One Google Ads click cost $335 after I fixed conversion tracking. What should I do now?

If another expensive click today would cost more than you can tolerate, pause the affected Search campaign now and confirm that it is paused. Then check which account-calendar day contains the $335 click. If its cost was billed today, it contributes to today's spend; if it belongs to an earlier day, today's daily limit starts separately. Either way, lowering the campaign's average daily budget now does not give the rest of today a new, lower billed ceiling: Google bases an edited day's billed limit on the highest budget selected that account-calendar day. Google's budget-change rule

That is the immediate decision. The forensic question—why one repair-cost click cost $335—can wait until another click cannot make the situation worse. A high price alone does not establish a bad match, an invalid click, or an error in the auction.

Protect the next dollar before diagnosing the last one

Check the account time zone, the click's account-calendar date, the campaign's budget history for today, today's billed cost and any other campaigns still spending. Google uses the account time zone for billing and daily-budget cycles. If today's highest selected average daily budget on this campaign was $200, reducing it to $60 now can affect future serving, but the ordinary billed limit for today is still based on $200: up to $400 for most campaigns. If the $335 was billed today, subtract it and any other billed cost from that $400 when assessing today's remaining exposure under this rule. If it was billed on an earlier account day, it does not use today's limit. In both cases, today's earlier $200 setting determines the limit for this campaign. The $400 is a daily billed total, not a maximum price for one click or a promise that pausing will be immediate. Google's time-zone guidance and same-day budget example

If you can tolerate some more exposure today, you could watch cost and pause at a lower trigger. Treat that as a monitored stop, not a guaranteed same-day cap. Cost reporting and the action can lag; Google says a scheduled rule can run within two hours of its scheduled time. Leave room for cost that arrives after your signal and for another click before the pause takes effect. If you cannot name an affordable buffer for that lag, leave the campaign paused. Google's data-freshness guidance and automated-rule timing

The phrase daily budget invites a perfectly reasonable mistake: thinking that a $60 setting means the day stops at $60. Google calls it an average daily budget. For most campaigns the billed daily limit can be twice that amount. It is a limit on the day's bill, not a maximum price for one click. Google's spending limits

Decide whether a bounded test can buy useful evidence

Suppose, purely as an example, you can lose $360 total on a three-day trial, and you want to run the existing eligible Search campaign for three complete account-calendar days after today. If it follows Google's ordinary daily spending-limit rule and its average daily budget stays at $60, with no higher budget selected on any of those days, the conditional billed upper bound is:

2 × $60 per day × 3 days = $360 across the trial.

In general, for a whole-test billed-loss allowance L over D later complete days, set an unchanged eligible daily budget B ≤ L ÷ (2 × D). Check the campaign's actual budget type and eligibility first, and include every campaign that can spend against the allowance. A budget edit, a partial first day, or a higher selected budget on any trial day breaks that simple bound. The account's monthly limit may also constrain spend; it does not make the daily rule a one-click stop. Google's daily limits and budget-change rule

Now ask what that protection lets you learn. Suppose you want six relevant clicks a day to inspect the bought query mix and see whether extreme prices recur—18 clicks in this three-day illustration. If all $360 were spent and all 18 clicks arrived, the test would average $20 per click. That is a feasibility yardstick, not an instruction to enter a $20 bid and certainly not a $360 guarantee from a bid limit. A $60 average daily budget might buy only about three $20 clicks on a typical day; the 2× billed limit is headroom for a high day, not a promise of six clicks every day. Real auctions can cost more or less, and a restrictive bid can keep the campaign from showing.

Do the traffic check before spending the $360. Pull recent, comparable Search-term rows for enquiries this business actually wants. Look at the middle and expensive end of their CPCs, not just one cheap click or the $335 outlier, and at how many relevant clicks those terms produced per day. Then ask whether the intended bid settings could still enter enough of those auctions and whether a $60 average daily budget could plausibly buy 18 relevant clicks within the $360 allowance. For illustration, if relevant clicks have typically cost about $30 and appeared four times a day, 18 at that price would cost about $540, while the observed volume also misses the six-a-day goal. Lowering a bid toward $20 might buy fewer impressions rather than fix either shortfall. The report omits some search terms, so use campaign and keyword traffic alongside it, and treat past CPC and volume as evidence to challenge a plan, not a forecast. Pre-repair history can help with this traffic check even though it cannot prove a lead or job rate. Google's search-terms report

Give the small trial a small job: check whether visible queries remain relevant and whether another extreme click price appears. If a genuine call or form arrives, check that it travels through the repaired tracking path once; if none arrives, the trial cannot verify that path. Set a review point before extending spend. Even 18 clicks cannot establish a stable click-to-lead rate, lead-to-job rate or account-specific Target CPA. If the existing traffic record suggests the allowance cannot buy enough relevant clicks for even this narrower check, leave the campaign paused and repair the plan or measurement first.

