Results · What I changed

What I changed.

And what it caused.

Not a portfolio and not a list of skills. Every account here carries its problem, what I found or changed, and what came out of it.

Book a no-obligation callWho is behind this
  • 50+clients over 8 years
  • 3+ yearslongest engagement
  • $75M+invested in PPC
  • 12+ marketsCZ, SK, DE, UK, US, AU and more
01From audit to result

Four accounts: what I found and what came out of it

Three studies without client names, the numbers come from Google Ads, with the window and the caveats under each study. Matchaday.sk is named and has no numbers, as the client asked.

  1. 01 · Online store for haberdashery with its own laser production+33%ad revenue at the same spend
  2. 02 · Online store for barefoot shoes (Czechia and Slovakia)+150%ad revenue, summer on summer
  3. 03 · Entertainment center (online booking)0 → 28tracked bookings in the first 4 days
  4. 04 · Matchaday.sk · Slovak online store selling matchaMarket or account?audit of falling orders with read-only access

01 / 04Online store for haberdashery with its own laser production

Same ad spend, 33% more revenue from ads.

+33%

ad revenue at the same spend

Starting point (took over on Apr 28, 2026)

  • For January to March 2026 the account reported a return of 3.32. The real figure was 1.66: every purchase had been counted twice from November 2025 to April 2026.
  • Over those six months that added up to 500,856 CZK in revenue that never existed.
  • A campaign on competitor names cost 5,944 CZK a month and brought not a single sale.

What I fixed

  1. Apr 29switched off the competitor campaign.
  2. May 21purchase tracking fixed, one purchase = one conversion.
  3. Jun 5new campaign structure built around margin, own production kept separate.
  4. Aug 13brought back 6,520 own-production products that had silently dropped out of ads after a category was renamed. Advertised products went from 1,532 to 7,645.

Result

Ad spend+0.64%
Jan–Mar 2026: 133,958 CZK, Jun–Aug 2026: 134,812 CZK
Revenue from ads+33.22%
Jan–Mar 2026: 222,349 CZK, Jun–Aug 2026: 296,205 CZK
Return on ad spend (ROAS)
Jan–Mar 2026: 1.66, Jun–Aug 2026: 2.20
Cost of sale (ad spend ÷ revenue)
Jan–Mar 2026: 60.25%, Jun–Aug 2026: 45.51%

From June to September, 23,775 CZK was not spent on the competitor campaign.

A note on the numbers: January to March revenue is corrected for the double-counted purchases. Winter months are seasonally stronger for this store, so the improvement is not explained by season.

Full study

02 / 04Online store for barefoot shoes (Czechia and Slovakia)

Summer 2026: ad revenue +150% on summer 2025, with costs under the agreed cap.

+150%

ad revenue, summer on summer

Starting point

  • The client wanted growth, but with ad costs under 20% of revenue.
  • Roughly two thirds of ad visits (65.05%) could not be attributed to a campaign in Google Analytics.
  • Ads on the brand name were also burning money inside the shopping campaign: 529 CZK over 30 days at a return of 1.11.

What I fixed

  1. Apr 22brand separated from the shopping campaign.
  2. May to Juneshopping campaign budget doubled, a new category campaign and the brand launched in Slovakia.
  3. Jul 10brakes on at the cap. More money no longer brought more orders, so I moved the budget back to where it earned.
  4. Jul 24fixed the measurement of visits from ads.
  5. Sep 9switched off campaigns that spent 13,487 CZK in 28 days and returned 13,832 CZK.

Result

Ad spend+289%
Jun–Sep 2025: 48,650 CZK, Jun–Sep 2026: 189,218 CZK
Revenue from ads+150%
Jun–Sep 2025: 436,151 CZK, Jun–Sep 2026: 1,090,330 CZK
Cost of sale
Jun–Sep 2025: 11.15%, Jun–Sep 2026: 17.35%

A note on the numbers: Cost of sale rose from 11.15% to 17.35%, still within the agreed 20%. You pay a higher cost share for 2.5 times the revenue. September is still being settled.

Full study

03 / 04Entertainment center (online booking)

From zero to 28 tracked bookings in the first four days.

0 → 28

tracked bookings in the first 4 days

Starting point (audit on Oct 5, 2026)

  • Over 90 days 26,831 CZK went into ads, with 0 tracked bookings.
  • 23.85% of the spend (6,397 CZK) went to campaigns meant to find bookings on their own. None of them could see a booking.
  • Tracking codes ran on the site three times side by side. One pointed at an account that does not exist.
  • The same query with the venue name ran in two campaigns: once at 9.26 CZK a click, once at 0.12 CZK.
  • Corporate events, the client's main goal, had no campaign at all.
  • All findings with their impact and the fix plan are in the sample audit of this account.

