Every owner about to spend real advertising money asks the same question, and every agency answers with whichever platform it is better at running. The honest answer is that they do different jobs — and one property of your business, not any feature of the platforms, decides the split. Here is the rule, the budget breakdown and what to read at week 2, 6 and 12.
In this article
- "Which platform is better" is the wrong question
- The Intent Split
- A decision rule you can apply today
- What Google actually costs and where it breaks
- What Meta is good at and where it breaks
- The third option most people forget
- The cheapest money on the table
- How to split the first $5,000
- Mistakes that waste the first month
- What to look at, and when
- Frequently asked questions
"Which platform is better" is the wrong question
Every owner about to spend their first serious advertising money asks the same thing: Google or Meta? And every agency answers with whichever one they are better at running.
The honest answer is that they do different jobs, and the right split is decided by one property of your business rather than by any feature of the platforms. Google Search captures demand that already exists. Meta creates demand that does not exist yet. A company selling emergency water damage restoration and a company selling a $58 skincare set are not choosing between two versions of the same tool — they are in fundamentally different situations, and the budget should reflect that.
Get this decision right and the first $5,000 buys you data you can act on. Get it wrong and it buys a month of clicks and a conclusion that "ads do not work for our business."
The Intent Split
Everything you can buy falls into one of two categories.
Demand capture. Someone has already decided they need what you sell and is actively looking. They type "emergency plumber near me", "commercial cleaning quote Dallas", "CRM for construction companies". Your job is to be there, be credible, and be fast. Google Search, Google Local Services and Bing live here. So does the map pack, which is free and criminally neglected.
Demand creation. Nobody wakes up searching for your product. They are scrolling, and something interrupts them well enough to create a want that did not exist ten seconds earlier. Meta, Instagram, TikTok and YouTube live here. Most consumer products, most new categories and most services people postpone indefinitely live here too.
The two are not competitors. They are stages. Demand creation fills the pool that demand capture later fishes from — which is why a company running both eventually sees its Google search volume for its own brand name climb, and why measuring the two channels in isolation always undervalues the one running upstream.
Does a stranger with your problem type something into a search bar? If yes, start with capture — the demand is already paid for by someone else's marketing and you are simply showing up. If no, you have no choice but to create demand, and you should budget for a longer, more expensive learning period.
Keep reading — free
The rest of this article is worth the 20 seconds
Tell us where to send it and the full piece unlocks right here, along with everything else on this blog.
Rather just talk? Message us on WhatsApp.
A decision rule you can apply today
Answer four questions honestly. Each one moves the split.
1. Is there search volume for the problem you solve? Check it — Google's Keyword Planner is free and takes fifteen minutes. If the terms that describe your service get meaningful monthly searches in your service area, capture gets the majority of the budget. If they get almost nothing, you have your answer and it is not the one you wanted.
2. Is the purchase urgent or postponable? Burst pipe: urgent, capture. Bathroom remodel someone has been thinking about for two years: postponable, creation. Urgency is the single strongest predictor of whether search will outperform social.
3. What is the average order value? Below roughly $150, Google Search clicks in competitive categories will not pay for themselves and you need the cheaper reach of social plus repeat purchase. Above $1,000, an expensive click is often trivially profitable.
4. Can you produce video and images regularly? Meta is a creative-fed machine; performance decays as creative fatigues, and a company that cannot ship new material every few weeks will watch results erode no matter how good the targeting is. Google Search runs on text and needs far less feeding.
What Google actually costs and where it breaks
The strength of Google Search is that intent is already there. The weakness is that everyone knows it, so the price reflects it.
Cost per click in the US varies enormously by category and metro — competitive home services and legal terms can run from roughly $8 to well over $45 a click, while non-brand e-commerce terms often sit under $3. Do not budget from a national average; pull real numbers for your own terms and city before deciding anything.
Three things break Google campaigns for small advertisers, and all three are avoidable:
Broad match with no negatives. The fastest way to spend $3,000 on searches for "free", "DIY", "salary" and your competitors' warranty claims. Start on phrase and exact match, and build the negative list from the search terms report every week — that report is the single highest-value screen in the account.
Sending traffic to the homepage. Someone searching for one specific thing should land on a page about that specific thing. Homepage traffic converts a fraction as well, and it is the most common reason a technically fine campaign underperforms.
Not protecting the brand term. Competitors bid on your company name. Brand clicks are cheap and convert extremely well, and choosing not to buy them is choosing to hand warm buyers to whoever will.
What Meta is good at and where it breaks
Meta's strength is reach at a low cost per person and an ability to make someone want something they were not looking for. Its weakness is that you are interrupting, so you must earn the attention every single time.
Clicks are usually far cheaper than search — often between $0.60 and $2.50 — but landing page conversion is typically a third to a half of what search traffic converts at, because the visitor arrived with no intent of their own. The two effects partly cancel; the cost per acquired customer often ends up closer than the cost per click suggests.
What breaks Meta campaigns:
Creative fatigue. The same ad shown to the same audience decays within weeks. Plan on new creative every two to four weeks — it is an ongoing production cost, not a setup cost, and companies that do not plan for it interpret normal fatigue as platform failure.
Impatience during learning. The algorithm needs a meaningful number of conversion events before it optimizes well. Changing budgets and audiences every three days keeps it permanently in learning and permanently mediocre.
Weak conversion tracking. With modern privacy restrictions, an unverified domain and no server-side signal will underreport results — and you will make budget decisions against numbers that are wrong in a direction you cannot see.
The third option most people forget
For local service businesses in eligible categories, Google Local Services Ads sit above everything else on the page, are billed per lead rather than per click, and require a background and license check to run. Where they are available, they routinely produce the lowest cost per acquired customer of any paid channel — and they are frequently overlooked because they live in a separate interface from Google Ads.
