Most operators do not have a marketing ROI problem. They have a marketing honesty problem. Spend $2,000 on Google Ads, see $7,000 in attributed sales, and call it a 250% ROI. The actual net, after a 60% prime cost and a $700 promo discount applied at the door, is closer to 40%. The Google Ad rep does not show that number. The platform does not show that number. The bookkeeper sees it three months later, at year-end, in a way nobody connects back to the campaign that caused it.
This guide is the version we use internally to score restaurant marketing channels. The formula is one line. The hard part is everything around it: which costs to load into the denominator, what to do with the multi-touch journey, when to declare a channel dead, and how much of the budget belongs in each channel given the size of the restaurant. The original frameworks below (the RV Portfolio Allocator, the Net-vs-Gross Honesty Gap, the Channel Payback Window) were built by running 32 client cohorts through the same scoring model in Q1 2026 and reverse-engineering what actually printed.
If you want the calculator only, scroll to the calculator section. If you want to know which number to plug in and why, read the part in front of it. The frameworks below are the same ones Restaurant Velocity applies on every onboarding.
The ROI formula most platforms hide
The formula is:
ROI % = ((Revenue x Contribution Margin) – Campaign Cost) / Campaign Cost x 100
Three pieces, in order of how often they are wrong:
Revenue. The platform-reported number is gross sales. It includes the promo dollars you discounted, the COGS-heavy items that were part of the offer, and (in most cases) repeat visits from people who would have come anyway. The honest revenue number is incremental sales: orders that would not have happened without the campaign.
Contribution margin. Most operators run a 28-32% food cost and a 28-32% labor cost. That puts prime cost at 56-64% and contribution margin (what is left after the variable cost of serving that order) at 36-44%. Use your own number. If you do not know it, use 40% and refine it next month.
Campaign cost. Not just the ad spend. Add the agency fee, the platform fee on bookings, the printed coupon redemption, the gift to the influencer, the staff overtime if the promo created a surge. We see operators understate this by 15-30% across the board.
The most expensive shortcut in restaurant marketing
Using gross revenue instead of net profit. A campaign drives $5,000 in sales on a 60% prime cost. The platform shows you a 525% ROI on $800 spend. The honest math: $5,000 x 0.40 = $2,000 contribution. Minus $800 spend = $1,200 net. Divided by $800 = 150% ROI. That is still profitable, but the decision to scale aggressively versus cautiously hinges on the difference between 150% and 525%. We have seen restaurants pour budget into a “525%” channel and slow down on a “150%” one when, after net adjustment, both were sitting at 130%.
The net-vs-gross honesty gap by channel
This is the original honesty gap we calculate for every restaurant we onboard. We take the channel’s headline ROI (the gross revenue figure the platform reports back), apply a 40% contribution margin (typical for a 60% prime-cost operator), then re-state the ROI. Three things become obvious that were not obvious before:
- Email and SMS still dominate, but the gap is smaller than the headline. $36 per $1 on email turns into $13.40 per $1 net. That is still the best-performing channel in the stack, just not 36x better than everything else.
- Meta cold audiences are the most-overstated channel. A 2x reported ROAS becomes a slight loss after prime cost. This is why operators who scale cold Meta without retargeting bleed cash for three months and cannot figure out why.
- Google Ads is honest enough that it survives the recalculation. A 4x reported ROAS becomes a 60% net ROI, which is the floor most operators should be willing to fund.
The practical rule: apply your real contribution margin to the platform-reported revenue before you make any budget decision. If a channel does not clear net 50% ROI after the adjustment, it is a candidate for cut, not scale.
The RV Portfolio Allocator
The allocator answers the question every operator actually asks: given the budget I have right now, how do I split it across channels to maximize blended ROI? The matrix is built from the 32-client cohort. Each row is a revenue band; each column is a channel. The percentage is the share of monthly marketing spend that maximized 12-month blended net ROI in that band.
Three patterns that contradict the standard playbook agencies sell:
Small restaurants over-invest in cold Meta. The under-$50K band frequently runs 40-50% of budget into cold Meta because cold Meta is what agencies pitch. The cohort math: 5% of budget into cold Meta, 30% into GBP and review velocity, drives more incremental revenue.
Email and SMS should compress as the restaurant grows. Counter-intuitive. The list is bigger so the ROI is higher in dollar terms, but the marginal $1 has lower returns once you are already mailing your full list weekly. Above $150K/mo, the smart move is to cap email/SMS at 15-18% and route excess budget to retargeting and acquisition.
Micro-influencer scales with revenue, not down. At $400K+/mo, micro-influencer climbs to 13% of budget. The reason: a $400K-revenue restaurant can absorb a $5,000 monthly creator program where the incremental customer acquisition cost is $35-45 (against an LTV of $480-720). At $40K-revenue, that same program eats 30% of budget.
How to use the allocator
Identify your revenue band. Multiply each channel column by your monthly marketing budget. That is your target spend per channel. If your current spend is more than 50% off the target in any column, that is the channel to rebalance first. Most operators discover they are 2-3x overweight on cold Meta and 4-5x underweight on Google Business Profile work and review velocity.
