Programmatic Advertising
CTV Ads for Restaurants: What $1,000 Buys
Part of the guide: Connected TV (CTV) Advertising
Here is the number that decides whether CTV works for your restaurant. At an assumed $30 CPM, $1,000 buys about 33,000 ad views. Show each household the ad four times and you have reached roughly 8,300 homes.
If each extra party leaves you $18 in contribution after food and variable labour, you need about 56 extra parties from those homes just to break even, which is 0.67% of everyone who saw the ad.
Every figure in that sentence except the arithmetic is an assumption, and the rest of this guide shows you how to replace them with your own.
CTV (connected TV) means video ads on streaming services watched on a television. This guide is for independent restaurants, cafes and bars in the US, UK, Australia and Canada. Self-serve tools are US-centred, so outside the US, ask each platform what is available in your country. We cover how targeting works, what to put in a 15-second spot, the break-even math, and how to measure walk-ins when nobody clicks anything.
What the Evidence Actually Says
Independent proof that streaming TV fills small restaurants is thin. Most published results are vendor case studies about big chains.
One example: Amazon's case study of 17 quick-service brands across 30+ streaming campaigns reports an incremental return of $1.74 per $1 spent and a 0.79% incremental sales lift, measured with partner Attain (Attain Incrementality Study, February 2026). Amazon also says campaigns combining live sports and streaming did better ($2.08) than single-platform ones ($1.17).
Read that with care:
- It is Amazon's own marketing page, and it does not publish the full method.
- The return is on sales, not profit. At restaurant margins, $1.74 of extra sales per dollar of ads may not cover the ad.
- The brands are national chains with large budgets and existing awareness.
The honest takeaway: CTV can create incremental visits, but the effect on any one restaurant is small relative to the ad spend unless your offer, targeting and tracking are tight. That is why the test plan below matters more than the platform.
How Targeting Works (and One Common Mistake)
Many guides say you can target "a radius around your restaurant". On Roku's self-serve tool, you cannot. Roku's location targeting help page lists states, postal or ZIP codes and designated market areas (DMAs), and says:
- ZIP codes should be residential ZIP codes only.
- You can include or exclude locations using only one geographic type at a time.
- ZIP targeting is not available if you declare a special ad category (credit, housing or employment).
Practical steps:
- Pull 50 recent reservations, delivery orders or loyalty sign-ups and note each ZIP code.
- Rank the ZIPs by share of customers. Pick the top three to six that sit within a short drive.
- If your restaurant is in a commercial or downtown ZIP with few homes, add the residential ZIPs around it, since the platform wants residential ones.
- Check the forecaster before you launch. Roku's audience targeting page warns that limiting your audience may impact performance, so expect a very small ZIP set to cost more per view.
Other platforms differ. Tubi announced a self-serve platform powered by Universal Ads in May 2025 (Fox Corporation), but the announcement does not list targeting options or minimums, so ask before you plan around it. Our CTV platform comparison goes through the self-serve routes side by side.
Is a tight ZIP set always right? Not always. Roku's car dealer case study describes Chapman Automotive Group dropping narrow ZIP targeting for a whole DMA, buying 45.5M+ impressions at a $7.36 effective CPM and reaching 1.7M+ households, in the team's words choosing "reach over repetition". That is a large dealer group with cheap inventory, not a single restaurant, and a $7.36 CPM is far below the vendor estimates below. For a restaurant where people will not drive 40 minutes for dinner, wide reach mostly buys waste.
What to Put in a 15-Second Restaurant Ad
Food is visual, so spend the first two seconds on it. A simple structure:
| Seconds | Show | Say or write |
|---|---|---|
| 0 to 3 | Your best dish, close up, steam or sizzle | Dish name |
| 3 to 8 | Room, patio or the team plating | One line on what makes it yours |
| 8 to 12 | The offer | "Tuesday: free dessert with any main" |
| 12 to 15 | Logo, street or neighbourhood, short URL or QR code | Phone number |
Rules that save money:
- One offer, one action. "Mention TUESDAY for a free dessert" can be counted. "Come see us" cannot.
