Twenty dollars a day in a category where average CPC sits near forty. Nine days later: a handful of clicks, no qualified conversations, and a firm opinion that paid search does not work for the ICP.
The channel never got a fair exam. The sample size failed first. The verdict arrived anyway.
That pattern shows up constantly. Approve a budget that cannot buy enough learning, then kill Search for reasons that had nothing to do with the auction.
This is for founders and marketing leads who want an honest budget floor before they greenlight or kill Search.
Three quiet days is not a test
Smart Bidding and honest conclusions both need conversion events. In B2B, CPCs are high and conversions are scarce. A token daily budget often produces a handful of clicks, zero qualified conversations, and a misleading dashboard.
If you need statistical comfort, you will be disappointed. B2B rarely gives clean A/B theater. What you can demand is enough spend to review search terms weekly, talk to sales about lead quality, and decide scale, refine, or pause with evidence.
From what I’ve seen, the teams that declare failure fastest are usually the ones who changed the offer, the page, and the bidding every three days while the budget was still too thin to teach anyone anything.
Size the floor from economics, not leftover software money
Work backward from reality:
- Expected CPC in your category (research auctions, do not invent optimism)
- A conversion rate you can defend, or a conservative placeholder until you have data
- Cost per opportunity you can afford given deal size and win rate
- Enough calendar time, weeks, not a long weekend, to read intent and quality
Rough planning math: if CPC is high and you need multiple qualified conversations to judge a theme, the monthly floor is not “whatever is left after software.” It is the price of learning.
If that math exceeds what you will fund, Google Ads may not be the first channel. That is a strategy decision, not a moral failing. Meta, outbound, or partnerships may be a better first motion. See when B2B should use Meta vs Google Ads.
Spend without structure just buys waste faster
More budget on broad themes, weak negatives, and form-fill optimization is not a better test. It is a faster burn.
Pair budget with intent themes and negatives you can defend, conversion tracking tied to pipeline, and a destination worthy of the click. Budget without structure is not a test of Google Ads. It is a test of how fast you can burn cash.
Enough, in practice, means you can review search terms every week with actionable volume, collect a meaningful sample of ICP conversations, see whether opportunity rate by theme is promising or poisonous, and decide without guessing from noise.
Also define kill criteria up front: what evidence pauses the program, what evidence earns more budget, and who decides. Otherwise every soft week becomes politics.
How “tests” get misused
Watch for these patterns:
- “Let’s try $20/day and see” in a high-CPC category
- Declaring failure before search-term hygiene had a chance
- Changing offer, page, and bidding every few days, then blaming the channel
- Optimizing to form fills so the test looks successful while sales rejects leads
A real test changes as few variables as possible and runs long enough for pipeline feedback, not just click feedback.
Automated bidding needs conversion density. Starving the account and expecting Target CPA magic is magical thinking. Either fund a floor that can produce the conversions you defined, or stay on simpler bidding while you gather structured learning with tight match types.
If conversions are scarce because the definition is honest (opportunities, not forms), that scarcity is a feature. Do not invent soft conversions just to unlock a bidding strategy. That reintroduces the problem from optimize for opportunities, not form fills.
Write the learning question before the first invoice
A useful test is not “turn ads on.” It is a 60–90 day learning plan with phases: build tracking, themes, and pages, then learn from search terms and lead quality, then decide scale, refine, or pause. Budget the whole arc, not week one.
Do not starve week two because week one looked quiet. Include creative and page iteration time in the plan. Media budget without destination work often fails for reasons that have nothing to do with Google’s auction.
Agree in writing before spend: primary conversion definition, budget floor and maximum, review cadence with sales, and kill or scale criteria. If those are unsigned, you do not have a test. You have a future argument.
Raise budget when themes show stable opportunity rates sales trusts, search-term waste is under control, and the constraint is volume, not quality. Do not raise budget to “give Smart Bidding more data” while conversions are still form fills sales hates.
Lower or reallocate when opportunities do not appear after a fair learning window, or when waste dominates despite hygiene. Pausing is a valid outcome. So is moving budget to a channel with better economics for your ICP.
One more trap: splitting a thin budget across too many campaigns to “test everything.” You will learn nothing about anything. Concentrate the floor on two or three sharp themes with clean tracking. Breadth is a luxury you earn after depth produces opportunities.
If procurement wants a smaller pilot, shrink scope: fewer themes, tighter geos, one offer. Do not shrink budget below the learning floor. A narrow test with enough spend teaches. A wide test with token spend only confirms noise.
One messy observation I still do not tidy up well: some categories need a higher floor than the spreadsheet suggests, because sales feedback arrives slower than clicks. The math said four weeks. Reality needed eight before anyone trusted the opportunity rate. I won’t pretend there is a universal multiplier. Build a little slack into the calendar, not just the daily budget.
Write the learning question on the brief: can Search produce qualified conversations at an acceptable cost per opportunity for ICP X? If the budget cannot answer that question, raise the floor or change the question. Do not pretend a token spend answered it.
We size B2B Google Ads programs on the strategy call from CPC and ICP, not a generic starter package. Book a free strategy call if you want a budget floor that can actually learn, plus a clear refine or pause rule before you spend.
