Here is a conversation I have about once a month. An owner has one channel that works, almost always Google Search. The phone rings. The jobs are profitable. So they do the obvious thing and raise the budget, and for a while it works: more spend, more jobs. Then something feels off. The account still looks healthy on the summary screen, but the business does not feel a third bigger, even though the ad bill is.
At that point they get two kinds of advice. Half the internet says scale what works and stop getting distracted. The other half says never depend on one channel and get onto Meta, YouTube and everything else. Both camps are half right, which makes them both wrong as a rule.
The useful question is narrower than either. Is Google Ads actually full, or is it just underfed, held back by something you can still fix? Those two situations look identical from the outside and call for opposite decisions. One says add a second channel. The other says do not touch a second channel yet, because the first one still has room you are not using.
You can tell them apart. It takes about twenty minutes in your account and one piece of arithmetic most reports never show you.
Your Average Cost per Job Looks Fine Right Up Until Google Ads Runs Out of Searches
Every ad platform reports averages. Cost per conversion, cost per lead, return on ad spend: all of them divide everything you spent by everything you got. That is a perfectly good way to grade last month. It is a terrible way to decide what to do with next month's extra money, because the extra money does not buy average customers. It buys the next ones in line, and the next ones in line are always more expensive than the ones before.
Here is how that plays out. This is an illustration, not a client account, but the shape is one we see constantly. A heating and air company spends $3,000 a month on Google Search and books 30 jobs. That is $100 a job.
They raise the budget to $4,500 and book 38 jobs. The report now says $118 a job. Up a little, nothing alarming. But look at what the extra money did on its own: the additional $1,500 bought 8 jobs. That slice cost $187.50 a job, almost double the original rate.
They raise it again to $6,000 and book 42. The report says $143 a job. Still a number most owners would happily take. The last $1,500, though, bought 4 jobs. That slice cost $375 a job.
Now set that against what a job is worth. Say the average job leaves $300 of gross profit after parts and labor. The account as a whole is comfortably profitable at $143 a job. The last $1,500 is losing $75 on every job it books. The average hides it completely, because the cheap jobs at the bottom of the budget are subsidizing the expensive ones at the top.
This is the real ceiling, and it is worth being precise about. It is not the point where your return on ad spend starts falling. That falls the moment you spend more, almost always, and a falling ratio can still mean more profit in total if every added dollar is still above break-even. The ceiling is the point where an added dollar stops adding profit. Below it, more budget is good business even when the ratio dips. Above it, more budget is a donation to Google.
You can work out your own version without any special tools. Pull two comparable months at different budget levels, subtract the lower month from the higher one, and divide the extra spend by the extra jobs. If your sales cycle is longer than a few weeks, compare jobs closed rather than leads, and leave a cycle's lag. Two months is a rough read, not a lab result, but rough is plenty to tell $187 from $375.
Lost Impression Share Tells You Whether You Ran Out of Money or Ran Out of Searches
A rising cost on the last dollar tells you something is limiting the account. It does not tell you what. There are three different limits, they feel the same from your chair, and only one of them is a reason to add Meta.
Google gives you the diagnosis in two columns that are not switched on by default. In the campaigns view, add Search impression share, Search lost IS (budget) and Search lost IS (rank). Google defines impression share as the impressions you got divided by the impressions you were eligible for, and the two "lost" columns split what is missing into two causes. In Google's words, lost IS (budget) is "the percentage of time that your ads weren't shown on the Search Network due to insufficient budget," and lost IS (rank) is the percentage of time they were not shown "due to poor Ad Rank in the auction."
Read them over the same months you used for the arithmetic, and one of three pictures appears.
You ran out of money. Lost IS (budget) is high, say 30 percent or more. Your ads are qualified to show and are being switched off partway through the day because the budget is spent. That is not a ceiling. There are still searchers out there that your ads qualify for. Before you fund a new channel, ask why this one is rationed, and test whether more budget here buys jobs at a cost you would accept.
