Google Ads Lead Form Assets: Cheap Leads, Questionable Leads, and How to Tell Which You Are Getting

A lead form asset is the easiest conversion event Google Ads has ever offered you. Someone searches, sees your ad, taps a button, fills in a name and a phone number without leaving Google, and you get a lead. No landing page to build, no page speed to worry about, no tag to install. The conversion rate looks wonderful. The cost per lead looks even better.

Then the phone calls start. A third of the numbers go nowhere. Some people don’t remember filling anything in. A few are real, but they wanted something you don’t sell. Your sales team starts to describe the leads from Google in a tone you’d rather not repeat, and you’re left holding a dashboard that says the campaign is your best performer and a team that says it isn’t.

Both of those things can be true at once. The lead form asset did exactly what you asked: it made submitting as easy as possible. The problem is that “as easy as possible” is a trade. This post is about that trade: what lead form assets are, what Google actually requires, what the quality settings do, how the leads reach you, and how to measure whether the campaign is producing customers or just contact details. I’ve drawn on Google’s own help documentation where I could find it and on practitioner write-ups where I couldn’t, and I’ll say which is which, because a lot of the numbers floating around on this topic come from vendor blogs with no method behind them.

What a lead form asset actually is

A lead form asset (previously called a lead form extension) attaches a Google-hosted form to your ad. When a person clicks the call to action, the form opens inside Google’s interface rather than sending them to your website. You choose which fields to collect, usually name, email and phone, and you can add questions of your own. When they submit, the lead lands in Google Ads and, if you’ve set it up, gets pushed on to your CRM.

Two details matter here. First, the form belongs to Google’s environment, not yours. That’s the source of both the convenience and the problem. You don’t control the page, you can’t add your own validation logic, and you can’t see anything about the visit other than what Google passes along. Second, a lead form submission is a conversion in its own right. It can be counted, it can feed bidding, and it can look fantastic in a report long before anyone has picked up the phone.

Google’s overview of the feature names Search and Performance Max as the campaign types that can use it, though some of its help pages still mention Display as well. If you run Search, the form attaches to your responsive search ads. Google’s help documentation says expanded text ads don’t qualify, which is not a live constraint for most accounts any more, since those ads can no longer be created.

If you’re deciding whether to use them at all, the useful question isn’t “do lead forms work?” It’s “what kind of person fills in a form without seeing my website, and what do I do about it?”

Who can use them, and what the rules are

This is the part of the topic where you need to check your own account rather than trust any article, including this one. The eligibility requirements have shifted, and the sources I found disagree with each other.

What is consistent across Google’s documentation: you need to be in an eligible vertical, you need a privacy policy for your business that the form links to, and you need to complete Google’s advertiser verification program. For a lead form asset to serve, the campaign needs a conversion-focused bidding strategy and has to be optimizing toward a Google lead form conversion goal. In other words, you can’t bolt a lead form onto a Manual CPC campaign that’s optimizing for clicks and expect it to behave.

What is not consistent is the spend threshold. For years the documented requirement was more than $50,000 USD in lifetime Google Ads spend. In mid-2026, several industry outlets reported that Google had removed that figure from its “About lead form assets” help page, leaving a path for reputable advertisers who have spent more than $1,000 USD in a single account, or more than $15,000 USD across accounts, and who complete advertiser verification. But the separate lead form requirements policy page was still listing the $50,000 figure when I looked, and at least one outlet noted that you can’t tell from the help center alone whether this was a text cleanup or an actual product change. If you’ve never seen the option in your account, don’t assume you’re ineligible, and don’t assume you are. Check the asset creation screen. If the option appears, you’re in. If not, the documentation won’t settle the question for you.

The practical point is that more small advertisers are likely to see this option than a year ago, which is part of why it’s worth understanding before you switch it on. A feature that used to be gated behind a large spend history is now reaching accounts that may not have the conversion tracking or the sales follow-up to use it well.

