The default answer is “probably not yet”
I get a lot of phone calls from operators asking whether they should run Google Ads. The honest first answer is usually no — at least not until the Google Business Profile is clean, the Maps citation work has had 12–18 months to settle, and the website has the basics (mobile-friendly menu, working reservation widget, a real catering page). Ads on top of broken fundamentals are a way to set money on fire faster.
That said, paid search has three real jobs for an independent Toronto restaurant. The trick is knowing which one you are doing and not paying for the other two.
When paid search is genuinely wasted
The classic waste pattern: a 60-seat Italian room in Yonge & Eglinton spends $1,200 a month on Google Ads and 70% of the spend is on its own brand name. The brand name was already winning the Map pack for free. Every paid click was a free click the ad budget just bought. This pattern is extremely common because the default Google Ads setup recommends branded keywords, and the metrics look great — high CTR, low cost-per-click, lots of “conversions” — while doing none of the work the restaurant actually needed.
A second waste pattern: “[cuisine] Toronto” broad-match. The geography is too wide, the intent is too generic, and the result is clicks from people in Etobicoke who were looking for a place near home, not a 25-minute drive away.
If your current Google Ads spend is going to either of those, the right move is to pause the campaign, fix the Maps work, and revisit in 90 days.
When paid search earns its keep
Three use cases hold up. None of them is “we want more visibility”:
- Catering and private events. B2B queries — “corporate lunch catering North York,” “private dining room Liberty Village 30 people” — are high-intent, low-volume, and expensive in a useful way. The Maps pack does not always serve them well, the inquiries close at higher margins, and a $30 cost-per-lead is reasonable when the average booking is $1,200+.
- “[Cuisine] near me” where your Maps presence is weak. A Filipino room that just opened in Mississauga is not going to rank in Maps for six to twelve months. Paid search for “Filipino restaurant Mississauga” and tightly geo-fenced “Filipino food near me” fills the gap while citations and reviews catch up.
- Pre-launch or new-location announcements. Two to four weeks of paid before opening a second room is defensible if the campaign is narrow — branded keywords for the new location only, geo-radius of 3–5 km.
If none of these three apply, the answer is probably “not yet.”
Negative keyword lists are 80% of the work
A restaurant Google Ads campaign without a negative keyword list will burn through 30% of the budget on irrelevant queries inside two weeks. The basic list should include, at a minimum:
- Jobs: “jobs,” “hiring,” “career,” “wage,” “tips” (yes, tips — people search “best tipping etiquette restaurant”).
- Supply chain: “wholesale,” “supplier,” “distributor,” “for sale,” “lease,” “POS system.”
- Education: “school,” “culinary,” “course,” “recipe,” “how to.”
- Reviews and ratings: depends on your goal, but if you do not want to pay for people researching, exclude “review,” “Yelp,” “Reddit.”
- Competitor brand names: unless you are explicitly running a conquesting strategy, which is a different conversation.
Build the list once, review monthly, and add anything weird you see in the search-terms report. The search-terms report is the most underused screen in Google Ads.
Tight geo-radius beats broad city targeting
The default “Toronto” target hits 2.7 million people, most of whom will never drive to your specific room. Run a 3–5 km radius around the restaurant for dine-in campaigns, and a separate larger radius (or list of postal-code targets) for catering.
For a catering campaign aimed at the financial district, target the M5H, M5J, M5K, M5L, M5V postal-code clusters rather than “Toronto.” Costs per click drop and the lead quality jumps because the searcher is actually somewhere your driver can reach during a lunch window.
Dayparting around your actual service
The mistake is running ads 24/7. A restaurant that opens at 5pm does not need to be paying for clicks at 11am on a Tuesday, when the searcher is making a decision for that night and will likely have moved on before dinner.
Two reasonable dayparting patterns:
- Dinner-led rooms: Wednesday–Sunday, 2pm–9pm. Lunch hours and Monday–Tuesday off unless you specifically want them.
- Lunch-led catering: Monday–Friday, 8am–2pm. Catering buyers are usually office admins making decisions in the morning for that day or week.
Saving 40–50% of the budget by not running ads when nobody is converting is one of the cheapest wins in the channel.
Max-click versus max-conversion bidding
For most independents under $1,500/month in ad spend, “maximize conversions” looks promising and does not work yet. Google’s machine-learning bidding needs roughly 30 conversions per month per campaign to optimize properly. A small restaurant rarely hits that.
The practical answer for the first 90 days: manual CPC or “maximize clicks” with a cost-per-click cap. Set the cap at what you can defend per visit (typically $1.50–$3.50 for restaurants). Once you have three months of conversion data and at least 30 conversions a month, then move to “maximize conversions” and monitor closely.
Attribution caveats that catch operators out
Google Ads will report conversions that did not happen the way it claims. The default conversion window is 30 days, which includes clicks that led to a Maps visit two weeks later that the diner would have made anyway. Treat the reported conversion count as an upper bound and pair it with at least one ground-truth signal: a unique phone number for the campaign, a UTM-tagged reservation widget, or a “how did you hear about us” question on catering forms.
A reasonable expectation: real attributable revenue is usually 50–70% of what Google Ads reports, and the gap is bigger for branded-keyword campaigns.
A reasonable next step
If you are currently spending on Google Ads and you cannot answer “which of the three jobs is this campaign doing?” — pause it, fix the Maps work, and revisit. We do a free 25-minute review of an existing Google Ads account for restaurants; see the contact page if a second look would help.