a page about the service you are advertising, compare it with the homepage as the final URL. It still carries the full navigation and may have been written for organic search rather than for someone who just clicked an ad. Even so, it is often a closer match and may be the fastest useful improvement while a dedicated page is being built.

Do that first, then decide whether a purpose built page is worth the money. Contractors stall on a build for months while the ads keep running to the worst available page. A decent page today beats a perfect page in the spring.

What a landing page is not: a homepage with the menu deleted

The predictable misread of advice like this is to strip the navigation off an existing page and consider the job done. If the page still does not answer the search, what you have now is a homepage with no exits, which is worse for the visitor and no better for you.

Matching the search is the job. Removing the exits is the supporting move, and it only starts helping once the match is there.

How Many Pages a Contractor Actually Needs

The unit is the service theme, not the company and not the keyword

Building one dedicated page for the whole business just relocates the homepage problem to a new URL. If a single page has to serve furnace repair, air conditioning installs and duct cleaning, it becomes a menu again, and you are back where you started.

Going the other way is worse. A page per keyword produces forty near identical pages nobody can keep current, and the differences between them are invisible to a homeowner. The unit that actually works is the service theme: the group of searches that all want the same answer and the same headline.

Repair and replacement are separate themes even inside one trade. The person with something broken right now and the person planning a replacement are in different situations, and one page cannot open with both.

Start with the services you spend the most on, not the ones you like most

Open the account, sort campaigns by spend, start at the top. That is the entire method. Contractors reliably want to begin with the service they enjoy or the one with the fattest margin, and neither is where money is currently leaking.

The second sort is by what a job is worth. A theme with moderate spend producing jobs worth several times your average deserves a page ahead of one burning more on small work. What each theme is worth varies by trade, which is the ground covered in the guides on getting roofing leads, HVAC leads and garage door leads.

Where near-identical city pages stop being useful and start being a risk

The other overbuild is a page per municipality, all identical except the town name swapped in. Past a handful they stop earning anything, and Google's destination requirements specifically name content replicated from another source without adding value, and destinations solely designed to send users elsewhere.

So this can become a policy exposure as well as a performance question when the pages replicate content without adding value. If a city page cannot say something true and specific about working in that city, it is unlikely to help the visitor.

Speed, and the Phone in Somebody's Hand

Assume one hand, a weak signal, and a person standing in front of the problem

Design for the worst realistic case, because the worst realistic case is the one that pays. One hand, a phone at arm's length, a basement or a driveway with one bar, and a person already annoyed about whatever is broken.

That drives small, checkable decisions. Tap targets in the thumb zone near the bottom of the screen rather than the top corner. A headline that fits a phone screen without truncating halfway through the service name. A phone field on the form that triggers the numeric keypad instead of the alphabet.

Someone whose page never finished loading does not show up in your account as a problem. He shows up as a click. He is indistinguishable from the visitor who read everything and chose someone else, which is why this goes unnoticed for years.

The numbers to put in a specification

Be clear about one boundary, because it gets blurred constantly by people repeating each other. Those are Google's web performance thresholds. They are not Google Ads Quality Score thresholds. Google Ads help tells advertisers to improve loading speed and attaches no number to it at all.

Hit the targets because they make the page work for the person holding the phone. Do not hit them expecting a specific score to move in response.

Page weight often comes from the theme and scripts

When a contractor page loads slowly, inspect image weight, fonts, third-party scripts, and the theme underneath. A general-purpose theme can load sliders, icon fonts and animation libraries before anything readable appears on screen, but measure before deciding which part is responsible.

Worth knowing before you ask anyone to fix it, because the honest answer is often that your page is fine and the scaffolding around it is not. A dedicated page built outside that scaffolding is frequently faster than any amount of tuning the existing site would ever make it.

What a Web Developer Needs to Understand

You do not need to prescribe the component tree. You do need agreement on the page's job: which service and area it answers, which action counts as success, what real proof the business can supply, how the mobile experience will be judged, and who owns measurement after launch. Those are acceptance criteria. The layout is an implementation choice.

A vague request for “a landing page” often returns a small general website because no narrower outcome was defined. A useful brief protects message match, a direct contact path, verifiable trust and measurement without publishing a universal build recipe that ignores the contractor's offer and traffic.

The tracking, which is a separate job from the build and fails silently

This is the item most likely to be skipped, and skipping it is expensive in a way that does not announce itself. Broken conversion tracking does not throw an error. It reports zero conversions, which looks exactly like a page nobody wanted to call.