The setting that decides what kind of leads you get

When you create the asset, Google asks you to choose between optimizing for more leads or for higher-quality leads. The default is “More volume.” The alternative is “More qualified.”

Per Google’s help documentation, “More volume” uses a form design with fewer steps. That tends to produce more leads, but some of the people submitting may be less interested in the product or service. “More qualified” uses a design with more steps, which yields fewer leads who tend to be more engaged. You can switch between the two at any time during the campaign, and either choice can change both your cost per lead and the number of leads you collect.

Read that again with a skeptical eye, because it’s an unusually candid description. Google is telling you that the default setting is built to produce more submissions from people who may care less. It’s not hidden, it’s in the help text. But most people never read the help text; they accept the default because it’s the default, and then judge the format by a number that the default setting was designed to inflate.

One blog I came across claims “More qualified” cuts lead volume by 15 to 25 percent. I can’t find that in Google’s documentation, and I wouldn’t plan around it. Volume will drop, since Google says so, but by how much will depend on your offer, your audience and your form. What you can do is run the two settings against each other, which I’ll come back to in the testing section.

If you’re only going to change one thing about a lead form campaign, change this. Starting on “More qualified” and loosening it later if you’re starved for volume is almost always a safer direction than starting on “More volume” and trying to work out later why the pipeline is full of people who don’t answer.

Where the junk comes from

It helps to be specific about why lead form leads can be poor, because the causes are different and so are the fixes.

Low friction. A landing page makes people do some work. They read, they scroll, they decide the offer is relevant, and then they fill in a form. That effort is a filter. A lead form asset removes most of it. A few sources I found make this point: one argues that a user who reads your page is more committed, so the friction on a landing page screens out weaker prospects, and another warns that the ease of submitting can raise the volume of low-quality leads. This is logical, though it isn’t proven by any controlled study I could find. I’ll come back to that gap in the evidence below.

Accidental taps. On mobile, a form that opens straight from an ad and asks for very little can be completed in a couple of taps. The submission is real. The intent often isn’t.

Mismatched expectations. Your ad headline made a promise in a few words. The form is the first thing a person sees after clicking, with no page to explain what happens next or what you charge. Someone who thought they were getting a quote may be filling in a form for a consultation. Someone who wanted a free resource may not realise a salesperson will call. The mismatch shows up later, on the phone, as “I didn’t ask for that.”

Bots and bad data. Spam is a separate problem from low intent, and I’ve written about it elsewhere, in the post on spam and junk leads in Google Ads. For lead forms specifically, the point worth making is that you have fewer tools than you do on your own site. You can’t add a honeypot field or throttle submissions by IP. Whatever filtering exists has to happen through Google’s own options or after the lead reaches you.

Bidding that rewards the cheap conversion. This is the one most people miss. If the conversion goal is “form submitted” and the bid strategy is Maximize Conversions or Target CPA, the system will go and find the people most likely to submit. Who is most likely to submit a form quickly and without reading anything? Not your best customers. Smart Bidding does what you’ve asked, which is find more of the cheapest event you’ve defined. If that event is a thin signal, you’ve told the algorithm that thin signals are the target. Which brings us to the most important decision in the whole setup.

Decide what a conversion is before you decide anything else

The advice that came up again and again in everything I read, and that I agree with, is this: don’t let raw form submissions be the thing your campaign optimizes for, if you can possibly avoid it.

Google’s own guidance, as relayed in the material I found, is to optimize toward the lowest-funnel conversion you can measure, ideally the eventual sale or its value, because that’s a closer read on what a lead is worth. In practice, most advertisers can’t send final-sale data on day one, so a realistic ladder looks something like this:

• Lead submitted. The raw event. Useful for seeing volume, risky as a bidding target.
• Lead qualified. The person answered a screening question the right way, or your team marked them as a fit after a first call.
• Opportunity or booked appointment. A real commitment from a real person.
• Closed sale, ideally with a value. The event the whole thing exists to produce.

