How Liquor Stars Grew to $15.4M in Google Ads Revenue at 10.4x ROAS

The Client

Liquor Stars is an online spirits retailer shipping nationwide across the United States from a Shopify storefront. The catalogue is built around allocated and hard-to-find bottles: Pappy Van Winkle, Blanton's, W.L. Weller, Eagle Rare, Clase Azul, Fortaleza, Johnnie Walker, and the top shelf of bourbon, scotch, cognac and tequila.

When the website launched in July 2021 the store was doing roughly $30,000 a month in revenue. It had no purchase history, no Merchant Center track record and no advertising data to learn from. Sixty-two months later the Google Ads account alone had produced $15,365,248 in revenue on $1,481,826 of spend, a lifetime ROAS of 10.4x across 20,820 orders.

"$30,000 a month to a $1.5 million month. Sixty-two months, one account, one team."

The Challenge: Google Ads for an Online Liquor Store

Alcohol is one of the hardest categories to advertise profitably on Google, and a brand-new store has none of the advantages an established one can lean on:

• A regulated category. Alcohol listings are flagged by policy by default in Merchant Center. Every product, every feed update and every new campaign has to survive policy review before it can earn a single click.
• No history to bid on. Smart Bidding and Performance Max learn from conversion data. A new store has none, so the first months go into building the signal the algorithm needs before it can be trusted with real budget.
• Eleven states the store cannot ship to. Alcohol cannot legally ship to UT, MS, AL, AR, DE, OK, MI, RI, TN, AK or HI. Every campaign had to exclude them from day one and keep excluding them as new campaigns were added, because a single positive city inside an excluded state will still serve ads.
• Scarcity economics. The most valuable bottles are allocated, which means they sell out, stay out and reappear without warning. Advertising has to move with inventory, not against it.
• Bigger competitors. National retailers with far larger budgets bid on the same bottles.

Our Approach: Structure that Protects Margin

Google Ads carried roughly 90% of the traffic and revenue, with Microsoft Ads and Meta Ads run alongside it. Inside Google, the account was built so that spend could be controlled at the level where margin is actually decided: the individual bottle.

Tiered Performance Max for Alcohol, by Price Band

Rather than one catch-all Performance Max campaign, the catalogue was split into tiers by price so that a $40 bottle never shared a budget or a bidding target with a $400 one: "300 PM | Pmax Campaign | Tier 1", "302 PM | Tier 2" and "304 PM | Tier 3", with a "Residual" catch-all behind them for long-tail catalogue demand.

Brand-Dedicated Campaigns for the Marquee Bottles

The bottles that drive the business got their own campaigns so they never competed with commodity spirits for budget: Pappy Van Winkle, Blanton's Original Single Barrel, Blanton's Gold, W.L. Weller, Weller 12, Crown Royal, Clase Azul, Fortaleza and Johnnie Walker, plus dedicated Tequila, Top Shelf and High Ticket campaigns. Standard Google Shopping for liquor ran for Blanton's, backed by a Supplemental Search campaign, a whiskey Search campaign and Demand Gen.

Dayparting, Bid Modifiers and Experiments

A dedicated "Pmax Campaign 2pm to 8pm" ran only in the afternoon and evening window where spirits buying converts. Every active campaign carried ad schedules and device, age and income bid modifiers. Campaign-level Google Ads experiments ran continuously on Bourbon, Cognac, Scotch, Tequila, Whiskey and High Ticket, so every structural change was tested against a control before it was rolled out.

What Was Built

The Results: Five Years, One Account

The account went from a $7,371 first month to a $1,548,807 December. Year by year, Google Ads only:

Efficiency Rose as Spend Grew

Most accounts lose efficiency when they scale. This one gained it four years running: 10.6x, 10.7x, 11.0x, 11.3x, while annual spend grew roughly eightfold.

The monthly view shows the same thing at higher resolution. Under 4x through the foundation years, then a sustained 7x to 18x range from 2023 onward, with the account's two best months, December 2025 and February 2026, both landing at 18.9x.

Year by Year

2021: A New Store with No History

Liquor Stars launched with no purchase history and no Merchant Center track record in a regulated category. The first six months went into the foundation: conversion tracking, a product feed that could survive policy review, and the state exclusions that keep an alcohol retailer legal. The 2.5x return for those months was the foundation, not the goal.

2022: Finding the Offer

Spend fell, from roughly $10,000 a month to roughly $4,000, while the account narrowed onto the allocated and hard-to-find bottles where a specialist retailer wins on availability rather than price. The turn came in November 2022: the first month above 7x, at 7.96x, followed by 6.6x in December.

2023: Scaling on a Proven Pattern

With the pattern proven, the tiered Performance Max architecture and the brand-dedicated campaigns went live, and spend scaled from $1,123 in January to $30,307 in November. Return went up as spend went up, from 12.45x to 14.29x across that climb. October brought $418,836 in revenue and November $433,049, the account's first two $400,000 months. The year closed at $2,075,572 on $195,260 of spend, 10.6x.

2024: Holding Efficiency Through Growth

Spend rose 26% over 2023 and the return improved with it, to 10.7x for the year. May peaked at 17.7x. This was the year the layers compounded: the negative keyword lists, the dayparting, the demographic and income bid modifiers, all working on top of a structure that already fit the catalogue. December delivered $428,540 on $52,791 of spend.

2025: The Best Year

Spend doubled to $486,098 and the account returned 11.0x, its best full year yet, with 11 of 12 months above 7x (May 11.6x, July 10.2x, November 10.9x). Then December: $1,548,807 in revenue on $81,903 of spend, 18.9x, across 1,204 orders. The gifting-season thesis, that a premium spirits store should lean hard into Q4, was proven in a single month.

