Google Ads Spending Money, No Conversions? Check This Order

"Spending with no conversions" is three problems, not one

An account that spent $2,400 last month and reports zero conversions is failing in one of three places: the conversion isn't being recorded, the clicks were never going to convert, or the site failed people who would have converted. The fixes have nothing in common, and picking the wrong one costs you another month.

The order below is not arbitrary. Each check only makes sense if the one before it came back clean. You cannot judge whether a search term is bad until you trust the conversion column. You cannot judge the landing page until you know the traffic hitting it was relevant. And you cannot sensibly change a bid strategy until you know which of the three problems you have.

Most of what ranks for this question is about a different problem — budgets that don't get spent. That's the opposite complaint. The pages that do address money going out with nothing coming back tend to arrive at bids, budgets and "test new ad copy" within three paragraphs. That advice isn't wrong, it's just last. It's the step you take after you've proved the other four are fine, and in the accounts we audit they're usually not fine.

One more thing before the list. Zero conversions on 40 clicks is not a finding. If your normal conversion rate is 2%, 40 clicks buys you an expected 0.8 conversions, and zero is an ordinary outcome. Zero on 900 clicks is a finding. Decide which one you have before you spend an afternoon on it.

Step 1: prove the conversion is being recorded

Before anything else, open Goals → Conversions → Summary and look at the status column next to each conversion action. Google tells you plainly: "Recording conversions", "No recent conversions", "Inactive", "Unverified". If your main action says "No recent conversions" and you know the phone rang last week, you have a measurement problem, not an advertising problem.

Five things go wrong here, roughly in order of how often we see them:

The action isn't set as Primary. Secondary conversion actions appear in "All conversions" but not in the "Conversions" column, and they don't feed Smart Bidding. A tracking change, a GA4 re-import or a new person in the account can flip an action to Secondary and nobody notices for six weeks. Compare the "Conversions" and "All conversions" columns side by side. If one is 0 and the other is 63, you've found it in ten seconds.

The tag fires on a page nobody reaches. The classic is a thank-you page that used to exist at /thank-you and now redirects to /thanks, or a form that submits over AJAX without a page load. Fill in your own form, on your own phone, and watch whether the conversion appears the next day.

The GA4 key event isn't imported, or was imported and then unlinked. GA4-imported conversions can keep showing historical data while quietly stopping at the import.

Consent mode is set up but modelling isn't kicking in. If a large share of your EU traffic declines cookies and you're on basic consent mode, you will genuinely lose conversions from the reports. They happened. They're not counted.

Conversion lag. Google reports conversions against the date of the click, not the date of the conversion. If your sales cycle is two weeks, the last fortnight will always look emptier than it turns out to be. Check the conversion window on the action itself before you conclude that last week was a disaster.

Here is the part that matters for anyone running automation, including us. PPCleaner never blocks a search term that converted in the last 90 days. That rule is only as good as your conversion data. If the tag broke in March, then from the software's point of view nothing converted in March, and the protection that's supposed to stop good terms being blocked has nothing to stand on. Smart Bidding has exactly the same problem, only worse, because it's also setting your bids. Fix tracking before you turn anything automatic loose on the account. We cannot fix tracking for you, and no negative keyword tool can.

Step 2: read the search terms report, sorted by cost

Once you trust the conversion column, open the search terms report for the last 90 days, add the cost column, sort descending, and read the top 50 rows. Not by impressions. Not by clicks. By cost. That list is where your money went, in order.

What you're looking for is not "low performing" terms. It's terms that were never going to convert no matter how good your ad or your site was. Three families do most of the damage:

• Job and salary intent. "plumber salary", "hvac technician jobs near me", "how to become an electrician". These people are not buying.
• DIY and information intent. "how to fix a leaking tap yourself", "boiler pressure too low fix", "what does a conveyancer do". Sometimes worth a click for a content play, almost never worth a $9 click on a lead-gen campaign.
• Competitor and retailer names you don't stock. "screwfix taps", "argos office chair", brand names that leaked in through broad match or through Performance Max.

In accounts we audit, the share of search spend sitting on terms that never converted and never plausibly would is commonly somewhere between 10% and 30%. It runs higher in accounts that lean on broad match with Smart Bidding, and higher again in Performance Max. We give a range because it genuinely varies that much; anyone quoting you a single figure is guessing.

If the top-cost rows look like the ones above, you have a traffic problem and you can stop diagnosing. Add negatives and move on. How to find wasted ad spend in Google Ads walks through the filters for pulling this list quickly at account level rather than campaign by campaign.

Two honest caveats about this report. First, it doesn't show you everything. Terms that too few people searched are withheld for privacy and rolled into an unlabelled bucket, so the visible terms are a subset of the spend. What the search terms report hides covers how much that bucket can be worth and what you can do about it. Second, the presence of junk terms doesn't prove your targeting is broken — Google's matching has got looser on purpose, and irrelevant search terms appear even on tight exact-match keywords. The question isn't whether junk appears. It's how much of the budget it's taking.

If the search terms look fine

This happens. You read 50 rows and every one of them is a person who wants what you sell, and none of them converted. That's a real result and it points you at step 4. Write it down and keep going — don't add negatives out of habit at this point, because you'd be blocking traffic that's doing its job.