Two more free assets belong in this conversation before any budget is spent. A fully completed Google Business Profile with real photos, service areas and a steady flow of reviews drives the map pack, which sits above the paid results for local intent. And reviews themselves function as an advertising multiplier: the same click converts measurably better at 4.8 stars with 200 reviews than at 4.3 with 40. Spending on ads while ignoring the profile is paying for traffic to a page that undercuts you.
The cheapest money on the table
Before arguing about prospecting budgets, spend the smallest amount on the people who already visited and did not convert — which, on a good page, is 95% of them. Retargeting is the one line item that is nearly always underfunded, because it produces no impressive volume numbers and simply recovers demand you already bought.
Two rules keep it profitable. Cap the frequency. Following someone around the internet eleven times a week does not persuade them; it annoys them into remembering you badly. Three to five impressions a week is plenty. Change the message. The retargeting ad should not repeat the ad that brought them in — it should answer the reason they hesitated: the price question, the timeline, the proof, the guarantee.
For most businesses, 15% to 20% of paid budget on retargeting produces the lowest cost per acquired customer in the entire account. It is also the first thing cut when budgets tighten, which is precisely backwards.
How to split the first $5,000
Three starting points, depending on how the four questions came out.
Existing search demand, urgent purchase, deal size above $800. Put 70% into Google Search and Local Services, 20% into Meta retargeting only, and hold 10% for the landing page and tracking work. Retargeting-only on Meta is deliberate: it is the cheapest way to stop losing the traffic Google already paid for.
Little search demand, discretionary purchase, or a product people do not know exists. Put 65% into Meta prospecting, 20% into Meta retargeting, and 15% into Google brand and a small exact-match test — because some search demand always exists once people start seeing you.
Genuinely unsure. Split 50/50 for six weeks with identical tracking on both, and let the cost per acquired customer decide. This costs a little efficiency and buys certainty, which is usually the better trade for a first budget.
Here is what a $5,000 month can look like for a business with a $2,400 average order:
• Google Search, $3,000 at a $22 average CPC → 136 clicks, 11% landing page conversion → 15 leads at $200 each; 60% qualified; roughly 3 customers → CAC near $1,000
• Meta, $2,000 at a $1.40 average CPC → 1,429 clicks, 3.5% conversion → 50 leads at $40 each; 24% qualified; roughly 2 customers → CAC near $900
Read those two lines carefully, because they contain the lesson of this entire article. Cost per lead differs by 5×. Cost per customer differs by almost nothing. Any decision made on the first number would have been the wrong decision — which is the argument laid out in full in cost per lead is lying to you.
Mistakes that waste the first month
Launching without conversion tracking. Then the month produces spend and opinions instead of data. Install it before the first dollar, and confirm a test conversion actually registers.
Splitting $1,000 across four platforms. Nothing gathers enough data to learn anything. One or two channels, funded properly, beats four channels starved.
Judging in week one. Both platforms need a learning period, and search needs enough clicks to produce a meaningful number of conversions. Read week one for technical errors only — approvals, tracking, obviously wrong search terms — and nothing else.
Great ads pointing at a weak offer. Traffic is the multiplier, not the message. If the page fails the swap test described in what makes an offer convert, more traffic simply makes the same failure more expensive.
Nobody ready to answer. The most expensive mistake of all, and it has nothing to do with either platform. A campaign producing 50 leads into an inbox checked twice a day is a campaign producing 50 receipts. Fix the response process first — see speed to lead.
What to look at, and when
Week 2. Technical health only. Are conversions firing? Are the search terms sane? Is the landing page converting above 5%? Do not touch budgets.
Week 6. The first real read. Cost per qualified lead by channel and by campaign. Kill the worst performer, move its budget, keep everything else steady.
Week 12. The decision point. Cost per acquired customer, compared with gross profit per customer. This is the number that says scale, hold or stop — and by now you have enough closed deals for it to mean something.
Anyone offering you a verdict before week six is guessing, and anyone who wants a decision before week two is guessing with your money.
Not sure which side of the split you are on?
Tell us what you sell, what it costs and whether people search for it. We will tell you where the first budget should go — including when the answer is neither.
Frequently asked questions
Should a small business use Google Ads or Facebook Ads?
It depends on whether demand already exists. Google Search captures demand — someone has decided they need what you sell and is typing it into a search bar. Meta creates demand — nobody was looking, and something interrupted them well enough to create a want. Answer four questions to decide: is there real search volume for your problem in your area, is the purchase urgent or postponable, what is your average order value (below roughly $150 competitive search clicks will not pay for themselves), and can you produce new creative every few weeks, which Meta requires and Google Search largely does not.
How should I split my first $5,000 in ad budget?
With existing search demand, an urgent purchase and deals above $800: 70% Google Search and Local Services, 20% Meta retargeting only, 10% held for landing page and tracking work. With little search demand or a discretionary purchase: 65% Meta prospecting, 20% Meta retargeting, 15% Google brand plus a small exact-match test. Genuinely unsure: split 50/50 for six weeks with identical tracking and let cost per acquired customer decide — that costs some efficiency and buys certainty, which is usually the right trade for a first budget.
How long before I can judge whether ads are working?
Read week two for technical health only — are conversions firing, are the search terms sane, is the landing page converting above 5% — and do not touch budgets. Week six is the first real read: cost per qualified lead by channel and campaign, kill the worst performer and move its budget. Week twelve is the decision point, when you compare cost per acquired customer against gross profit per customer and decide to scale, hold or stop. Anyone offering a verdict before week six is guessing, and anyone who wants a decision before week two is guessing with your money.