← Back to case studies
Google Ads · Premium wheels & tires · Kyiv

WheelShopROAS from 6.0 to 9.7 on Google with 23% less spend

A Kyiv store that sells premium wheels and tires and has its own forged-wheel brand, WS Forged. Before we started, two thirds of its Google Ads budget went into one Performance Max campaign. From December 2025 we replaced it with a structure where each campaign does one job. This covers seven months, December 2025 to June 2026, compared with the three months before (August to October 2025).

WheelShop, Google Ads · Premium wheels & tires · Kyiv
6.0 → 9.7

ROAS, before vs the seven months with WeBrandLab

−61%

Cost per conversion, ₴6,686 → ₴2,610

5.7 → 11.3

Conversions a month, at spend of ₴38,087 → ₴29,485 a month

6 of 7

Months above the 3.33 target ROAS (May: 2.8; strongest month: 16.3 in March)

Google Ads, December 2025 to June 2026 vs August to October 2025, UAH. A conversion is a purchase counted by the GA4 purchase event, in the same way in both periods. November 2025 is excluded: it had ₴12,145 of spend and no conversions. These are ad platform figures; store data is not included.

Monthly average · before vs. seven months with WeBrandLab

Aug–Oct 2025 → Dec 2025–Jun 2026

Spend, ₴ a month-23%
38,087→29,485 (−23%)
Conversions a month+98%
5.7→11.3
Conversion value, ₴ a month+24%
230,411→284,936
ROAS+62%
6.0→9.7
Cost per conversion, ₴-61%
6,686→2,610 (−61%)

Context

WheelShop is a Kyiv store of premium cast, forged and flow-formed wheels and tires, with a showroom for fitting, painting and seasonal storage. What sets it apart is WS Forged, its own forged-wheel brand made to order for a specific car. Buyers own expensive imported cars and buy slowly: they check fitment, compare prices and call the store. WeBrandLab has run its Google Ads since November 2025; Meta Ads is a separate case.

The challenge

From August to October 2025, 66% of spend (₴74,883 of ₴114,262) went into a single campaign, Performance Max Start New. A search campaign for wheels by radius spent ₴12,743 without a single conversion. Together that gave 5.7 conversions a month at ROAS 6.0. The target agreed with the client was ROAS of at least 3.33.

What we did

  • Replaced the old set of campaigns in December 2025 with a clear structure: Performance Max on the whole wheel catalog, a separate PMax for Mercedes, and a PMax for all of Ukraine except Mercedes and BMW.
  • Launched Shopping in March 2026 with a target ROAS, so budget can follow the product and its margin.
  • Kept a separate brand search campaign for people who already know the store, and a remarketing campaign for BMW and Mercedes.
  • Closed the tests that did not work.

Outcome

Over seven months the account returned ₴1,994,550 in purchases on ₴206,392 of spend: ROAS 9.7 and ₴2,610 per conversion, against ROAS 6.0 and ₴6,686 before. Monthly spend is 23% lower than in autumn 2025, and conversions doubled, from 5.7 to 11.3 a month. The three strongest months (December 2025, February and March 2026) averaged ROAS 12.8 and 16.0 conversions a month. Six of seven months are above the 3.33 target; May is the exception at 2.8. After the case period the result dropped. ROAS was 3.4 in July 2026 (9.0 conversions), 2.0 in August (2.0 conversions), 3.7 in September (1.0) and 4.4 in the first 15 days of October (1.0). In August clicks were 16% above July while conversions fell, and a quarter of August spend went to YouTube without conversions. After an audit on 29 September we raised budgets to about ₴1,100 a day, cleaned the conversions and widened the geography of Shopping and PMax to 13 locations. It is too early to call a recovery: October so far has one order. Next: close the double counting of purchases, bring the budget to ₴40,000 a month for the high season, move forged wheels into a separate search campaign and connect call tracking from ads.

Chapter 01 · Structure

One campaign carried two thirds of the budget. Now each campaign has a job

The main sales engine was PMax on the whole wheel catalog; Shopping took over from it in spring. Brand search serves buyers who already know the store. Each tile shows spend, ROAS and conversions for December 2025 to June 2026.