2026: A Category Correction and a Business that Held

The American whiskey market turned in 2026 (see the obstacles below). Even so, January produced $1,163,177 on $82,140, 14.2x, and the eight months from January to August returned 11.3x on $443,750 of spend, the highest annual figure in the account's history.

The Obstacles

Five years in a regulated category is not a straight line. These are the problems that had to be solved to keep the numbers above.

1. Merchant Center Disapprovals in a Regulated Category

Alcohol listings are flagged by policy by default. At one point 2,290 products carried alcohol policy flags. On 24 June 2026 a Google-side change disapproved around 104 products overnight with nothing but "product page unavailable" as the reason, and it hit other liquor stores the same day. Keeping a spirits feed live is continuous work, not a setup task.

2. Shopify Was Rewriting Product Titles

Around 90% of product titles had "| Liquor Stars" appended by Shopify's own Google sales-channel sync, confirmed through the raw Merchant Center API, which recorded the modification source as third party, Shopify. Google Shopping keys performance history to the title string, so every renamed product became a new item with zero history, splitting years of conversion data across two labels for the same bottle. We diagnosed it to the source and corrected it with a supplemental feed.

3. Paid Traffic Reporting as Organic

Shopify's Google & YouTube channel app injects its own tracking parameters (utm_medium=product_sync and utm_campaign=sag_organic) into feed URLs. Those collided with the account-level tracking template; analytics read the first duplicate parameter, so Shopping and Performance Max traffic was being reported as organic search in the store's own reporting. We removed the account-level suffix and re-applied it at Search-campaign level only. After the fix, Google-sourced sessions were flat to up, roughly 1,350 a day to 1,400, and the apparent "drop" in the cpc column was simply Shopping traffic finally landing in its own bucket.

4. Stockouts on the Best Sellers

High-value SKUs repeatedly went to zero with no resupply: W.L. Weller Single Barrel, Eagle Rare 10 Year, Old Rip Van Winkle 10 Year, the Blanton's Full Lineup bundle, Macallan 30 Double Cask, Crown Royal XR, Hibiki 21, Weller C.Y.P.B. and Blanton's Black. Separately, some listed SKUs absorbed ad clicks for a full year with zero sales: Blanton's Gold 750mL took 197 sessions and sold nothing in 365 days, and Patron XO Cafe 50ml took 112 sessions for the same result. We built a product-level report cross-referencing ad spend against sell-through so dead SKUs could be excluded before they burned more budget.

5. Conversion Rate Work Outside the Scope

Advertising can only deliver the click. Landing page structure, above-the-fold clutter, product page trust signals and checkout friction were audited and advised on beyond the advertising engagement, at no additional cost, because a better-converting store makes every dollar of ad spend work harder.

6. Google's 17 August 2026 Bidding Change

Google changed its target-based bid strategies so that budget-limited campaigns track their stated target instead of overdelivering above it, a change covered by Google Ads Help, Search Engine Journal and Optmyzr. Seven of the account's eight live campaigns at the time were budget-limited, so the change reset the bidding model. Targets were re-established campaign by campaign, never in one move, because a single-step jump earlier in 2026 had nearly killed volume.

7. The Bourbon Bubble Deflated

Kentucky distillers were holding a record 16.1 million barrels of aging whiskey. US whiskey volume was down roughly 4.9% and revenue 5.1% in the twelve months to July 2025. MGP reported whiskey sales down 59% and described the category as structurally oversupplied (WhiskeyPulse). The secondary market fell around 11% (Bourboneur) and Blanton's started turning up locally at around $90 (Bourbon Culture). Tariffs cut Canadian exports and pushed product back into the US market, and category price/mix went from +0.3% in January to roughly -5% by spring: industry-wide discounting. Meanwhile craft tequila grew 28.5% while mass-market brands declined (Park Street).

The response was strategic rather than tactical: diversify out of mid-tier allocated bourbon and into the categories still growing, tequila above all. That is the market the account returned 11.3x in.

Further reading on the 2026 whiskey market: Forbes on the headwinds facing American whiskey, Forbes on 11 trends to watch and The Bourbon Road on where bourbon is headed.

Why It Worked

• Structure protected margin at every level. Price tiers, brand-dedicated campaigns, dayparting and a catch-all for the long tail meant every dollar was bid in the context of what that particular bottle could return.
• Return rose four straight years while spend grew roughly eightfold. 10.6x, 10.7x, 11.0x, 11.3x. Scaling was done on proven patterns, tested in experiments, never in one jump.
• The record year happened in the worst whiskey market in a decade. The December 2025 peak and the 11.3x for 2026 came after the category turned, not before it.

Key Takeaways

1. Performance Max for alcohol works when it is tiered. One catch-all campaign cannot bid a $40 bottle and a $400 bottle intelligently. Splitting the catalogue by price band, and giving the marquee brands their own campaigns, is what makes spend controllable.

2. A spirits feed is never finished. Policy flags, platform-side disapprovals, title rewrites from the sales-channel sync and tracking collisions all happened after launch. The account held because they were caught and fixed, every time.

3. Efficiency and scale are not a trade-off. Most agencies will tell you that scaling means accepting a lower return. This account grew spend roughly eightfold while its annual return rose four years in a row.

4. When the category turns, the answer is strategic. Bidding tweaks do not fix an oversupplied market. Diversifying into the categories still growing did, and the account posted its highest annual return in that market.

About This Campaign

This account was managed by Live PPC Ads from launch through August 2026 across Google Ads, Microsoft Ads and Meta Ads.

Industry: E-Commerce: Wine & Spirits Platforms: Google Ads (Search, Shopping, Performance Max, Demand Gen, Display), Microsoft Ads, Meta Ads Location: United States (nationwide shipping) Account Duration: July 2021 to August 2026 (62 months)