The further down that list you can optimize, the less the bidding is chasing noise. The cost is data volume: there are far fewer closed sales than submissions, and Smart Bidding needs a certain amount of signal to work with. I went through the numbers in how much conversion data Smart Bidding needs, and the short version is that there’s a tension between a good signal and a plentiful one. Where you land depends on your volume. A campaign producing thirty leads a month can’t realistically train on closed deals. A campaign producing three hundred perhaps can.

The setup most of the practitioner write-ups converged on is a sensible middle: keep raw submissions as a secondary conversion so you can still see volume, and make a qualified lead the primary goal that drives bidding. That’s a judgment, not a rule, but it’s the arrangement I’d start with.

Qualifying responses: the built-in way to get a quality signal

This is the most useful feature on the lead form asset and the one most accounts don’t use. Google calls it qualifying responses, and according to its help page it’s available for Search campaigns only.

It works like this. You add a multiple-choice question to the form and mark which answers count as qualified. When someone submits a form with a qualifying answer, Google tags the lead as qualified automatically. Qualifying questions can’t be marked optional, which makes sense, since an optional question would let anyone skip the screen. On the conversion side, this creates a “Lead form – Response qualified (Google-hosted)” action under the Qualified lead goal, which you can use for bidding.

The appeal is obvious. You get a quality signal without needing a CRM integration, and it’s available instantly, at the moment of submission. For a small business with no way of passing back sales data, it’s the only lead quality signal that arrives fast enough for Smart Bidding to use.

Where I’d be careful is in the design of the question. A qualifying question is only as good as the thing it separates. Here are the mistakes I’d expect to see:

• Questions with an obvious right answer. “Are you ready to buy now?” Everyone says yes. The question filters nothing, and the algorithm now optimizes toward people who say yes to anything.
• Questions that filter by what the lead says rather than what they are. A budget range tends to work better than a statement of intent, because it’s easier to answer honestly and harder to game by accident. A timeline works. A service type works if you only want some services.
• Too many qualifying answers. If four of the five options count as qualified, you’ve built a form that looks like a screen and behaves like an open door.

The Google Ads API documentation for the lead form asset, as summarised in the sources I read, caps custom questions at five per form. That’s plenty. Two or three well-chosen questions will usually do more than five generic ones, since each extra field costs you completions.

There’s a second thing about qualifying responses that’s easy to miss: they’re self-reported. A person who ticks the qualifying answer is claiming to be a good lead. That’s better than no signal, but it isn’t verification, and it won’t tell you whether the person ever answers the phone. Treat it as a first filter, not as the truth.

What about phone verification?

Third-party guides describe an option that sends a text message to the number a person enters, so that only valid numbers get through. I couldn’t find this feature described in the Google help pages I could reach, so I can’t tell you with confidence where it is available, whether it applies to your account, or exactly how it behaves. One guide notes it reduces the total number of leads, which is what you’d expect. If the option appears in your asset settings, it’s worth testing on a campaign where bad phone numbers are a real cost, such as when your sales team calls every lead. If you can’t find it, don’t spend time hunting for it, and put the effort into the qualifying question and the follow-up process instead.

How leads get to you, and the 30-day trap

Where do the leads go? This sounds like an administrative question, and it’s responsible for more lost money than any bidding setting.

According to Google’s help documentation, you can get leads in three ways: download them as a CSV, receive them through a webhook, or pull them through the Google Ads API. A webhook sends each lead to your CRM or another system when it arrives. The API route lets you export lead data in bulk.

The CSV download is where the trouble sits. The help pages I could find say leads available for manual download cover the last 30 days. If nobody is downloading regularly, leads disappear from the download screen. The retention numbers are not completely consistent across sources: some newer documentation says Google stores lead data for 60 days, and API exports can reach up to 60 days, while the manual download window is 30. So the safe reading is that 30 days is your practical limit for a human with a CSV and 60 days is the outer bound for automated access. Neither is a place to leave leads.