Channel-by-channel net ROI (what to expect)
Email and SMS (warm list)
Email returns $36 in gross revenue per $1 spent on a clean, segmented list. SMS returns $71. After 40% contribution margin, net ROI lands at 1,340% (email) and 2,740% (SMS). These numbers are real, but they apply only to the list you already own. The cost of acquiring the list (paid social, in-store sign-up flows, Wi-Fi gate) belongs in a different line item.
The mistake we still see in 2026: blasting one Tuesday-morning coupon to the entire list. Segment by RFM (recency, frequency, monetary). Send the premium offer to top-decile spenders on Thursday/Sunday evenings. Send the reactivation offer to 60+ day lapsed customers on Wednesday afternoons. The 32-client cohort shows a 2.4x lift in conversion when the same offer is segmented versus blasted.
Google Ads (search and brand)
Average CPC for restaurants in 2026: $2.40-3.20 in tier-1 metros, $1.40-2.00 in tier-3. Conversion to order or reservation: 5-8% on brand terms, 1-3% on generic (“italian restaurant near me”). At $1,500/mo with a 6.5% conversion rate and $48 AOV, the math: 600 clicks, 39 orders, $1,872 revenue, $749 net contribution at 40% margin, $749 – $1,500 = -$751 net loss. To clear net-positive on Google Ads at this AOV, you need either a 9%+ conversion rate or a higher AOV. The ad copy and landing page get you there. A neglected campaign sits at 1.5% and loses money for months.
Meta retargeting (warm)
Net ROI 100-200% on a properly built audience: site visitors past 60 days, engaged followers past 30 days, customer match from email list. Cold targeting in restaurants is a higher-risk play. The single best decision a $50-150K-band operator can make is to refuse to run cold Meta until retargeting clears net 100% ROI for two consecutive months.
Google Business Profile (zero-spend)
The “infinite ROI” label is a shortcut. There is a real cost: the time it takes to maintain Posts, Photos, Q&A, and review responses, plus the licensing cost of any tool that automates it. A solo operator handling GBP manually spends 4-6 hours/week (call it $200-300 in opportunity cost). Done well, GBP drives 15-35% of new-customer foot traffic for local restaurants. This is why the allocator weights it heaviest in the under-$50K band: it is the highest-leverage cheap channel for a restaurant that cannot yet absorb paid acquisition.
Micro-influencer (5,000-50,000 follower local creators)
Cost per post $200-1,000 depending on follower count and location. The 32-client cohort shows net ROI lands in the 100-180% range when influencers are vetted for genuine local audience (check the comment-to-like ratio, ask for last-90-day-city-breakdown screenshots). The same channel runs near zero net ROI when the creator has a national audience that does not convert to a single-location restaurant.
Use this calculator right now
Plug your numbers into the calculator below. It applies the formula honestly: net contribution divided by total campaign cost, expressed as ROI%.
Calculate your restaurant marketing ROI
If the channel clears 150% net ROI after this calculation and has held that level for at least the full payback window, scale it. If it clears 50-150%, keep it running and optimize. Below 50%, prepare to cut. Operators who do not want to run the GBP and reviews channel manually can start the free trial and let the autopilot handle it.
Channel payback window (when to declare a channel dead)
Most channels get killed too early. The payback window is the time it takes for the first attributable revenue to land. Below the upper bound, you do not have enough data to make a decision.
The hardest call is Meta cold (21-60 days). Operators get nervous at day 14, kill the campaign, and never see the curve flatten. The 32-client cohort shows the median Meta cold campaign hits net-positive at day 38. The median Google Ads campaign hits net-positive at day 22. Email shows up in 48 hours.
The rule we follow: do not change spend, audience, or creative on any channel during its payback window unless the channel is bleeding more than 2x the planned spend. Set the budget, set the kill threshold, walk away.
The attribution problem nobody solves
A customer sees your Google Ad on Tuesday. Scrolls past. Sees your Instagram post Thursday. Saves it. Checks your Google Business Profile Friday for hours. Then gets your email coupon Saturday and orders. Which channel worked?
Last-click attribution (what most platforms show you by default) gives email 100% of the credit. Google and Instagram get zero. But all four channels contributed. Google created the awareness. Instagram kept you top of mind. GBP verified the hours and the menu. Email closed.
This is not a theoretical problem. We see operators over-invest in email (which already has high ROI on its own) and under-invest in Google and social (which are higher-friction but essential for reaching new customers). The result: the warm list is well-served and slowly shrinks because nothing new is being funneled into it.
Pick a multi-touch model and stick with it
First-touch credits the channel that originated the customer (usually Google). Last-touch credits the closer (usually email or direct). Linear gives each touchpoint equal credit. Time-decay weights later touches more heavily (email 50%, Instagram 30%, Google 20%, for example).
For restaurants, the cohort shows linear or time-decay produces budget decisions closest to what actually printed at year-end. Time-decay is slightly better in markets with heavy paid-search competition. Linear is more forgiving in smaller metros.
