- Fake HVAC leads and click fraud are not rare accidents. Independent data puts the average invalid-click rate on Google Ads at 11.4%, and 15 to 25 percent of internet leads carry fake or misleading information.
- You pay for the fraud either way. The only choice is whether you catch it before the charge (verification) or chase a credit after (disputes), and disputes rarely pay out.
- Recycled and incentivized leads come mostly from shared pools that resell one lead to three to five contractors.
- The cheapest fix is a verification gate that fires before you pay: reject dead numbers, spoofed calls, and bot form-fills at the door.
- At a 30 percent close rate, a junk rate of 25 percent pushes your true cost per booked job up by a third.
Here is the short answer. A fake HVAC lead is any contact you paid for that was never a real buyer in your market. That covers recycled records, incentivized form-fills, bots, spoofed calls, and clicks a competitor or a script fired at your ad. You spot it by checking the number, the timing, and the behavior before you spend a minute on it. And you stop paying for it by moving the point of detection upstream, so junk gets blocked before it bills instead of disputed after.
Most contractors treat this as the cost of buying leads. It is not. It is a tax you can measure, and most of it is recoverable if you catch it early enough.
Quick answers:
- What is the fastest way to spot a fake HVAC lead?
- How much of my HVAC ad budget is lost to click fraud?
- Can I get a refund for invalid clicks on Google Ads?
- Can I still dispute bad Local Services Ads leads?
- Are shared HVAC leads always low quality?
- What is OTP lead verification and how does it stop fraud?
- How do I calculate the real cost of unverified HVAC leads?
What a fake HVAC lead actually looks like
Fake leads leave fingerprints. Once you know them, you can score a lead in under a minute.
Start with the contact data. Watch for disconnected or wrong numbers, reserved area codes like 555, and obvious patterns like 123-456-7890. Watch for phone numbers whose area code does not match the stated city. On the email side, flag disposable services like 10minutemail or guerrillamail, and tidy patterns like firstname.lastname.numbers@domain.
Then read the behavior. A lead-fraud guide from HowToWorkLeads notes that fraudulent submissions often look too perfect: every field filled, answers matching the dropdown options exactly, and the form completed faster than a real person could type. Real homeowners leave optional fields blank and take their time.
Timing is the next tell. Large batches of leads that land at the same moment, especially outside business hours, are a classic sign of scripted submissions. The same guide reports that 15 to 25 percent of internet leads carry fraudulent or misleading information, and that some verticals run as high as 40 percent.
Incentivized leads are subtler. These are real people who filled out a form to win a gift card or enter a sweepstakes, not to fix a furnace. They answer the phone and then have no idea why you called.
The last signal is silence. A real prospect engages somewhere, by call, text, or voicemail. A dead lead never responds to anything. A practical rule: flag a lead after three calls, two emails, and one text across five days with zero response. By then you are chasing a ghost. By that same guide’s count, the average operator spends about 23 minutes per lead just trying to reach one.
Click fraud is the other half of the bill
Fake leads are what you see in your inbox. Click fraud is what drains the budget before a lead ever forms.
Google defines invalid clicks as clicks that are not the result of genuine user interest, including intentionally fraudulent traffic and accidental or duplicate clicks. In its own words, the company uses a multi-layered approach to filter invalid clicks from your reports and payments, so you are not charged for the ones it catches.
The problem is the ones it misses. According to a Search Engine Land analysis of Fraud Blocker data, the average invalid-click rate across Google Ads is 11.4 percent, up from 5.9 percent in 2010 to 12.3 percent in 2024. High-cost verticals like legal services and insurance are the primary targets, because the payoff per wasted click is higher. HVAC sits in that same high-CPC neighborhood.
Who fires these clicks? Sometimes a competitor draining your daily budget so their ad shows more. Sometimes a bot farm. Sometimes a click ring monetizing junk traffic. The intent varies. The result is the same: you pay for a visit that was never going to call.
The decision rule here is simple. If your search terms report shows spend with no calls and no form-fills from specific times, IPs, or geographies, treat that as invalid traffic and act on it. Do not average it away across a good month.
