The short answer: your Google Ads budget isn’t a number someone recommends — it’s a calculation from four inputs you mostly already have: what a customer is worth to you, how often enquiries become customers, what clicks cost in your auction, and how much data you need before numbers mean anything. This guide walks the arithmetic with worked examples, then covers the part most guides skip: how to know when your budget is wrong — too small to learn from, or scaled past its ceiling.
One caveat before any numbers: click prices, conversion rates and seasonality vary enormously by industry, geography and offer. Every figure below is illustrative arithmetic, not a promise about your market. The method is the takeaway; the audit fills in your actual values.
The four inputs
1. Customer value (the ceiling-setter)
Not revenue — margin, and ideally lifetime. A dental clinic’s new patient might bring RM300 on the first visit and several thousand over years of care; a renovation contractor might net five figures from one job. Write down: gross margin per new customer, first transaction and 12-month view.
2. Close rate (the honesty check)
Of the enquiries you receive — calls, WhatsApp messages, forms — what fraction become paying customers? Most SMEs guess high; check your last 30 enquiries if you can. Write down: enquiries-to-customer percentage.
These two produce your maximum affordable cost per enquiry:
Max CPL = customer margin × close rate
Example: RM1,200 margin × 25% close rate = RM300 per enquiry, break-even.
You don’t want break-even — take a working target at half to two-thirds of the maximum, leaving room for profit and imperfect months.
3. Auction prices (the reality tax)
What a click costs for your keywords, in your locations. Google’s Keyword Planner gives free directional estimates; a week of live data beats it. Malaysian context, directionally: consumer-service clicks in the Klang Valley commonly run from under RM2 (long-tail, BM-language, suburban) to RM15–30+ (legal, aesthetics, property, finance-adjacent head terms). Secondary cities and specific niches can sit well below that; a few run far above.
4. Conversion rate (the multiplier)
The share of ad clicks that become enquiries — a property of your landing page more than your ads. Cold-traffic service pages commonly land between 2% and 10% depending on intent-match and page quality. Until you have data, run scenarios at 2%, 4% and 6%; after a month you’ll know. (This number is also the cheapest one to improve — that’s CRO.)
The worked example
A Klang Valley aircond services company:
- Average job margin: RM350; close rate on enquiries: 40% → max CPL RM140, working target ~RM70
- Auction estimate for their service terms: RM4/click; landing page assumption: 4% conversion
Cost per enquiry = RM4 ÷ 4% = RM100 — above the RM70 target: the plan is viable but needs the optimisation phase to close the gap (better search-term pruning, better page, better match).
Now size for learning: to judge anything you need volume. 30 enquiries in a month is a reasonable floor for early decisions → 30 × RM100 = RM3,000/month media budget to start, before management.
Notice what the arithmetic just did: it turned “what should we spend?” into a falsifiable plan — RM3,000 buys roughly 750 clicks and ~30 enquiries at assumed rates, and month one exists to replace assumptions with data.
The minimum-data rule (why tiny budgets fail)
The most common Google Ads failure isn’t overspending — it’s spending too little to ever learn. A budget producing three enquiries a month cannot distinguish luck from strategy; every decision becomes a coin-flip narrated confidently.
Rules of thumb that hold up:
- Budget should buy at least 20–30 conversions a month on your primary goal — or concentrate until it does: fewer keywords, one suburb, one service. Narrow-and-learning beats broad-and-blind at any spend level.
- A month of stable data before structural judgments. Week-one numbers are weather. (Smart Bidding also needs conversion volume to work with — sparse accounts fight the machine.)
- If the arithmetic says you can’t afford meaningful volume on your dream keywords, don’t abandon ads — change the battlefield: long-tail terms, BM-language variants, cheaper adjacent intents, tighter geography.
Reading whether the budget is right
After the first month or two, the account tells you — if you know where to look:
Signals you’re under-budgeted: Search Impression Share lost to budget (Google literally reporting “you were affordable but capped”); campaigns exhausting budget by mid-afternoon; strong conversion rates on tiny volume. In these cases spend is the constraint, and raising it buys more of something already working.
Signals you’ve hit a ceiling: rising cost per enquiry as spend rises; impression share lost to rank rather than budget; conversion volume flat while clicks climb. More money here buys worse clicks — the fixes are structural (better Quality Score, better pages, new keyword territory), not financial.
Signals the problem was never budget: clicks arriving, enquiries not. That’s a landing-page or offer problem wearing a media costume — audit the funnel before adjusting a single bid.
Scaling without breaking
When unit economics hold and you want more:
- Raise in steps, not leaps — 20–30% increments with a week or two between, watching CPL for bend. Doubling overnight resets algorithmic learning and floods past the auction’s efficient depth.
- Expand territory before intensity — new keyword clusters, neighbouring geographies, additional services usually scale cleaner than bidding harder on the same ten terms.
- Feed the machine cleaner signal — offline conversion imports and verified tracking let Smart Bidding optimise toward customers instead of clicks; this often is the scaling unlock.
- Mind the calendar — Malaysian auctions inflate around festive commerce peaks; a fixed monthly budget overspends troughs and starves peaks. Curve it.
Budgeting mistakes that repeat
- Copying a competitor’s rumoured spend. Their margins, close rates and lifetime values aren’t yours; their number is trivia.
- Treating management fees as optional. A well-run RM3,000 beats an abandoned RM6,000 — the weekly optimisation loop is where budgets earn or rot.
- Judging the account on revenue you didn’t track. WhatsApp enquiries and phone calls that never got counted make good accounts look bad — and get them cut. Track first.
- Setting-and-forgetting the number. Your budget is a hypothesis. Recalculate quarterly as close rates, auctions and seasons move.
Bottom line
Work backwards from margin, size forwards for learning: customer value × close rate sets your affordable CPL; auction price ÷ conversion rate predicts your actual CPL; the gap between them is the optimisation job; and volume requirements set the monthly number. Start concentrated, judge on month-scale data, scale in steps toward the ceiling the data reveals.
Next step: if you’d like this arithmetic run on your real auction — with Keyword Planner data, honest scenario ranges and a recommendation you can hold us to — request an audit. It’s the first thing we build for every Google Ads engagement anyway.