So you sit there for six weeks deciding the page failed, when the page was fine and the tag was never firing. Worse, every decision you make in those six weeks, pausing keywords, cutting budget, rewriting ads, is made on numbers that are wrong.

Treat verification as its own deliverable with a named owner. The acceptance evidence should show that each intended contact path produces the correct record after the normal reporting delay. “The tag is installed” is not the same result.

The rules Google enforces, separately from quality

There is a floor underneath everything discussed so far, and it is not about performance. It is about whether your ads are allowed to run at all.

Two of those clauses are the ones a hastily built landing page trips over. A thin page whose only real function is to push the visitor onto your main site is a destination designed to send users elsewhere. Forty near identical town pages are content replicated without adding value.

What a Landing Page Will Not Fix

The page is one part of the system, and pointing at it as the cause of everything is how contractors end up rebuilding the wrong thing twice.

A perfect page cannot rescue the wrong traffic

Lead quality starts upstream, at the keyword, the match type and the geography. If the searches bringing people in are homeowners looking for a DIY fix, students looking for a job, or people three towns outside your area, a page rewrite cannot turn them into the intended customer.

That is worth being blunt about because the fix is in a different place entirely. Negative keywords, tighter match types and honest geographic targeting are the levers. Rebuilding the page in response to junk traffic changes the wrapper on the same problem.

Landing pages and targeting discipline are complements, not substitutes. A matched page makes good traffic convert better. It does nothing at all to traffic that should never have arrived.

If nobody answers the phone, the money is spent and the job is gone

Invoca's 2026 home-services benchmark report, based on calls running through its own platform, reports answer rates across sub-industries ranging from roughly a third to about three quarters. Treat that as vendor data from its own customers rather than a neutral census. The spread is the useful part: call handling differs sharply between businesses.

This belongs in a page discussion because it changes the design, not just the operations. If you genuinely cannot answer between eight and five, the form has to be the primary path during those hours, not a courtesy at the bottom of the page.

And if you are going to miss calls, know what happens to them. A missed call with no voicemail, no callback and no text back is a paid click that produced a lead you then threw away.

If the leads are bad, the page is usually not the reason

Contractors say "the leads are bad" to mean several different things, and the fix differs for each. Wrong service, wrong area, tyre kickers, people who wanted a price and vanished, people who never answer the callback.

Work out which one you actually have before touching anything. Wrong service and wrong area are targeting. People who wanted a price are usually an offer or a qualification problem. People who never answer the callback are often a speed problem, because a lead goes cold faster than most contractors believe.

None of those are solved by a new headline. If you are still working out whether the channel makes sense for your business at all, that is the prior question, covered in whether Google Ads are worth it for contractors.

Why This Is Ongoing Work, Not a One Time Build

The page has to stay aligned with ads and keywords, and both change

The page is only matched to the search until something moves, and something always moves. You add keywords. A search term report shows people asking in words you never used. A competitor starts running a different offer and yours reads flat next to it.

Seasonality moves it hardest. A page built in April with air conditioning as the headline is the wrong page in October, and the ads pointing at it are spending money on the wrong first sentence for months before anyone notices, because nothing breaks. It just quietly does worse.

The trust content ages faster than you think

Every trust element on the page has a shelf life. Review counts and ratings drift. Photos start showing work from four years ago. Certifications need renewing. Guarantee wording changes when the business changes. WSIB says clearances are valid for up to 90 days, so a published clearance needs a recurring check.

Any one of those is a five minute job. That is exactly why none of them get done. Multiply five minutes across every service page and every season, add the ad copy that has to stay matched to them, and the five minute jobs are the job.

Most contractors do not have the traffic to test their way to an answer

Here is the honest part, and it is the reason this work leans on judgment rather than experiments.

At many single-location contractor volumes, a controlled test can take a long time to accumulate enough conversions for a stable result. Small samples can make an ordinary swing look decisive. Test when the traffic supports it, but do not treat an early lead or a short run of zeros as proof that one variant won.

That means many decisions begin with message match, usability evidence, and accumulated account data rather than a quick winner declared by a tool. If you are weighing up handling this yourself against paying someone, this is the part worth thinking hardest about: running Google Ads yourself or hiring someone.

The least expensive next decision

Compare the ad's promise with the first screen of its actual destination on a phone. If an existing service page answers that promise more directly than the homepage, it may be the right interim destination while a purpose-built page is evaluated. A new build is justified by a clearer customer path and reliable measurement, not by the label “landing page.”

If the larger issue is keeping destinations, ads and measurement aligned after launch, that is part of the Google Ads work I own for contractors.