Ad spend → ROAS

PMax, whole wheel catalog

₴64,054→ROAS 12.2

25.5 conversions · ₴782,986 value

PMax, Ukraine except Mercedes and BMW

₴39,642→ROAS 9.0

24.6 conversions · ₴354,885 value

Shopping

₴42,746→ROAS 9.8

15.0 conversions · ₴417,105 value

PMax Mercedes

₴17,507→ROAS 10.7

4.0 conversions · ₴186,840 value

Remarketing BMW and Mercedes

₴20,602→ROAS 6.9

5.0 conversions · ₴141,835 value

Brand search

₴9,609→ROAS 11.5

5.0 conversions · ₴110,900 value

PMax on the catalog gave 39% of all purchase value in the period; with Shopping and brand search, 66%. Brand search costs ₴1,922 per conversion, against ₴2,610 across the account.

Chapter 02 · Shopping takes over

When PMax stopped selling, budget moved to Shopping

From April to June PMax on the catalog spent ₴12,444 and brought no conversions (in March: 10.0 conversions, ROAS 21.2). Budget moved to Shopping, which in the same months gave 13.0 conversions at ROAS 9.2. Its monthly spend grew from ₴4,001 in March to ₴17,920 in June without a drop in return.

Chapter 03 · Tests that did not work

₴12,231, or 5.9% of the period's spend

Demand Gen, DSA, Shopping Best-sellers and Search Carfest together spent ₴12,231 and brought no conversions. We switched them off.

Chapter 04 · Tracking

The algorithm should learn on purchases

Until 29 September, dead goals from the old GA4 setup counted as primary conversions, including Begin Checkout with a value of 1. We moved them to secondary; the primary actions are now the purchase and calls from ads that last 60 seconds or longer. Calls are the main way orders come in at this store, so a separate calls-from-ads conversion was added. In October one ₴71,044 order was found counted twice, by GA4 purchase and by Google Shopping App Purchase; the decision on which action stays primary is still open.

Method: in the period of this case the Google Ads Conversions column also included Google Shopping App Purchase (11 purchases) and one add-to-cart action, so the account shows ROAS 11.4. This case counts GA4 purchase only, in both periods.

Timeline · How it actually unfolded

From takeover in November 2025 to the audit in September 2026.

Nov 2025

Start

WeBrandLab takes over Google Ads.

Dec 2025

New structure

PMax on the catalog, PMax for Mercedes, PMax for Ukraine. First month: 13.3 conversions, ROAS 9.8.

Feb-Mar 2026

Peak

35 conversions in two months, ROAS 12.4 and 16.3. Shopping launched in March.

Apr-Jun 2026

Steady level

5 to 9 conversions a month, ROAS 2.8 to 7.4; Shopping brings 13 of the 20 conversions.

29 Sep 2026

Audit

Budgets raised, conversions cleaned, geography widened (after the case period).

Proof · Real dashboards

Account screenshots with notes. Check the numbers yourself.

Screenshots from the client's Google Ads account for the case period. The account number is not shown, and each screenshot has a note.

WheelShop, Google Ads · Account overview (Dec 1, 2025 to Jun 30, 2026): The account counts all conversions that were primary in the period: ROAS 1,136%. Purchases by GA4 only, as in this case: ROAS 9.7, 79 conversions, ₴2.61K per conversion.
Google Ads · Account overview (Dec 1, 2025 to Jun 30, 2026)The account counts all conversions that were primary in the period: ROAS 1,136%. Purchases by GA4 only, as in this case: ROAS 9.7, 79 conversions, ₴2.61K per conversion.
WheelShop, Google Ads · Campaign list (same period): In the account, PMax on the catalog shows 15.34 and Shopping 11.49; counting GA4 purchases only, they are 12.2 and 9.8. Brand search is the same (11.5).
Google Ads · Campaign list (same period)In the account, PMax on the catalog shows 15.34 and Shopping 11.49; counting GA4 purchases only, they are 12.2 and 9.8. Brand search is the same (11.5).

Want this for your brand? Book the 30-min call.

You leave with a 90-day plan whether we work together or not.