What I fixed (Oct 8 to 13)

  1. I linked Google Analytics to Google Ads and set the currency.
  2. Google Ads now sees every online booking.
  3. One campaign for the venue name instead of two.
  4. A new campaign for corporate events and one for the venue itself.
  5. Remarketing that brought nothing is paused.

Result so far

  • Oct 12 to 15: 28 tracked bookings for 1,979 CZK of ad spend, which is 70.69 CZK a booking.
  • I will add the one-month result in November.

A note on the numbers: Bookings were not tracked before the fix, so 28 bookings cannot be compared with an earlier period. It is the first four-day window and conversions are still being settled.

Full study

04 / 04Matchaday.sk · Slovak online store selling matcha

Why orders were falling: weakening demand and frequent changes in Google Ads

Market or account?

audit of falling orders with read-only access

What the client was dealing with

  • Online store orders had been falling month after month since spring.
  • It was unclear how much of the drop came from the market and how much from changes in the account.
  • The client needed an answer backed by data that would show what to change in the account. And in time for Black Friday and Christmas.

What I found

  1. MarketDemand for matcha is shrinking year over year after last year’s boom. Search volume has been below last year since spring, and it is not just seasonality: in the same months last year it grew. The store is holding its market share.
  2. AccountWithin a few months the account went through a series of changes to return targets and budgets, all meant well to improve return on ad spend, but together they threw the bidding algorithm off. After a target change, performance takes one to two conversion cycles to settle, and the campaigns never got that time between changes. Strict return targets and lower budgets held the campaigns back exactly when the market was weakening and the store needed to invest its way out of it.
  3. What did not cause the dropNeither purchase tracking nor errors in campaign settings explain the drop. The account does not need to be rebuilt.

What follows from it

  • Clear causes: the client knows why orders fell and can separate the impact of the market from changes in the account.
  • No rebuild: the brakes were targets, budgets and the pace of changes, not errors in the account.
  • What to change: the client knows what to change in the account and handed the recommendations to the team that runs the campaigns.
„Potrebovali sme oddeliť vonkajšie vplyvy trhu od zmien v účte. Karel nám za pár dní dodal krátky a prehľadný report podložený dátami a na spoločnom hovore ho do hĺbky vysvetlil. Teraz vieme, prečo objednávky klesli a čo treba v účte zmeniť. Za nás stopercentná spokojnosť.“
Mgr. Martin Šottníkowner and founder of Matchaday.sk

A note on the numbers: At the client’s request, the study does not include any numbers from the account or the store. The audit was done with read-only access. Changes in the account are rolled out by the team that runs the campaigns.

Full study

See what the audit this starts with looks like
02Monthly report

What you get from me every month

Just the core numbers and, above all, what changed: how many crowns I saved, and how much efficiency or revenue grew.

Sample data, the numbers are made up

Monthly Google Ads report

September

September: 38,400 CZK more revenue from ads at the same spend.

  • Ad spend62,000 CZK+1% vs August+12% vs last September
  • Revenue from ads248,000 CZK+18% vs August+31% vs last September
  • Cost of sale25%−4 pp vs August−4 pp vs last September
  • Orders from ads214+15% vs August+22% vs last September

What I changed this month and what it did

  1. When: Sep 3Change: Switched off a campaign that had been losing money for 2 monthsImpact: saved 9,800 CZK a month
  2. When: Sep 12Change: Brought back 1,200 products that had dropped out of the feedImpact: +21,600 CZK revenue in 18 days
  3. When: Sep 24Change: Higher return target for regular goodsImpact: cost of sale on regular goods 31% → 24%

Revenue and ad spend by week, last 6 months

  • Revenue from ads
  • Ad spend
123AprMayJunJulAugSep

The numbers in the circles match the changes listed above.

What I am working on in October

  • A campaign for the Christmas range, starting Oct 20.
  • A test of new product titles on 300 items.
03Main case study
HappySport logo

HappySport / Freeskier

Outdoor equipment store

Growth above 30% with unchanged cost of sale

Result

+30%

Cost of sale unchanged

1Problem

Campaign structure did not match the range, and the feed lacked the attributes people search by.

2Change

Audit, structural rebuild, then product feed optimization.

3Result

Growth above 30%, cost of sale unchanged. Two more stores followed after the first year; we are in the fourth season.