What this means in practice:

• If you’re using CSV downloads, someone has to own that task and do it on a schedule. A lead that sits for two weeks before anyone calls is a lead you’ve mostly paid for in vain.
• If you’re using a webhook, check that it’s actually delivering. A webhook can fail without anyone noticing (a changed endpoint, a renamed field, an expired credential), and if it does, Google Ads will keep showing leads on its end while your CRM sees nothing. I also found a reference to a limit of one webhook delivery method per form, so wiring it up twice won’t get you a second copy.
• If a human is manually forwarding leads, you’ll lose some. Not as a matter of principle; it’s just what happens when speed depends on one person’s inbox.

Speed matters more with lead forms than with most lead sources, for a reason that’s easy to forget: the person didn’t visit your site, so they have no relationship with you. They tapped a button in an ad. If your call comes three days later, they may not remember doing it. A call within the hour comes from the business they just saw. A call the next week comes from a stranger.

Connecting the lead back to the click

Everything above gets you a lead that is better filtered and gets to you faster. The step that actually changes how the campaign performs over time is telling Google what happened to the lead afterward.

That’s offline conversion import, and for lead forms there’s an additional flavour called enhanced conversions for leads, which uses hashed customer details such as an email address to supplement imported offline data. Google’s API documentation describes it as an upgraded version of offline conversion import and recommends starting with it, or upgrading to it. I covered the mechanics, and the 2026 changes that could stop lead data from reaching Smart Bidding, in the post on offline conversion imports. If you want the underlying tracking to be sound first, how enhanced conversions fail silently is worth reading before you build anything on top.

Two caveats from what I found, both of which match what people tend to hit in practice:

• One CRM vendor’s help page names a mismatch between the GCLID Google generates and the one pushed to the CRM as a common cause of failed imports. If your CRM doesn’t store the click identifier, or truncates it, the sale can’t be matched back to the ad. That is a problem to find before you have a year of leads with nothing to match.
• Google’s API documentation notes that imported conversions are reported against the date of the original click’s impression, and can take time to show up. That makes the numbers look wrong at first: last month’s campaign shows more conversions this week than it did last week. It’s not a bug. It’s the reporting model.

The part about lag is where many advertisers lose their nerve and switch strategies. If a sale takes six weeks, a campaign you judge at two weeks will always look worse than it is. I wrote about the general version of this in conversion lag, and it applies doubly when your best signal arrives late.

If tracking is the part you suspect is shaky, auditing it comes before tuning the lead form. The walkthrough in why your conversion numbers don’t add up covers the usual failure points.

Lead form or landing page?

The honest answer is that nobody has published a clean, independent comparison, and you should be suspicious of anyone who quotes you one.

When I went looking for data on lead forms versus landing pages, I found vendor blogs. Several say lead forms convert better and cost less per lead. One puts the saving at 30 to 60 percent. Another reports that leads from the form have a 20 to 40 percent lower contact and qualification rate than leads from landing pages. A third claims form completion rates of 70 to 85 percent versus 15 to 30 percent on landing pages, though that source was describing native forms on Meta, LinkedIn and TikTok rather than Google, and it didn’t explain how it measured anything. None of these figures come with a method I can check. I’d treat them as the shape of an argument, not as numbers.

The shape is still informative. The consistent claim is that lead forms produce cheaper leads and worse leads, and that the question that matters is the cost per qualified lead or per opportunity, not the cost per lead. That’s a claim I’d bet on even without the numbers, because it follows from the mechanism. Remove friction and you get more of everything, including the people who weren’t going to buy.

Who should lean toward which?

• Lead forms suit simple, low-consideration offers. A quote request for a service with a clear price range, a callback for an appointment, a free estimate. The decision is easy and the form is short.
• Landing pages suit complex or high-consideration sales. If someone needs to read about what you do, compare options, or understand pricing before they’re a viable lead, a form that skips all that will mostly collect people who haven’t done the thinking. A B2B software demo is the classic example. If that’s your situation, our page on Google Ads management for B2B and SaaS is the closer fit.
• Local services sit in the middle. Plumbers, electricians and similar businesses often win on speed of response, and a short form fits that. The caveat is that a missed or late call wastes the lead completely. If that’s your business, the local service businesses page is the one to look at.