Spending $360 for three thin days just to complete a test is not a scientific virtue. A new eligible campaign can use a fixed-period campaign total budget as a hard billed total, but Google offers that budget type when creating a campaign; this existing daily-budget campaign cannot simply switch to it. A new setup also changes the comparison you were trying to make. Google's campaign-total-budget guidance

Keep a bid yardstick separate from profitable average CPC

A maximum-CPC bid is a different control. It limits the bid a strategy sets in eligible circumstances; it does not cap the campaign's total spend, guarantee the actual number of clicks, or convert a poor query into a good one. For Search, Google documents a maximum bid limit on portfolio Target CPA, not standard Target CPA, and warns that a tight limit can restrict optimization. If you consider Manual CPC or another capped strategy, inspect the live bidding settings and Search-partner settings before assuming the displayed maximum is an absolute per-click invoice ceiling: Google documents circumstances, including bid adjustments and manual bidding for Search partners, in which the actual CPC may exceed a stated maximum. Target CPA bid limits · Maximum-CPC exceptions

No honest calculation from “a $335 click” and “jobs sell for $150–$600” yields an account-specific $35 or $50 bid. Those tickets are revenue; they say nothing about labor, parts, overhead, repeat work or profit. The number that matters later is the advertising cost you can afford per won job.

Here is a separate, explicitly hypothetical economics example. Suppose a job can bear $120 of acquisition cost after the business accounts for its other costs and required return. Suppose half of verified leads become jobs, and one in ten relevant ad clicks becomes a verified lead. Across enough comparable traffic, the sustainable average CPC is:

$120 per won job × 0.5 jobs per lead × 0.1 leads per click = $6 per click.

The two rates must be measured for the same kind of leads, jobs and traffic. If the click-to-lead rate improves while the other inputs hold, that average rises; if fewer leads become jobs, it falls. This is a break-even-style average, not a rule that every auction above $6 is a mistake. One outlier could coexist with a profitable average if enough other clicks are cheaper and produce jobs; the outlier itself supplies no such evidence.

For the same hypothetical business, if a verified lead is the conversion action used for bidding, $120 × 0.5 = $60 per verified lead is an affordability anchor for an average Target CPA. Our guide to choosing a Target CPA works out the same affordable cost per counted lead, measuring the share of leads that become jobs only on leads old enough to have finished their sales process. If Google is instead counting every form submit, recalculate with the fraction of those submits that become jobs. Compare that business limit with the cost per actually counted action after allowing for conversion delay; with only weeks of clean data, treat the observed rate as provisional. Do not step a target down through arbitrary numbers and call the result a margin calculation. If you do not yet know the close rate, you do not yet know a defensible numeric Target CPA. The $360 trial allowance still gives you a way to decide how much you are willing to risk while you learn.

Google permits starting Target CPA without conversion history; its recommendation to evaluate performance over a period with at least 30 conversions is evaluation guidance, not a requirement to first buy 30 leads. A Target CPA is an average cost per counted conversion, not a ceiling on an individual click. If the newly counted conversion is merely a form submission, a target based on qualified leads or won jobs uses the wrong denominator. Check which actions are primary and actually enter the bidding “Conversions” column, then compare them with answered calls, real enquiries and booked jobs. Old spend and search-term history can still show which traffic was purchased; pre-repair spend cannot magically become verified lead-to-job history. Google's Target CPA guidance · Primary and secondary actions

What would explain the $335 click?

Start with the date of the click, the Search campaign and ad group, the actual search term if Google reports it, the keyword that matched, match type, network, location setting, landing page and bid strategy. The term “sliding glass door repair cost” could come from someone pricing a repair or someone who will never hire you. Its wording alone cannot decide which. The search-terms report distinguishes a person's search from your targeting keyword, although Google omits some low-volume queries for privacy. If the visible term or matching pattern repeatedly attracts services you do not offer, narrow the match or add a carefully scoped negative; if it fits your service, a high CPC may be an expensive eligible auction rather than a matching error. Google's search-terms report

Then check the lead path. Did a call arrive outside office hours? Did a form submit, and did the repaired tracking record it once as the intended primary action? Was it a qualified repair enquiry, and did it become a job later? Three clicks with no recorded lead in one day cannot establish either that tracking works end to end or that this search term never converts. Keep the click date and conversion window in view before calling the day a final zero. Google's conversion-window guidance

Finally, separate price from validity. Google says actual CPC follows auction rank and thresholds; a surprising price does not, by itself, meet its definition of invalid traffic. Check the campaign's “Invalid clicks” column and Billing adjustments for any already-filtered activity. If there are independent suspicious patterns—repeated visits, unusual devices or locations, server logs and click IDs that connect the traffic to the dates in question—collect those records and request an invalid-traffic investigation within Google's stated 60-day window. Google says it may credit invalid traffic found later, but it will not confirm the validity of a particular click in its response. A credit is possible if the traffic is found invalid; it is not owed just because $335 bought no job. Actual CPC · Invalid-traffic investigations and credits

The next decision can be made without solving that auction. Stop exposure you cannot afford today. If a later full-day test has both an acceptable billed bound and a credible chance of enough relevant clicks for a narrow learning goal, run it with verified lead tracking and a prewritten review point. If the numbers cannot meet both conditions, leave it paused. The receipt for the $335 click is evidence to investigate; it is not a bidding strategy.