Step 3: find out which campaign type actually took the money

Before you blame the Search campaign, check whether the Search campaign is where the money went. Segment the account by campaign type for the same 90 days. Three patterns show up again and again.

Performance Max is eating the budget. PMax will happily spend most of its budget on Display and YouTube placements that produce cheap clicks and no revenue, while the Search portion looks fine. The search terms it does report show up in the same place as everything else — campaign_search_term_view covers Search, Shopping and Performance Max — but the Display and Video side of PMax reports no search terms at all, because there are none. There's nothing to negate there. You control it through asset group targeting, audience signals and campaign settings, not negatives.

Where negatives do help in PMax, know the limits: a Performance Max campaign accepts up to 10,000 campaign-level negative keywords and does not accept shared negative keyword lists. That means the tidy account-wide exclusion list you maintain for Search does nothing for PMax unless it's been pushed into each campaign individually. Negative keywords in Performance Max and the 10,000 cap goes through what fits and what to do when you approach the ceiling.

Shopping is spending on queries with no buying intent. Standard Shopping does accept negative keywords and shared lists, like Search, so the mechanics are familiar. What isn't familiar is that you have no positive keywords to work against, so the negatives are doing all the steering. Shopping negatives work differently explains why the same list behaves differently there.

Display, Video or App campaigns are quietly running. These report no search terms whatsoever. If a Display campaign has been spending $30 a day since a consultant set it up in 2022, no amount of search term analysis will find it. Look at the campaign list sorted by cost, not at the report you're used to opening.

Also check two settings on your Search campaigns while you're here: whether Search Partners is on, and whether any older campaign still has Display expansion enabled. Both can move a meaningful share of spend somewhere you weren't looking.

Step 4: check the landing page — but only now

If tracking works, the search terms are good, and the money went where you intended, the problem is on your site. This is the one step nobody can automate for you.

Start with the query-to-page match rather than with page speed. Someone searching "emergency plumber leeds" at 11pm who lands on a homepage with a services carousel and a contact form below the fold is going to leave. The same person landing on a page that says "24-hour emergency plumber, Leeds, call now" with a tappable phone number converts. Pull the landing page report, sort by cost, and actually open the top ten pages on a phone.

Then the mechanical checks, in this order because that's roughly how often they're the culprit:

1. Device split. Segment by device. If desktop converts at 3% and mobile at 0.1% with 70% of the spend, your mobile experience is broken, not your keywords.
2. Does the form submit? Fill it in yourself from a real phone on mobile data. Twice. Once with a valid email, once with a slightly odd one — required-field validation that rejects + in an email address or a mobile number with spaces kills more leads than most people believe.
3. Phone number. If you take calls, is the number a click-to-call link, and is anyone picking up during the hours the ads run? Dayparting mismatches are real and invisible in the ads interface.
4. Page speed on mobile, and only after the three above.

PPCleaner does nothing about any of this. No negative keyword tool does. If step 4 is your problem, close the ads interface and go fix the site.

Step 5: bids, budget and bid strategy, last

Only now is it reasonable to change bids. And when you do, remember what a bid change actually does: it changes the price and pace of the traffic you're already buying. It does not change the composition. A campaign spending 30% of its budget on job-seeker queries at $4 a click will spend 30% of its budget on job-seeker queries at $2.60 a click after you cut bids. Same waste, slower.

Things worth checking here, once the first four steps are clean:

Is the bid strategy starved of data? Google's own guidance is that Smart Bidding needs a reasonable volume of conversions in the window to learn from. If you're getting two a month, tCPA is mostly guessing, and every change you make resets what little it had. Consider moving up the funnel to a cheaper conversion action with more volume — as long as that action is genuinely correlated with revenue, and not just a "viewed contact page" event that flatters the report.

Did you set a tCPA the market won't meet? A $15 tCPA in a vertical where clicks cost $6 means Google will find you the cheapest possible clicks, which usually means the vaguest possible queries. Aggressive targets and loose matching feed each other.

Are you changing things faster than they can be measured? If your conversion lag is ten days and you adjust bids weekly, you're reacting to incomplete data every single time.

And if you're adding negatives at this stage, add them carefully. Broad negatives look convenient and behave unpredictably across match types — negative match types do not mirror positive ones, and a single careless broad negative can silently suppress a whole profitable theme. When you're blocking at volume, phrase and exact are the safer defaults.

The 60-minute version

If step 2 is where your problem is — and in the accounts we look at it usually accounts for a meaningful slice of it — the work after the diagnosis is repetitive rather than difficult. New junk terms appear every week because matching is loose by design, so a one-off cleanup decays. That's the job PPCleaner does: read the search terms every night across Search, Shopping and Performance Max, add negatives for the ones that were never going to convert, never touch bids, budgets, ads or keywords, and keep every block reversible in one click for 90 days.

It won't fix your tag. It won't fix your landing page. And it can only act on terms Google chooses to report. But if you've worked through the five steps and the answer was "the money is going to queries that were never going to buy", that's the whole problem, and it's solvable. Just be aware that solving it means your negative list grows — how large it can safely get is its own question, and the answer is larger than most people assume.