Hospitality | Multi-Location Restaurant Group
From 225 legacy campaigns to 15 focused winners.
$2.50M
Reported Value*
9 to 10x
Return on Ad Spend
225 to 15
Campaigns Focused
+22.6%
Catering Sales, Booked Orders†
*The $2.50M is platform-reported conversion value across Google, Meta and LinkedIn for February through July 2026 on $211,756 of managed spend, 11.8x blended. The 9 to 10x counts purchase conversions only, $1.93M at 9.09x, and is the figure we stand behind. The window ends July 31, the last month in the archive. †Catering sales on a trailing 30-day basis through mid-August, measured in the client's booked orders rather than platform attribution. The two windows sit on slightly different store bases; the like-for-like same-store comparison ran +28.1% and we quote the lower unmatched figure. An earlier 30 to 32x catering return figure was retired in July after an audit of our own server tags found general orders labelled as catering. Spend and value figures cover February through July 2026; catering figures run through mid-August.
The Situation
A paid account no one could read
A 30+ location fast-casual restaurant group across the Northeast had cycled through four prior agencies. What they handed us was the result. It included 225 paused legacy campaigns, roughly 14 loosely connected live campaigns, and five competing naming conventions in a single account. There was no clear priority, no incrementality testing, and no honest read on what was actually working. Money was going out every day. One earlier stretch under a prior operator had burned tens of thousands of dollars on fragmented traffic campaigns at close to zero return. The winning pattern was buried in the noise, and nobody could see it.
The Approach
Find the pattern, archive the noise, scale what works
We did not start by spending more. We started by reading three years of account history to find the pattern, then rebuilt the account around it.
Diagnose the winning pattern
A full pull of three years of Google and Meta history showed the same result every time. Catering and value offers, optimized for purchase, in a single metro, with food-forward creative, returned three to four times what traffic and awareness plays did. Prior teams had touched these wins but never replicated them consistently.
Canonicalize the account
Archived 210 legacy campaigns in place, stood up 15 focused campaigns, and applied one client-readable naming system using offering, market, channel, builder, month. The whole team can now scan the account by what it sells and where, not decode it.
Build the high-intent foundation
Stood up a non-brand catering search foundation aimed at high-intent demand like corporate catering and office lunch. Standardized seasonal creative into every catering campaign across markets.
Wire attribution to reality
Rebuilt conversion tracking so reported value matches real sales, separated modeled in-store visits from online purchases, and put an offline-order value feed in motion. The goal was a number leadership can trust, not a flattering one.
Test for true lift
Designed a geo-holdout incrementality test so the group can estimate how much of the return is genuinely caused by the ads, not just credited to them. The honest question, asked on purpose.
The Results
A focused account that pays its way
$211,756 of managed spend across Google, Meta and LinkedIn reported $2.50M in platform value from February through July 2026, 11.8x blended
9 to 10x return on ad spend on the strictest basis, purchase conversions only, with modeled in-store value excluded
Catering campaigns first reported 30 to 32x on the platform. A July audit of our own server tags found general orders labelled as catering, so we retired that figure and rebuilt the view. Measured in booked orders rather than platform attribution, catering sales ran +22.6% on a trailing 30-day basis by mid-August (the like-for-like same-store comparison over the same window was higher, +28.1%; we quote the lower figure)
A new-store launch play, built one market at a time, proved the hyper-local structure and became the template for every launch after it
225 legacy campaigns consolidated into 15 focused, consistently named active campaigns
An account leadership can read by offering and market, with weekly performance reporting
CRM rebuilt alongside paid, with a 150,000-contact cleanup and lifecycle automation now feeding the funnel
A note on honesty. Reported return on ad spend in restaurant accounts is inflated by modeled in-store visits and platform attribution. We report the conservative online-sales number here. In July our own audit found server tags labelling general orders as catering, which had inflated the catering campaigns' reported return; we retired that figure and rebuilt the view on booked orders. The headline result is the restructure, the measurement discipline, and the willingness to correct our own number, not a single flattering multiple.
Engagement Included
Full scope of work
Multi-year Google and Meta account audit
Account restructure and canonical naming system
Non-brand catering search build
Performance Max and Meta management
Creative standardization across markets
Conversion tracking and attribution rebuild
Geo-holdout incrementality testing
CRM cleanup and lifecycle automation
Weekly performance reporting to leadership
Engagement Context
Industry
Multi-location restaurant group (fast-casual)
Footprint
30+ locations, Northeast US
Inherited State
225 legacy campaigns, four prior agencies
Primary Challenge
Fragmented paid media with no trustworthy read on performance
Channels
Google Ads (Search + PMax), Meta, LinkedIn
Engagement Type
Fractional CMO + paid media management
Status
Ongoing
Find out what your ads actually returned.
Start with the free reconciliation check. Three steps, thirty minutes, your own accounts, and no pitch until you have seen your own numbers.
Get the free reconciliation check