You ran out of rank. Lost IS (rank) is high. Plenty of searches, but your ads are losing the auction to competitors with a better combination of bid and quality. Ad Rank weighs your bid together with the quality of the ad and the landing page, so the fix lives in the ads, the page and sometimes the bid, not in a new platform. We have written about how Google decides who gets the top spot if you want the mechanics.
You ran out of searches. Impression share is high, both lost columns are small, and the extra money is mostly buying pricier clicks on the same searches you were already winning. This is the true ceiling. There are only so many people in your service area typing "furnace repair" this week, and you are already in front of most of them. Spending more cannot create searches. It can only bid harder for the ones you already get.
A word of caution Google itself gives: impression share is an estimate, and small swings from week to week do not mean anything needs doing. You are looking for a pattern across a couple of months, not reacting to a Tuesday.
If Meta is the channel that is working instead of Google, the same question has a different signal. Meta does not run out of searches, it runs out of new people, and the number that shows it is frequency, which Meta defines as "the average number of times each person saw your ad." Frequency climbing while cost per lead climbs on an audience that is not growing means you are paying to show the same people the same message again. We go through what to do about that on each platform in the Meta scaling piece and the LinkedIn version. The decision underneath is the same: underfed, outranked, or out of people.
Most Owners Who Think Google Ads Has Maxed Out Just Have a Weak Offer
Before we get to adding Meta, the honest counter-argument deserves a fair hearing, because for most small businesses it wins.
The case against a second channel goes like this. You have a finite amount of money and a finite amount of attention. Split them across two channels and both get less of each. Every new channel needs its own ads, its own landing page and its own learning period, and most of the businesses that run four channels at once run all four badly. One growth consultant with a widely shared video on this puts it bluntly: he has seen a single acquisition channel carry a business into seven figures, and his advice to companies under about half a million in revenue is to stay the course and push harder on whatever is already working. We do not share every part of his model, but on that point the evidence is on his side.
The reason is that a demand ceiling is rarer than it feels. Most accounts that look maxed out are actually limited by something upstream of the ad platform, and those limits move the ceiling up when you fix them:
- The offer. If the ad promises the same thing as every competitor's ad, you win the auction only on price and position. A stronger reason to call, a guarantee, a faster appointment, a clear price range, raises the share of searchers who pick you. Same searches, more jobs.
- The landing page. A page that converts 4 percent of visitors instead of 8 percent doubles your cost per job, and no amount of budget fixes that. Every improvement here lowers the cost of every dollar, including the last one.
- The ads themselves. Better ads raise Ad Rank, which is precisely what fixes a high lost IS (rank).
- The service list and the map. If you only advertise one service, or one city, you have not hit the ceiling for the business. You have hit it for one product in one place. A new service line or a neighboring town is often a cheaper second step than a new platform.
So here is the test, and it is a strict one. Write down the fixes you have not tried yet on the channel that works. If the list has three real items on it, the next dollar belongs to those fixes, not to Meta. If you genuinely cannot name any, and the impression share columns say you are already in front of most of the searches, you are at the ceiling. That is when the second channel stops being a distraction and becomes the only way to grow.
Meta Earns Its Budget by Reaching the People Google Ads Never Sees
The most common way a second channel fails is that nobody decided what it was for. It gets launched with the same goal as the first channel, gets judged on the same numbers, loses the comparison, and gets switched off after six weeks.
The marketing manager of a large Ohio landscaping company recently described the rule she uses when she plans the year, and it is the best one we have heard: every channel has to have one clear purpose. Can it bring in design-build leads? Maintenance leads? Lawn care? If a channel's purpose is no longer clear, she cuts it or shrinks it. Her other habit is worth stealing too. She runs paid social in the same ZIP codes as her postcard drops and treats the overlap as reinforcement, not waste.
Applied to Google and Meta, the purposes nearly write themselves, because the two platforms reach different moments in a customer's life. Google Search catches people who already know they need you. They typed the problem in. Meta reaches people who will need you but are not looking yet: the homeowner whose furnace is twelve years old, the family that just moved in down the street. We cover this split in more depth in which ad channel to start with.