Why shared-lead pools quietly manufacture junk
If you buy leads instead of running your own ads, the fraud has a different source: the resale model itself.
Shared leads are sold to three to five contractors at roughly 25 to 50 dollars each, while exclusive leads run 75 to 200 dollars and up, per a breakdown from LeadsHook. When one homeowner is sold to five shops, four of them lose. The lead is real, but it is recycled the moment it is created. The same source describes recycled leads as having been shopped around more than a used car, and warns against anyone promising 100 percent exclusive leads for 20 dollars.
That economics is why incentivized and aged leads exist. A pool operator paid per lead has every reason to pad volume with contest sign-ups, six-month-old records, and form-fills that were never qualified. Your close rate on that pool tells the story better than any pitch.
None of this means shared leads are worthless. It means the sticker price is not the real price. A 40 dollar shared lead that reaches you fifth is not the same product as a 40 dollar exclusive lead that answers on the first ring, even though your card gets charged the same.
What unverified leads really cost at a 30% close rate
The tax only becomes real when you put it against your close rate. Here is the math on a batch of shared leads, assuming a 30 percent close rate on the real ones you can actually reach.
Say you buy 100 shared HVAC leads at 50 dollars each. That is 5,000 dollars. Now apply a junk rate, using the 15 to 25 percent fraud range from the data above, plus a worst-case 40 percent for a bad pool.
| Junk rate | Real leads you can work | Jobs booked at 30% | True cost per booked job | Spend wasted on junk |
|---|---|---|---|---|
| 0% (ideal) | 100 | 30 | $167 | $0 |
| 15% | 85 | 25 | $200 | $750 |
| 25% | 75 | 22 | $222 | $1,250 |
| 40% | 60 | 18 | $278 | $2,000 |
The sticker never moved. It was 50 dollars a lead the whole time. But at a 25 percent junk rate, your real cost per booked job jumps from 167 to 222 dollars, a third higher. You handed 1,250 dollars to leads that were never going to close.
For context on what a booked job should cost, SearchLight Digital’s 2026 HVAC benchmarks put the blended cost per lead at 104 dollars and non-branded search at 149. On non-branded campaigns, the cost per paying customer reaches 804 dollars. When your inputs are that expensive, a 25 percent junk rate is not a nuisance. It is the difference between a profitable channel and a break-even one. This is the same close-rate logic behind our exclusive vs shared HVAC lead math.
Verify before you pay: the upstream gate
Everything above points to one move. Catch the junk before it bills, not after.
Verification is how you do that. The strongest gate is OTP verification, where the lead confirms a one-time passcode on the phone number they submitted. A bot cannot pass it. A spoofed number cannot receive it. A recycled record with a dead line fails it.
That single step filters out most of the fake and unreachable contacts before they ever reach a salesperson.
For calls, the gate is duration and intent. A verified-call layer only counts a call that clears a minimum length and reaches a real conversation, so a 12-second misdial or a spam robocall never bills as a lead. Pair that with bot and spam detection on the form and the click, and you filter three fraud types at once: fake form-fills, invalid clicks, and junk calls.
This is where Elevarus operates. We run the paid channels and the verification stack together, so the leads a client pays for are already screened by OTP, call-duration thresholds, and bot detection. We are the operator running this stack, not a tool you bolt on.
The point is not the software. The point is that the filter fires before the charge, which is the only place it saves you money instead of chasing it. Attribution back to the booked job then closes the loop, which is the job of a proper HVAC call-tracking and attribution stack.
When the charge already hit: clawing back clicks and LSA leads
Verification is not always in place. When the junk already billed, you move to recovery. Just know the odds are worse downstream.
On Google Ads, start in your billing summary, where Google posts automatic Invalid Activity credits for fraud it detects on its own. For fraud it missed, an account admin can file the Click Quality Form to request an investigation. You have a 60-day window from the suspicious activity. Google wants technical evidence: IP addresses, server logs, and GCLIDs, not just a high bounce rate. FraudBlocker’s own tracking puts the manual approval rate around 10 percent, which tells you how much better it is to block clicks than to argue about them later.