Several of the sources I read recommend running both: a lead form to broaden the top of the funnel and a dedicated landing page for the campaigns where you need people to read before they raise a hand. That seems right to me. The mistake is treating it as a choice, when you can run them side by side and compare what each produces.

How to test it without fooling yourself

The biggest risk with lead forms is a flattering comparison. The cost per lead is lower, so it must be better. Here’s a way to avoid that trap.

Define the success measure first. Pick one number you’ll judge on, such as cost per qualified lead, cost per booked appointment, or cost per sale, and write it down before you look at any results. If you pick it after seeing the data, you’ll pick whichever number makes your preferred option look good.

Track the same outcome for both. If landing page leads are marked qualified by your sales team after a call and form leads are marked qualified by a form answer, you’re comparing two different definitions. Use the same definition, applied at the same point, for both.

Split volume cleanly. Google Ads has an experiments feature built for exactly this kind of comparison, and the post on Google Ads experiments goes through how to run one without contaminating it. Running a lead form campaign and a landing page campaign against the same keywords at the same time, with no cleaner split, means they compete in the same auctions, and the results blur.

Compare More volume against More qualified. Since you can switch at any time, you can run the two settings in sequence. The risk with switching back and forth is that the bidding system has to relearn each time, and a sequence confuses cause and effect. If you have the traffic, a proper split is better than a before-and-after.

Wait for the outcome, not the form. Whatever you’re comparing, give it long enough for the downstream outcome to show up. If your sales cycle is a month, a two-week test tells you about submissions and nothing about customers.

If volume is too low to test at all, you may be better off picking the more cautious configuration and watching the downstream numbers than pretending a small test is conclusive. Not every account can run a clean experiment, and acknowledging that is better than running a bad one.

Settings that quietly make things worse

A few specifics worth checking on any lead form campaign you inherit or set up.

Automatically created assets. One of the sources I read recommends turning this off for lead generation, on the grounds that it can encourage low-intent submissions. I can’t verify the mechanism from Google’s documentation, but the principle stands regardless: for an account where lead quality is the problem, you want fewer places where Google decides how your offer is presented, not more.

Call to action wording. The call to action button is the first thing a person sees. Phrases that promise something free or instant (“Get Quote”, “Download”) draw different people than phrases that describe a consultation or booking. Match the button to what the lead will actually receive. If a salesperson is going to call, a button that implies a PDF is a setup for a bad conversation.

The thank-you message. What you say after submission shapes what happens next. Tell people what’s coming: who will contact them, how, and roughly when. This does little for conversion numbers and quite a lot for contact rate, because people who know to expect a call from an unfamiliar number are more likely to answer it. Third-party guides also mention adding a call to action to the post-submit message, such as visiting your website. I couldn’t confirm Google’s rules for that message in the official material, so check the asset editor for what’s allowed.

Search terms. Lead form leads still come from queries, and the queries matter. A lead form attached to a broad campaign will happily collect submissions from searches that have nothing to do with what you sell. Review the search terms behind your leads regularly, using the approach in how to read a search term report, and add negatives for the ones that produce leads you wouldn’t want.

Segment before you conclude. One source recommends breaking leads down by campaign, device, location and hour of day before changing anything. I’d add day of week. It’s common to find that most of the junk comes from a narrow slice, such as late-night mobile submissions in one region, and that cutting that slice fixes more than rebuilding the form.

What to watch week to week

A lead form campaign is not something you set up and leave. The failure is rarely a sudden collapse. It’s a gradual drift: submission volume holds, qualified rate slips, the follow-up gets slower, and nobody notices until the quarter ends.

