Google Ads vs Meta Ads for small budgets
If you can only run one, which one. The answer depends on where demand lives.
When budget is tight (under $5K/month), running both Google and Meta at half-strength usually beats running neither well. Better to pick one and run it properly.
Pick Google Ads when demand exists
If people are actively searching for what you sell, Google Ads is almost always the better starting channel. Search ads capture existing demand — the user has already decided they need something. You just need to be there at the moment they look.
Examples where this works: emergency plumbers, accountants, B2B SaaS for known categories, e-commerce in categories with searchable keywords.
Test budget: $30-50/day on Google Search for 4-6 weeks gives you enough data to evaluate.
Pick Meta Ads when demand has to be created
Some products solve problems people don't know they have, or sell on visual appeal more than functional search. If nobody is Googling for what you offer, search ads have no inventory to bid on.
Examples: fashion, lifestyle products, novel SaaS categories, anything with strong impulse purchase dynamics.
Test budget: $50-100/day on Meta for 3-4 weeks.
The hybrid case
Most service businesses have both: people searching for the category (Google catches them) and people who'd buy if shown the right ad (Meta catches them). The ideal is running both. With a $5K/month combined budget, that often means $2K Google, $3K Meta, neither at full strength.
If you can't do both well, pick the one that matches your demand profile and commit.
What kills small budgets
Spreading across too many campaigns. With $1K/month on Google split across 5 campaigns, each campaign learns slowly and converges late.
Frequent strategy changes. Both algorithms need time to learn. Changing targeting or creative every week resets the learning. Set, wait, measure, then adjust.
Ignoring landing page. The ad gets the click. The page closes the deal. If the page is bad, no amount of ad optimisation helps.
Not having a real conversion event. Both algorithms optimise toward whatever event you tell them is valuable. If your only event is "page view", you're optimising for low-quality traffic.
Realistic expectations at this budget
Small budget paid ads usually take 6-12 weeks before you have enough data to know if they work. Many businesses cancel after 4 weeks of "it's not working" — exactly when the data was about to stabilise.
If you can't commit to 12 weeks at the same budget, you probably can't afford paid ads yet. Focus on organic and referral until budget allows real testing.