- Name the neighbourhood. Viewers in your ZIPs should know it is nearby.
- Use sound. CTV is usually watched on a TV with the audio on. Do not rely on captions alone.
- Film on a phone in good window light if you have no budget. Most self-serve platforms accept a file at their stated specs; check the spec sheet before you shoot.
- Make a second version after four weeks (new dish or season) so the same homes do not see an identical ad all quarter.
Budget: Work Backwards From a Break-Even
CTV is bought by CPM, the cost per 1,000 ad views. No platform publishes a price list. Vendor estimates put Roku CPMs at about $20 to $60 and quote campaigns starting around $500 in total (Vibe, Roku advertising cost; the vendor notes Roku does not disclose a minimum, so confirm in the tool). Our CTV cost guide covers the market-wide ranges.
The Math
Every input below is an assumption for illustration: the CPMs sit inside the vendor range above, and the frequency, ticket and margin are placeholders. Replace each with your own numbers.
Assumptions: $1,000 for the month, each household sees the ad 4 times, an extra party of four spends $60 and leaves you $18 in contribution after food cost and variable labour.
| Cheaper | Middle | Pricier | |
|---|---|---|---|
| CPM (assumption) | $20 | $30 | $45 |
| Ad views | 50,000 | 33,333 | 22,222 |
| Households at 4 views each | 12,500 | 8,333 | 5,556 |
| Extra parties to break even ($1,000 / $18) | 56 | 56 | 56 |
| Share of reached households that must visit | 0.44% | 0.67% | 1.00% |
| Cost per extra party at 1% response | $8 | $12 | $18 |
How to read it: break-even parties is the budget divided by contribution per party. If your real contribution per party is $30, you need 33 parties. If it is $10, you need 100. The last row shows what you pay per extra party if 1% of reached households visit, which is an optimistic guess, not a benchmark.
Three things follow:
- Frequency is a dial. Cutting from 4 views to 2 doubles households reached for the same money, but fewer people see it enough times to remember it. Test, do not guess.
- Lifetime value changes the answer. If a new regular visits 10 times a year, the first party does not have to pay for the ad. Use your repeat rate, not one visit.
- Compare with search. If your Google Ads for "restaurant near me" cost less per extra party, spend there first. See our restaurant Google Ads guide.
A 60-Day Test Plan
| Phase | What to do |
|---|---|
| Before launch | Record baselines: covers on your two slowest nights, weekly branded searches for your restaurant name, and your Business Profile calls and direction requests |
| Weeks 1 and 2 | Launch one 15-second ad with one offer in your chosen ZIPs; check delivery and completion rate in the dashboard |
| Weeks 3 to 6 | Compare slow-night covers and offer redemptions against baseline; swap creative if completion is low |
| Weeks 7 and 8 | Decide: scale, change the offer or ZIPs, or stop |
Sixty days is my rule of thumb, not an industry standard. It gives you roughly eight weekly cycles to see a pattern past one rainy Friday. Run the test off-peak, when you have the empty seats to fill.
Measuring Walk-Ins When Nobody Clicks
CTV viewers rarely type your URL into a television. They search on their phone, ask their partner, or just turn up. So set up several weak signals and read them together:
- Offer code. "Say TUESDAY for a free dessert." Have servers log it in the till or a notes app.
- Source question. Add "How did you hear about us?" to online booking and the phone script, with "Streaming TV ad" as an option.
- Branded search. Compare weekly searches for your restaurant name in Google Search Console before and after launch.
- Business Profile. Watch calls, direction requests and website clicks in your profile's performance report. Make sure the profile has current hours, menu and photos first; see our Google Business Profile checklist.
- Phone calls. If you use a call tracking number in your ad, see call tracking for local business.