That difference is exactly why Meta is the natural second channel once search runs out. You have already caught nearly everyone who is looking. The only way to grow is to reach people before they look, so that when the furnace finally fails, your name is the one they search for.
Before any money moves, write Meta's purpose down in one sentence. Not "brand awareness," which is a goal nobody can grade. Something like: "Meta keeps us in front of homeowners in our five ZIP codes with older homes, so they call us before they search." A sentence like that tells you who to target, what the ad says, and what success looks like. If you cannot write it, you are not ready to add the channel.
This is also where our piece on running two channels well comes in. Its test for adding a channel is that you have exhausted the ones you have. Everything above is how you actually check that, with numbers rather than a feeling.
The Budget for Meta Is Already Sitting in Your Most Expensive Google Ads Jobs
Once you have decided to add Meta, there is a funding question, and the instinctive answer is the wrong one. The instinct is to split: take the $6,000 and send half to Meta. That cuts Google back to $3,000 and throws away the 8 jobs at $187.50 along with the 4 at $375. You have given up good customers to fund an experiment.
The arithmetic from the first section points to a better move. The cheapest money to take away from Google is the money that was buying the most expensive jobs. In our example, the last $1,500 was booking jobs at $375 against $300 of profit. Taking it out of Google loses about 4 jobs that were losing money anyway. Put that $1,500 into Meta and you have funded the second channel at almost no cost to the first.
Two honest qualifications. First, when you lower a budget, Google's bidding decides which auctions to drop, and you will not get a perfect reverse of the climb. Watch it for a few weeks and confirm the cost per job at the lower budget settles where your arithmetic said it would. Second, if the top slice of your budget is still profitable, then you are not funding Meta out of waste, you are funding it out of good money, and the honest framing is that Meta has to beat that slice to deserve it.
Then fund it properly. A new channel needs a full runway before its numbers mean anything, usually around ninety days, for reasons we lay out in how long ads take to work. Six weeks of Meta is not a smaller test. It is a test that ends just before the answer arrives. If the slice you can move only covers a month, either find more or wait.
Last-Click Reports Will Undersell Meta for Months After You Add It
The final trap is in how you judge the result. Most reports give the whole credit for a sale to the last thing someone clicked. That suits Google, because a search is almost always the last step. It is brutal on Meta, because Meta's whole purpose is the first step.
Here is the sequence that gets misread. A homeowner sees your Meta ad in March, does nothing, and in May, when the AC dies, types your company's name into Google and calls. Your Google report records a new customer from a branded search. Meta records nothing, or a view it cannot prove mattered. A B2B agency podcast made the same point recently: when owners grade every channel on the last click, they end up cutting exactly the channels that introduce new customers and keeping the ones that harvest them.
So judge the pair, not the newcomer. Before launch, write down three numbers for the previous three months: total jobs from all sources, total ad spend across both channels, and how many searches for your business name show up in your Google Ads search terms report or in Google Search Console. Then after ninety days, ask:
- Did total jobs go up by more than Meta cost to buy them?
- Did the combined cost per job, all spend divided by all jobs, hold or improve?
- Did searches for your name rise in the areas Meta covers?
If the answers are yes, Meta is working, whatever its own dashboard says. If total jobs did not move, it is not, however good the in-platform numbers look. We go deeper on why no single report can settle this in You Will Never Know Exactly Which Ad Made the Sale. Set the keep-or-kill rule before the money goes out, while you can still be objective about it.
A Maxed-Out Google Ads Account Is Good News. It Means Search Worked.
If you have read this far and your numbers say you are at the ceiling, that is not a problem. It means the first channel worked well enough to be fully used, which most businesses never get to. The ceiling is just the point where the next step changes, from spending more on the same searches to reaching people before they search.
Most of the time, though, the numbers will say something else: an underfed budget, a rank problem, or an offer that is doing less than it could. Those are cheaper to fix than a new channel and they raise the ceiling for whatever you add later.
This is the kind of decision we work through with owners every week, and if you would rather have someone read the columns with you, we are happy to. If you would rather do it yourself, everything you need is in your account right now. Pull the two months, do the subtraction, and look at what your last dollar actually bought.
