Most companies that ask us to launch paid ads should spend their first month not launching them.
Paid media is an amplifier. It takes what your offer, your message, and your pipeline math already do, and it does more of that, faster, for money. When those three things work, ads compound them. When they do not, ads buy data nobody in the room can interpret, plus a monthly invoice nobody can defend.
This is for a founder or marketing lead at a B2B company with real revenue, a budget that is approved or close to it, and a plan to switch on Google Ads or LinkedIn this quarter. Five checks decide whether the right time to start paid ads is now, or ninety days from now.
The short answer: ads amplify, they never diagnose
Advertising platforms report clicks, costs, and conversions against the definitions you hand them. Hand them weak definitions and they return very confident numbers about nothing.
So the readiness question is not "can we afford a test." It is "do we already know enough for the test to mean something." A company with no acquisition cost baseline, no funnel numbers, no win-loss record, and no evidence that free users ever pay is not running a test when it turns on ads. It is postponing the actual work at three to ten thousand dollars a month.
I say this to prospects with money in hand, which is a strange way to run an agency and a good way to keep clients past month three.
Ready for paid ads that pay off?
Book your free auditCheck 1: You can say what one new customer is worth, out loud
You are ready when you can state three numbers without opening a spreadsheet: average deal size, the rate at which qualified conversations become closed business, and the most you are willing to pay for a qualified lead.
Work an example. A $40,000 average contract with a 25% close rate from qualified demo to won makes each qualified demo worth roughly $10,000 in expected revenue. If you want acquisition to sit near 15% of first-year revenue, you can pay about $1,500 for that demo, which at a 20% demo rate from lead means roughly $300 per lead. Now you have a number to manage against on day one instead of arguing in month four about whether $340 is good.
If those three numbers do not exist yet, the honest answer is that no cost per lead we report will be judged as good or bad. It will be judged as expensive.
Check 2: The message exists somewhere other than one person's head
The most common version of not-ready looks healthy from the outside. Revenue is fine, the pipeline is full, and almost all of it came from the founder's reputation and word of mouth. Nothing about why people buy has ever been written down, because it never had to be.
Paid search punishes that immediately. You get thirty characters for a headline that has to say what you do, match what the buyer typed, and make the wrong buyer decide not to click. A founder can improvise that on a call. An ad cannot improvise.
Before any spend, pull the language out of recorded sales calls: the sentence that makes prospects nod, the objection that always shows up, the comparison they make to the alternative they are already using. That raw material is what makes ad copy that repels the wrong buyers possible, and there is no way to generate it from a keyword tool.
Check 3: The buyer in your ads is the buyer who pays you
Product-led companies fail this one quietly. A free tier picks up individual users for years, those signups keep arriving on their own, and revenue quietly shifts to small teams who need shared workspaces and admin controls.
Those are two different buyers with two different reasons to care, and the home page usually still speaks to the first one. Ads pointed at that page will produce plenty of cheap signups from people who will never pay, the automated bidding will learn to find more of them, and the cost per paying customer will drift up while the dashboard looks fine.
Fix the target before you fund the traffic. Decide which buyer the program is for, then make sure the page a paid click lands on argues for that buyer, because the landing page is the biggest lever in paid search and it is the cheapest thing on this list to change.
Check 4: One conversion you actually trust is defined and recorded
Not five conversion actions. One, defined in business terms, firing reliably, and visible in the system your sales team maintains.
Inherited accounts are full of the opposite: site-wide page-load triggers that count every visitor as a conversion, image pixels from a form tool nobody uses anymore, lifecycle events superseded two CRMs ago, and an integration that silently disconnected in the spring. Every one of those numbers is still in the reports, and several are still feeding the bidding.
The check is small and specific. Pick the moment that means a real buyer raised their hand, confirm it fires once per real event, name it so a stranger can tell what it is, and reconcile the count against your CRM for a full month. If the counts disagree, you have found the reason paid leads appear to drop to zero long before you spend a dollar looking for it.
Check 5: Your conversion volume can survive the learning math
This is the check almost nobody runs, and it is arithmetic rather than opinion.
Google's own documentation says a bid strategy can take around 50 conversion events or three conversion cycles to calibrate after a change, where a conversion cycle is the time from click to conversion. Target ROAS has a stated eligibility requirement of at least 15 conversions in the past 30 days for Search and Shopping campaigns.
Now put a real B2B account against those numbers. Four qualified leads a month with a sixty-day lag from click to qualified means fifty events is a year of spending, and three conversion cycles is six months. The automation is not going to rescue an account like that, and neither is a target you set in the first week. Thin volume is also why value-based bidding usually has to wait even when the deal sizes genuinely vary.
Low volume does not disqualify you from advertising. It disqualifies you from expecting the platform to figure the account out. Accounts like this need a tightly scoped keyword set, an upper-funnel signal the account generates weekly rather than monthly, and a human reading search terms.
When the checks fail, buy evidence instead of impressions
The money does not have to sit idle. It buys different things for four to six weeks.
1.) Interview eight customers who bought in the last year and five prospects who chose someone else. Ask what they typed into Google, what they compared you against, and what nearly stopped them.
2.) Write the win-loss pattern down as five sentences. These become headlines later.
3.) Build one page for one buyer and one offer, with the proof visible rather than hidden behind a carousel nobody opens.
4.) Define and validate the single conversion from Check 4, and reconcile it against the CRM for thirty days.
5.) Then start, with one channel, one segment, and a number you already agreed is the ceiling for a qualified lead.
That sequence costs a fraction of a quarter of media and it ends with a campaign that can be judged. The reverse order ends with a spreadsheet of costs and an argument.
The exception worth taking seriously
There is a case where starting early is correct even with half of this unfinished. It applies when one qualified lead is worth many multiples of the monthly spend.
A niche business where a single win is worth six figures can run a small, deliberately narrow search budget and be right to do it, because two qualified conversations pay for a year of spend. Judge that account on deal value, not on lead volume, and accept that the reporting will look terrible by normal standards. Low volume with high value is a real strategy. Low volume with no value math is just spending.
What we do when a company is not ready
We say so, which costs us the retainer sometimes. Our own fit criteria say plainly that companies still hunting for product-market fit, expecting results without three or more months of investment, or hoping ads will fix churn are not a fit, and our pricing page recommends starting around $3,000 a month in media on Google or LinkedIn for a reason. Below that, a B2B account cannot generate enough signal to learn anything in a reasonable window.
The teams that pass these checks first are the ones still with us two years later, running B2B Google Ads at several times their original budget. The teams that skipped them mostly spent a quarter proving that their page did not convert.
Before your next budget meeting, write four things on one page: average deal size, close rate from qualified lead to won, the maximum you will pay for a qualified lead, and the single conversion event your sales team agrees is real. If any of the four is blank, that blank is this quarter's project, and the media budget can wait until it is filled in.