These are the numbers worth looking at each week, in roughly this order of importance:

• Valid lead rate. What share of leads have a working phone number and a real person behind them. A fall here suggests spam or accidental submissions.
• Contact rate. What share of leads your team actually reached. A fall here often means slow follow-up, not worse leads.
• Qualified rate. What share of reached leads were a genuine fit. A fall here points to targeting, the form, or the offer.
• Cost per qualified lead. The number you really care about, which will often differ a lot from the cost per lead in the main column.
• Time to first contact. Not a Google Ads metric, but the one most likely to explain a weak result.

Watching a split like that is also a good way to see where the problem sits. If valid rate is fine but contact rate is dropping, look at your process, not your campaign. If valid rate is dropping, look at traffic and form settings. Keeping those categories separate stops you from rebuilding an ad campaign that was fine because the real problem was a webhook that stopped working on a Tuesday.

Where continuous monitoring fits

Most of what’s in this post is checking things that drift: a webhook that stops delivering, a conversion action that quietly changes status, a qualified rate that slides over a few weeks, a campaign that starts spending on queries it didn’t match last month. These aren’t dramatic failures, and they don’t announce themselves. They’re the kind of thing that gets noticed late, because nobody looks at them daily. This is the reason we built Growera’s Google Ads manager to watch the account continuously rather than on a monthly review cycle: not to make decisions for you, but so the slow changes get flagged while they’re still small enough to fix cheaply.

A sensible starting configuration

If you’re setting up a lead form for the first time, and your offer is simple enough to suit one, here’s a setup I’d be comfortable defending. It isn’t the only reasonable one.

• Check that you’re eligible and that the asset is available to you, rather than relying on any article’s reading of the spend rules.
• Start on “More qualified,” and move to “More volume” only if volume becomes a real problem and the downstream numbers support it.
• Add one or two qualifying questions that actually separate good leads from bad ones, using budget, timeline or service type rather than a question everyone answers yes to.
• Make the qualified lead, not the raw submission, the primary conversion goal, and keep submissions as a secondary measure.
• Wire up a webhook to your CRM, then test it with a real submission, and check it again a week later.
• Make sure your CRM stores the click identifier, so sales can be imported back later.
• Set a response target for your team, and measure it.
• Review search terms and segment the leads weekly for the first month.

Not all of this is glamorous, and none of it is complicated. It’s the part of the job that takes attention rather than expertise.

Summary

Lead form assets are a trade of friction for volume. Google’s own documentation says so: the default “More volume” setting uses fewer steps and may bring in people who are less interested, while “More qualified” adds steps and gives you fewer, more engaged leads. Neither is the right answer in general. Which one suits you depends on how simple your offer is and how well you can tell good leads from bad ones afterward.

The decision that matters most is what you tell bidding to optimize for. A raw submission is the weakest signal available, and a bid strategy built on it will go looking for the people who submit most easily. Qualifying responses, available on Search campaigns, give you a quality signal at the moment of submission. Offline conversion import and enhanced conversions for leads give you a better one later. Use as much of that as your volume supports.

Get leads out of Google quickly and reliably: a webhook you’ve tested beats a CSV someone forgets to download, particularly given the 30-day limit on manual downloads. Then follow up fast, since the person has no relationship with you beyond one tap on an ad.

Finally, be wary of any comparison between lead forms and landing pages that doesn’t measure cost per qualified lead, including every figure I quoted above that came from a vendor blog. The numbers that matter are in your own account, in the gap between leads submitted and customers acquired. Measure that, and the right setup becomes much easier to see.

If you want to see what continuous monitoring looks like in practice, the pricing page and the ERA AI overview are the places to start, and 5 signs your account is quietly wasting money is a good companion read if you suspect lead quality is only part of the problem.

Sources referenced: Google Ads Help pages on lead form assets, qualifying responses and lead delivery; Search Engine Land’s report on the removed spend requirement; and practitioner write-ups from LeadsBridge, Clixtell and others, treated as unverified where noted. Google changes these features often, so confirm the details in your own account.