Est. reading time: 6 minutes
Most Google Ads budgets get set by one of three methods. Whatever last year’s number was, whatever’s left after other line items, or whatever feels survivable to the owner. All three produce the same account, one that’s simultaneously too small to let bidding stabilize and too large to be an experiment, drifting along at a spend level nobody can defend because nobody derived it. A budget worth defending comes from three inputs, the market’s actual size, your revenue, and arithmetic, and the process runs in that order.
The market sets your ceiling
You can’t outspend the search box. Your budget is bounded by the real query inventory for your keywords, in your geographies, and the first job is sizing it with Keyword Planner, Trends, and Auction Insights. If your category and geo produce 8,000 qualified searches a month, that ceiling is finite, and the goal becomes capturing a profitable share of what exists rather than budgeting for reach that doesn’t.
Competitors set the clearing price inside that ceiling. CPCs reflect auction pressure, which keeps climbing in the expensive verticals, legal, home services, SaaS, B2B tech, and the Impression Share and Top-of-Page columns tell you what participation actually costs. When incumbents hold 70 percent or more of impression share on the core terms, the honest choices are two. Bring real budget, sharp creative, and aggressive bids to break in, or specialize into the sub-niches the leaders ignore and win those outright. Splitting the difference buys expensive irrelevance.
Small and local advertisers still face a floor, and it’s worth computing before committing. Smart Bidding needs conversion volume to stabilize, and our working rule is 30 to 50 conversions per campaign per month. At a $60 target CPA, that’s a $2,400 to $3,600 monthly floor just to participate properly. If the budget can’t support that in one campaign, the answer isn’t accepting instability, it’s consolidating campaigns or narrowing geography until one campaign clears the floor. Fragmented structures on starvation budgets are where good intent goes to die quietly.
Revenue sets your guardrail
The working benchmark we hold growth-minded small businesses to is 7 to 12 percent of topline revenue invested in marketing, with Google Ads taking a serious slice wherever search intent runs high in the category. Below 5 percent is maintenance spending, defensible for a mature brand harvesting existing demand, fatal for one that wants to grow. Above 12 percent shows up legitimately in earlier stages and long-payback categories, provided the unit economics actually close.
Two common escape hatches deserve closing. Thin margins are a pricing and operations problem, and shrinking the acquisition budget treats the symptom while starving the cure. And seasonality justifies surging, not disappearing, because going dark in the off-season surrenders Quality Score history, brand recall, and the audience pools your peak depends on. A lighter year-round presence with in-season surges beats the on-off switch on both cost and momentum.
Within the guardrail, decide what Google specifically should own. In search-intent-driven categories, Google Ads can reasonably command 40 to 80 percent of the paid media line. In discovery-driven categories it takes less, but branded search protection, competitor defense, and Shopping or PMax coverage keep a non-negotiable allocation regardless, and the tactical playbook for stretching that allocation is in our 10 best Google Ads strategies for small businesses.
How to set a Google Ads budget: the arithmetic
The core equation has two terms. Target CPA times required volume equals monthly budget. A business that needs 200 leads a month at a $50 target CPA needs $10,000 a month, and there’s no clever structure that changes it. Run the result against the revenue guardrail as a sanity check, and when the two conflict, the problem is in your expectations or your pricing, not in Google.
The CPA itself gets derived, not chosen. For lead gen, back into it from unit economics. Target CAC equals LTV divided by your target LTV-to-CAC ratio, so a $900 customer LTV at a 3-to-1 target gives a $300 allowable CAC, and if one in six qualified leads closes, the target cost per lead is $50. For ecommerce, the anchor is contribution margin, where break-even ROAS equals one divided by margin. A 40 percent margin after COGS, shipping, and variable fees puts break-even at 2.5x, and a growth target lands at 3x to 4x depending on how much payback patience the business can afford. That margin-first framing is the whole argument of optimizing Google Ads for profit instead of ROAS.
Then translate the monthly number into daily budgets the learning can live on. A campaign targeting 45 conversions a month at a $60 CPA needs roughly $2,700 a month, about $90 a day, and funding it at $20 a day while expecting a stable cost per lead is asking the algorithm to learn from data you’ve decided not to buy. Where the math gets tight, the levers are consolidation of themes and SKUs, broad match run with disciplined negatives, offline conversion imports, and value signals, all of which concentrate learning instead of diluting it. Whether that learning should run on automated or manual bidding at your volume is its own decision, and we made the case in Smart Bidding vs manual bidding for small businesses.
Scaling by rule, not by adrenaline
Raising spend is where derived budgets either compound or unravel, so put rules on it. When a campaign holds its efficiency guardrails, CPA within roughly 10 percent of target or ROAS above it, for 7 to 10 days and shows impression share headroom, raise the budget 15 to 30 percent. When efficiency slips past the guardrails for a week, freeze and diagnose before spending another growth dollar, checking conversion rate, feed health, policy status, and auction pressure in that order. The full decision tree is in the budget scaling framework that actually maintains ROAS.
Sequence the expansion rather than spraying it. Saturate profitable non-brand queries and Shopping or PMax inventory in core geos and hours first, then widen match types and geographies, then layer new segments, refreshing creative and feeds every four to six weeks along the way since stale assets raise CPC while sinking CTR. And measure like the budget depends on it, because it does. Enhanced Conversions, Consent Mode v2, offline imports from the CRM, value-based bidding on predicted LTV rather than raw order value, data-driven attribution, and the patience to judge changes across two or three cycles of your conversion lag rather than by yesterday’s dashboard.
Run the sequence honestly, market ceiling, revenue guardrail, CPA times volume, rules for growth, and the budget stops being a monthly negotiation with your own nerve. It becomes a number the business derived, which is the only kind that survives a bad week, and the only kind worth scaling, which is where our budgeting strategy for scaling without overspending picks up.