None of these proves cause. Together, a lift in all of them on the weeks the ad ran, with a flat line before, is a decent signal. A lift in none is a clear one.
Which Restaurants Should Skip It
- Under $500 a month to spend. You will reach too few homes to see an effect. Put the money into search and your profile.
- No slow nights to fill. If you are full every evening, extra awareness has no seat to sell.
- A weak basics layer. If your profile photos are old, reviews are unanswered or the menu link is broken, fix that first. Our Instagram marketing for restaurants guide covers a cheaper way to build the same visual awareness.
- Sports bar? Live sport is bought differently; see Peacock ads for local businesses.
Do This Today
- Pull your last 50 bookings, orders or sign-ups and list the ZIP code of each. Pick the top three to six residential ZIPs.
- Work out your contribution per extra party (average spend minus food and variable labour) and your break-even party count for a $1,000 test.
- Choose your two slowest nights and one offer that you can count, such as a code for a free dessert.
- Film three 15-second clips of your best dish and the room, then open one self-serve tool (Roku Ads Manager or Tubi) and write down its minimum and the forecast reach for your ZIPs.
- Record your baselines this week: slow-night covers, branded searches and Business Profile direction requests. You cannot prove a lift without them.
FAQ
Q: Can I target a radius around my restaurant on CTV?
A: On Roku's self-serve tool, the location options are states, ZIP codes and DMAs, per its help page, and ZIP codes should be residential. Other platforms differ, so ask before you buy.
Q: What does CTV cost for a small restaurant?
A: There is no price list. Vendor estimates put Roku CPMs at about $20 to $60, and campaigns often start around $500. Use the math above with your own quote.
Q: Is CTV better than Google Ads for a restaurant?
A: They do different jobs. Search catches people already looking for dinner; CTV tries to put you in their head before they look. Most single-location restaurants should prove search and their profile first.
Q: How long before I see results?
A: Plan on about 60 days, which is my rule of thumb. Watch offer redemptions and branded searches first, because they move before covers do.
Q: Do I need an agency?
A: One location with one offer can run through a self-serve tool. Multiple locations or complex measurement may justify help. Ask any agency for fees and minimums in writing. See our CTV advertising service.
Sources
- Amazon Ads: Amazon DSP streaming TV QSR results (Attain Incrementality Study, February 2026)
- Roku Ads Help: Location targeting
- Roku Ads Help: Audience targeting
- Roku Advertising: Hyperlocal CTV, car dealer network case study
- Fox Corporation: Tubi self-serve ads platform announcement, May 2025
- Vibe: Roku advertising cost (vendor estimates)
Related reading: CTV advertising platforms compared, DOOH advertising for restaurants, and our restaurant marketing page. Want a test planned and tracked? Contact DataLatte.
Free for local businesses
Want this applied to your business?
I'll review your Google presence, local SEO, and ad accounts — and send you a specific action plan within 48 hours. No pitch, no pressure.
Want hands-on help?
See how DataLatte handles CTV Advertising for local businesses.
Nataliia
Local marketing strategist with 10+ years at global agencies — OMD, Dentsu, GroupM, and BBDO. Now helping small businesses get the same data-driven edge. Based in Europe, working with clients in the US, UK, Australia, and beyond.
About NataliiaRelated articles
Programmatic Advertising
CTV Ads for Yoga Studios: Streaming TV Advertising to Fill Your Classes
13 min read
Programmatic Advertising
Connected TV Ads for Pet Groomers: How to Get in Front of Pet Owners on Hulu & YouTube TV
9 min read
Programmatic Advertising
CTV Ads for Nail Salons: Streaming TV Advertising That Fills Your Books
13 min read
Programmatic Advertising
CTV Ads for Coffee Shops: Reach Local Customers on Streaming TV
9 min readWant this applied to your business?
Let's review your current marketing setup together — free, no obligations.
Get Your Free Marketing Audit