Local Services Ads changed even more, and most articles have it wrong. Google eliminated the old manual dispute flow in mid-2024. As Google now explains in its automated LSA lead credits documentation, machine learning reviews charged leads and posts credits to your balance within 30 days, and you cannot appeal those decisions. Critically, Google no longer credits leads for job type not serviced or geo not serviced.
You can still influence the review. Open the lead, rate it Very dissatisfied, pick a reason like spam or wrong service, and submit within 30 days. One contractor-marketing breakdown puts credit rates on flagged leads around 15 to 25 percent.
The takeaway is not that disputes are useless. It is that recovery caps out at a fraction of what verification prevents. A dollar of fraud you block is a dollar kept. A dollar you dispute is a coin flip you usually lose.
Draw your detection line as far upstream as it goes
Every fake HVAC lead and every fraudulent click gets caught somewhere. The only question is where on the timeline you catch it, because that decides how much of your money survives.
Catch it at the click, and you never pay. Catch it at the form or the call with verification, and it never bills. Catch it in your CRM after a week of dead dials, and you have already spent the money and the labor. Catch it in a dispute, and you are begging for a credit that lands only a fraction of the time. The line moves in one direction only, and every step downstream is more expensive than the one before it.
So pick your gate by where it fires, not by what it is called. A verification layer that screens the number, the call, and the click before your card is charged is not a feature. It is the difference between paying for buyers and financing everyone else’s fraud.
If you would rather not build that stack yourself, that is exactly what our HVAC lead generation program is built to run. The broader HVAC lead generation playbook covers how the channels fit around it. When you do vet an outside vendor, judge them on the same line. Do they verify before they bill, or do they leave the fraud on your side of the ledger, the way most HVAC lead generation companies do?
Frequently Asked Questions
What is the fastest way to spot a fake HVAC lead?
Check the phone number first. Reserved area codes like 555, sequential patterns like 123-456-7890, and area codes that do not match the stated city are immediate red flags. Then look at timing and response: leads that arrive in identical batches outside business hours, or that never answer three calls, two emails, and one text over five days, are almost always junk.
How much of my HVAC ad budget is lost to click fraud?
There is no HVAC-specific figure, but the average invalid-click rate across Google Ads is 11.4 percent, and high-CPC verticals like the ones HVAC competes in are the top targets. Assume a meaningful slice of any unmonitored search budget is going to clicks that were never real buyers, and check your search terms and placement reports for spend with no engagement.
Can I get a refund for invalid clicks on Google Ads?
Sometimes. Google automatically credits invalid clicks it detects, posted as Invalid Activity credits in your billing summary. For fraud it misses, an account admin can file the Click Quality Form within 60 days with technical evidence like IPs and GCLIDs. Manual approval is uncommon, so treat disputes as a backstop, not a strategy.
Can I still dispute bad Local Services Ads leads?
Not the way you used to. Google replaced manual disputes with an automated credit system in mid-2024. It now auto-credits spam, wrong numbers, and duplicates but no longer credits job type not serviced or geo not serviced. You can still rate a lead Very dissatisfied with a reason within 30 days to trigger a review, and a minority of flagged leads get credited.
Are shared HVAC leads always low quality?
No, but they are structurally weaker than exclusive leads. A shared lead is typically sold to three to five contractors, so speed to contact matters far more, and recycled or aged leads are common in these pools. Judge a shared source by your actual close rate on it, not by the sticker price, and demand the vendor’s dead-number rate in writing.
What is OTP lead verification and how does it stop fraud?
OTP verification asks the lead to confirm a one-time passcode sent to the phone number they submitted. Bots cannot complete it, spoofed numbers cannot receive it, and dead recycled numbers fail it. Because the check happens before you pay, it filters out most fake and unreachable contacts at the door instead of after they have cost you time and money.
How do I calculate the real cost of unverified HVAC leads?
Take your spend, subtract the junk you cannot work, then divide by the jobs you actually book. Say you buy 100 shared leads at 50 dollars, with a 25 percent junk rate and a 30 percent close rate. You work 75 leads and book about 22 jobs. Your true cost per booked job is roughly 222 dollars, not the 167 the sticker implies. The higher the junk rate, the wider that gap.