Five practical attribution moves you can run this month
- UTM parameters on every link. Format: ?utm_source=email&utm_medium=newsletter&utm_campaign=may-special. Google Analytics 4 ingests these automatically.
- Unique coupon codes by channel. EMAIL20, INSTA20, FB20, GBPVIP. Redemption rates tell you which channel converted.
- POS-to-GA4 connection. Toast and Square push order data to GA4 directly. You see source attribution in your sales report.
- Call tracking by channel. Use 555-0001 in Google Ads, 555-0002 in Facebook, 555-0003 in print. CallRail or Twilio handles the routing for $20-50/mo.
- Hostess “how did you hear about us” question. Imperfect, but catches the gaps the digital attribution misses. Code the answer into the POS as a custom field.
Five ways restaurants mess up ROI tracking
1. Gross revenue instead of net profit
Covered above. The single most common error.
2. Quitting before the payback window closes
Google Ads and Meta need their upper-bound days to print. Email shows results in days. Local SEO takes 6 to 12 months. Kill a campaign before its window closes and you will never know what works.
3. Running everything at once
Launch Google Ads, Facebook ads, and email simultaneously and you cannot tell which is contributing. Sequence them. Run one channel for its payback window. Stabilize. Add the next.
4. Obsessing over vanity metrics
10,000 impressions, 200 clicks. Sounds great until none became customers. Impressions and clicks do not pay your rent. Net contribution does. The dashboard should not show CTR. It should show net ROI per channel.
5. Ignoring seasonality
A campaign in December (holiday eating-out peak) will run higher ROI than the same campaign in August (slowest month for full-service in most metros). Always compare to the same channel’s performance during the same season the year prior, not to a different channel during a different month.
Building your own tracking system
You do not need a $400/mo attribution platform to do this well. A Google Sheet, UTM parameters on every link, and POS data gets you 90% of the way there.
Eight fields to track per campaign
- Campaign name and channel
- Start date, end date
- Total campaign cost (ad spend + fees + production + staff overtime triggered by the promo)
- Attributed revenue (multi-touch model of choice)
- COGS on those orders (food + packaging)
- Variable labor on those orders
- Net profit (revenue contribution minus campaign cost)
- ROI percentage (calculated)
Update monthly. By month four you can see which channel allocation pattern is printing and which one is not. By month twelve, you have a defensible budget for next year.
If you’d rather not run this manually
The Restaurant Velocity app handles the GBP and reviews layer of the stack: Posts on autopilot, AI review replies within 60 minutes, photo cadence, ranking audits, and the Maps grid scan. The output drops into the same ROI ledger as your paid channels, so the “infinite ROI” Google Business Profile column becomes a real, line-itemed cost you can compare against everything else. Start your 14-day free trial if you want that piece off your plate.
Frequently asked questions
What is a good ROI for restaurant marketing?
After applying contribution margin to the platform-reported revenue, target net ROI 150% as the scale floor, 50-150% as keep-and-optimize, below 50% as cut candidate. The blended net ROI across all channels for a well-run restaurant marketing budget lands at 120-180% in our 32-client cohort.
How much should I spend on marketing per month?
3-6% of monthly revenue is the standard industry guidance. A $100,000/month restaurant budgets $3,000-6,000. New restaurants in the first 18 months often run 8-10% to accelerate brand awareness. Established restaurants with strong word-of-mouth can hold at 2-3%. Use the RV Portfolio Allocator to split that budget by channel.
Should I test all channels at the same time?
No. Sequenced launches produce signal. Simultaneous launches produce noise. Run one channel through its payback window, document the net ROI, then add the next channel. The whole portfolio should take 4-6 months to spin up from scratch.
When should I pause a failing campaign?
After the upper bound of the channel’s payback window has passed and net ROI is below 50% with no upward trend across the last two reporting periods. Killing during the payback window is the single most common reason restaurant marketing budgets get wasted.
How do I track walk-in attribution?
Walk-in is the hardest attribution problem in restaurants. Three combined practices get you within 15% of true: a hostess “how did you hear about us” question coded into the POS, a unique coupon code on every promotional asset, and a week-over-week foot-traffic delta compared to the channel’s launch and pause dates.
When does local SEO pay off?
6 to 12 months. The early signal (3-6 months) is GBP impressions and ranking position; the revenue signal lags by another 3-6 months. Local SEO sits at the bottom of the payback window because the channel does not produce a clean, attributable click-to-order trail the way paid search does. The ROI, once it prints, is the highest in the stack because there is no ongoing per-click cost.
Is the Restaurant Velocity app a marketing channel or a tool?
It is the tool that runs the GBP and reviews layer of the stack. The eight workflows (review replies, Instagram and Facebook comment replies, weekly Google posts, photo library, profile audit, local rank grid, competitor tracking, and control modes) are the recurring work that powers Google Business Profile, which is itself a channel in this allocator. Budgeted as a tool line, the subscription cost sits in the campaign-cost denominator for the GBP and reviews channel column.
