Why does Google Ads show more conversions than you have leads?
In a 2026 Adimize study of 321 recorded junk removal sales calls, Google Ads reported 558 conversions against 293 real job inquiries, an overcount of 1.9 times. If your Google Ads conversions run well ahead of the leads you can actually name, your reports are counting things that aren't leads, and every number built on them is off.
What we found
Google Ads reported 558 conversions over six months. Reading every call and form turned up 293 real job inquiries. Google's count was 1.9 times the real one, and every campaign was overstated.
A real inquiry here means a unique caller or form asking about a job. Repeat callers count once. Vendors, spam and existing customers don't count.
| Campaign | Google Ads conversions | Real inquiries | Overcount |
|---|---|---|---|
| Residential junk removal | 322 | 175 | 1.8× |
| Junk hauling | 182 | 88 | 2.1× |
| Specialty (hot tubs, sheds) | 27 | 17 | 1.6× |
| Brand | 18 | 7 | 2.6× |
| Dumpster rental | 7 | 5 | 1.4× |
| Cleanouts | 2 | 1 | 2.0× |
We didn't audit how the old conversions were set up, so we can't say exactly which setting caused which extra count. What we can say is that the gap was in every campaign, not one bad apple.
The study was a junk removal company, but nothing about this is specific to hauling. Plumbers, roofers, HVAC companies, cleaners, anyone running Google Ads to a phone number can have the same gap. The full study has the rest of the numbers, and the headline figures are on our statistics page.
Why it matters
The overcount matters for two reasons: it makes leads look cheap, and it teaches Google the wrong thing.
First, the cost per lead in the reports looked about half of what it really was. An owner reading those reports would think the ads were working better than they were, and might raise the budget on that basis.
Second, Google's automated bidding learns from conversions. If a wrong number counts the same as a customer booking a job, the bidding goes looking for more of whatever produced the conversions. In this account, any automated bidding was chasing the wrong signal.
And even real inquiries aren't jobs. In this study, only 14% to 23% of real inquiries became booked jobs, and ad spend per booked job was $615 to $1,045. We break that down in what a booked junk removal job costs from Google Ads.
Why it happens
The usual causes are counting every call however short, counting the same person's repeat calls, and counting clicks on the phone number as leads.
Each one is easy to set up and easy to miss:
- Every call counts. A wrong number, a hang-up, a robocall and a quick "do you take mattresses" all count the same as a booked job.
- Repeat callers count again. A customer who calls three times to reschedule shows up as three conversions.
- Phone clicks count as leads. Someone taps your number on their phone, changes their mind and never dials. Google can still count that tap.
- Existing customers count as new. A past customer who finds you through a brand search and calls about a new job is real business, but not a new lead from the ads.
Nobody sets it up this way on purpose. The defaults are generous, the reports look good, and nobody compares them with the phone.
How to audit your own
You can check this in an afternoon. Pick the last 30 days.
- Pull Google's number. In Google Ads, note total conversions for the period, and which conversion actions are counted.
- Count real people. From your call tracker and form inbox, count the unique people who asked about a job. Leave out repeat calls, vendors, spam and existing customers.
- Divide. Google's conversions over your real count. If it's close to 1, you're fine. If it's much higher, your cost per lead is fiction.
- Find the cause. Look at which conversion actions are counting: phone number clicks, calls of any length, form page views. Each one is a candidate.
What to do about it
- Count only calls of 60 seconds or longer. Short calls are mostly wrong numbers, hang-ups and quick questions. Set a 60-second minimum on call conversions.
- Count real form submissions only. A submitted form, not a page view or a button click.
- Stop counting phone number clicks as leads. A tap isn't a call.
- Track booked jobs as their own conversion. Tag booked jobs in your call tracker or CRM and send them back to Google Ads. That's the conversion that tells you what's working.
- Judge the account on cost per booked job, not cost per conversion. Divide spend by booked jobs and compare it with your average ticket.
If your conversions are clean but the jobs still aren't there, the leak is usually on the phone. See how many calls junk removal companies miss.
How we measured this
We studied six months of Google Ads calls and forms for an owner-operated junk removal company in Southern California, late March to the end of September 2026: about $42,000 in search ad spend, 327 calls, 52 forms and 321 call recordings. Each transcript was read and coded with AI assistance for job type, price quoted and how the call ended, and every call coded as booked was checked a second time. One company, one market, six months: treat it as one detailed example, not an industry average. Read the full study.
Frequently Asked Questions
Why does Google Ads show more conversions than I have leads?
Conversions often count short calls, repeat calls from the same person and clicks on a phone number. This account reported 558 conversions against 293 real inquiries.
How do I know if my conversions are inflated?
Count the unique people who asked about a job over the last 30 days and compare it with Google's conversions for the same period. If Google's number is much higher, your conversions are inflated.
What should count as a conversion for a home service business?
Calls of 60 seconds or longer and real form submissions. Track booked jobs as their own conversion and judge the account on cost per booked job.
Does an inflated conversion count hurt my ads?
It can. Google's automated bidding learns from conversions, so counting the wrong things teaches it to find more of the wrong things.
Want us to read your calls? We'll go through your last 90 days of sales calls and show you where the money leaks. Request a call audit → · Read the full study →