4Measurement conditions

Tracked metric
Cost of sale unchanged
Scope
Two more e-shops after the first year
Source
Petr Křivánek, Owner, 2026
This is our fourth season working with Karel. We started with an audit and changes to campaign structure, then moved on to optimizing the product feed. The result was growth of over 30% at an unchanged cost of sale. After the first year we handed him our two other stores as well.”
HappySport logo

Petr Křivánek, Owner, 2026

HappySport / Freeskier

Petr Křivánek, Owner, 2026 · HappySport / Freeskier
04More interventions

Changes and their effect

Each row is one account. The problem is where it started, the change is my work, the result is a number and the period it covers.

3.6 → 4.7Winter sports storeROAS 3.64 → 4.72 for the same months two years later. Orders from ads +51%, revenue +55%.
Problem
In early 2024 the account stopped tracking purchases, and the season ended at a ROAS of 3.64 for October to December.
Change
Took over management before the 2024 season.
3×Store after eighteen months elsewhereThree times the orders at a ROAS above 5.
Problem
Budget poured into a structure that could not spend it well.
Change
Rebuilding the structure before touching the budget.
7.58Store, 14 months inROAS 7.58 with orders tripled over the year.
Problem
Budget split evenly, regardless of where a click pays for itself.
Change
Tactical campaigns where the margin carries the cost of a click.
78 → 141New store, starting from zero78 orders in month one, 141 in month two at a ROAS of 4.27.
Problem
No history and no data for the algorithm to learn from.
Change
Building the account, the tracking and the campaigns from scratch.
3,366Catalog with thousands of products3,366 products rewritten in a day. More accounts run the same setup.
Problem
Feed titles and attributes did not match how people search, and editing the store meant waiting on a developer.
Change
A supplemental source for Merchant Center, with no change to the store.
3,000International marketplace3,000 granular PLA campaigns across 12 countries in a week. After a structural rebuild and feed work, revenue from US PLA campaigns grew by roughly 40%.
Problem
Hundreds of thousands of items across more than ten markets. Manual management makes no sense at that scale.
Change
Building campaigns through the Google Ads API instead of clicking through the interface.
3 of 5Five inherited accountsThree of the five were optimizing toward a number that did not mean revenue. Not any more.
Problem
The Conversions metric mixed purchases with store visits and add-to-cart.
Change
Sorting the conversion actions out before any optimization.
05Case studies

Case studies from feed enrichment

Four client accounts, measured against a control group of products from the same catalog. Each study links to the full report, which is in Czech.

Case study #1

Client PMax account

48 products22 days

Measured metric

Conversion value

+719%

The +719% is conversion value, not clicks. Clicks went from 480 to 1,582, gross; net of the control group trend, +260%. The enriched products grew well above the rest of the catalog.

Open the full report (in Czech)

Case study #2

Client PMax account

138 products22 days

Measured metric

Clicks

+144%

I'm not reporting revenue yet. The conversion sample is too small; I'll add it once the data is statistically significant.

Open the full report (in Czech)

Case study #3

Client PMax account

193 products22 days

Measured metric

Clicks

+394%

Enrichment generated broader search queries. One subcategory maintained exact matches, while the other attracted more generic queries at the cost of a lower CTR. Conversion data will be added later.

Open the full report (in Czech)

Case study #4

Client PMax account

620 products13 days

Measured metric

CTR in the winning subcategory

+41%

The average masked two contrasting results. Splitting by subcategory showed one drove volume up, while the other was diluted by more generic search queries. This is why I trust the details more than the average.

Open the full report (in Czech)
1 / 4
06Experience

Where I learned this

Besides my own clients, two large brands.

01

Alza.cz

Consumer electronics, accountable for revenue and return. A large catalog and margins where a bad decision shows within a week.

02

International marketplace

Today I lead Shopping campaigns and product data across more than ten markets. Hundreds of thousands of items and thousands of granular campaigns built through the API.

03

My own clients

Online stores with budgets from $10k to $450k+ a month. Fashion, electronics, automotive and supplements.

Why there are no names and logos

Most clients have a confidentiality clause. So I quote the numbers without naming the store. A client is named only where they explicitly allowed it.

For the feed enrichment case studies I measure against a control group of products from the same account and subtract the seasonal trend. Where the data will not carry that, I say so next to the number instead of claiming credit for it.

07Contact

Want numbers like these on your account?

I reply within 24 hours. The call takes 30 minutes.

On the call I will tell you what I would fix first and whether it is worth doing at all.

This consultation is non-binding. By submitting, you agree to the processing of personal data.

Prefer to reach me directly?