Case Study: The Leaking Google Ads Account
This is a case story. Names and sector details are hidden; numbers are rounded, but the ratios are real. We tell it because what we found in this account repeats, with small variations, in most accounts we audit — and because it answers the question foreign businesses ask us most: “how would we even know if our Turkish account is wasting money?”
The client is a mid-sized service business. Three years on Google Ads; a six-figure monthly budget in lira. The complaint was familiar: “The phone rings, but the numbers don’t add up. We don’t know if this is worth the money.”
The key phrase was we don’t know. Nobody had neglected the account deliberately; over the years it had settled into a routine nobody fully examined. The ads ran, the invoice was paid, the questions were not asked — a pattern that grows even faster when the account’s language is one the owner does not read.
Our first move was not touching the ads but taking a photograph: a two-week digital audit. This story covers the photograph, the repairs, and the third-month numbers.
Before the Case: The Missed Signals
BU BÖLÜMÜN ÖZETİ
- No report had ever been requested
- No single number was tracked
- The invoice was never questioned
- The “someone’s watching” assumption
Looking back, the leak had been signalling for years. Four signals foretell the same story in any account — and cross-border, each is easier to miss.
No report had ever been requested
In three years, not one written monthly report existed; the knowledge lived in phone calls. Unwritten knowledge cannot be audited. A client who does not demand a report buys, without noticing, the absence of oversight.
No single number was tracked
Nobody knew the cost per enquiry; the feeling of “the phone rings” had replaced the measure. Regular tracking of even that one number would have exposed the leak years earlier — in any language.
The invoice was never questioned
With spending steady month to month, nobody descended into the line items. But a stable total says nothing about a healthy distribution: waste hides comfortably inside a flat invoice — doubly so in a foreign currency the owner mentally rounds.
The “someone’s watching” assumption
Someone external did look at the account occasionally; but scope, hours and responsibility had never been written. Everyone assumed someone else was watching. Unwritten responsibility is nobody’s responsibility.
The Audit Photograph: Four Leaks
BU BÖLÜMÜN ÖZETİ
- Leak 1: Intent-free searches
- Leak 2: The display network left open
- Leak 3: Out-of-area impressions
- Leak 4: A single, tired ad text
The audit showed roughly a third of the budget escaping through four holes. None was exotic; all were neglect.
Leak 1: Intent-free searches
In the search terms report, a serious slice of the budget flowed to Turkish queries meaning “free”, “salary”, “how to”, “job vacancy”. The negative keyword list had barely been fed in three years. It was the most expensive finding — and the easiest to close. It was also the leak the owner could never have seen: the terms were invisible to anyone not reading Turkish.
Leak 2: The display network left open
A setting surviving from setup day kept part of the search campaigns running on the display network too. Impressions on irrelevant sites to irrelevant eyes: click rate at the floor, conversions near zero. Nobody had opened it on purpose; nobody had closed it.
Leak 3: Out-of-area impressions
The business served two cities; the ads showed in seven. Geographies left over from an expansion trial had been forgotten after the trial ended. Every lira spent on a city the service could not reach was, by definition, waste.
Leak 4: A single, tired ad text
The same ad copy had not changed in over a year. The click rate had slowly eroded, and with it the click price had climbed. A testing culture had never existed: there was no second ad racing, so no winner could ever be chosen.
The Repair Plan: Order and Reasoning
BU BÖLÜMÜN ÖZETİ
- Tracking was repaired first
- The bleeding stopped in one week
- The structure was simplified
- A testing culture was built
The findings went on the table; the plan spread across four weeks. The order was not random: stop the bleeding first, improve second.
| Week | Action | Reasoning |
|---|---|---|
| 1 | Tracking repair + negative list (first 200 terms) | Decision data and the fastest saving |
| 1 | Display network closed, geography cut to two cities | Instant end to defined waste |
| 2 | Structure simplified: 14 campaigns → 6 | Budget control and readability |
| 3-4 | New Turkish copy + A/B routine, bid strategy review | The engine of lasting improvement |
Tracking was repaired first
The account’s conversion counting double-counted some enquiries and missed others entirely. Until the measurement was honest, every improvement would happen in the dark. Week one therefore went to the tracking more than the panel.
The bleeding stopped in one week
Negative list, display network, geography: three moves ended most of the defined waste within the first week. None of these moves requires genius; they require maintenance. The difference was three years without it.
The structure was simplified
The scattered fourteen-campaign build was reduced to a readable six. The goal was not aesthetics: seeing where the budget flows and piling it onto the winners. The logic followed the account structure guide — and gave the foreign owner, for the first time, a panel they could navigate themselves.
A testing culture was built
Every ad group received a second, racing Turkish text; a monthly winner-stays routine went on the calendar. The first month’s winners beat the old copy by a clear margin. The tired text had been costing more than anyone assumed.
Three Months Later: What the Numbers Said
BU BÖLÜMÜN ÖZETİ
- The waste share fell to a five percent band
- Cost per enquiry fell about forty percent
- The report shrank to three lines
- The gain was tied to a routine
At the end of month three, the same photograph was taken again. The budget was not raised; the comparison ran on identical spend.
The waste share fell to a five percent band
Irrelevant spend in the search terms report retreated from a third to around five percent. Not zero — zero is not the target. A small discovery share was left deliberately, so new Turkish keywords could still be found.
Cost per enquiry fell about forty percent
Freed from the waste, the same budget flowed to the winning keywords; enquiries multiplied to 1.7 times on identical spend. The client’s three-year question was answered with a number for the first time: whether it is worth the money is now visible — in a one-page English report.
The report shrank to three lines
A monthly reporting routine was built: spend, enquiries, cost per enquiry; beneath it, the month’s actions and next month’s plan, in English. The client’s comment was worth recording: “For the first time, I read the report and understand it.”
The gain was tied to a routine
The real success is not the third-month number; it is the routine: weekly term scans, monthly tests, quarterly reviews. Every unmaintained account drifts back to the same holes; the routine is what blocks the drift. Its shape sits in the management guide.
Lessons to Take from This Case
BU BÖLÜMÜN ÖZETİ
- Looking alive is not being alive
- The leaks are ordinary — which is why they are everywhere
- Photograph first, repair second
- Ask your own account the same question
The case is specific; the pattern is general. Carry these four lessons to your own Turkish account.
Looking alive is not being alive
This account “worked” for three years: the phone rang, the invoice was paid. The leak never showed because nobody looked. Trust the report, not the appearance — and if no report exists, that is the first problem. Distance and language make the appearance even more convincing.
The leaks are ordinary — which is why they are everywhere
None of the four findings is rare: an unfed negative list, an open network, forgotten geography, tired copy. The ordinariness is good news: both the diagnosis and the repair are known work.
Photograph first, repair second
Repair without an audit is surgery in the dark. The two-week photograph put the four-week repair in the right order. The order was critical: any move made before the tracking was honest would have been a move with unmeasurable effect.
Ask your own account the same question
If “we don’t know if it’s worth the money” sounds familiar, the answer lives in an audit, not a guess. Have your own photograph taken; let a number make the repair decision. Our audit and management structure sits openly on the Google Ads page.
Frequently Asked Questions
Sık Sorulan Sorular
Yes; identities are hidden and numbers rounded, but the ratios are genuine audit findings.
An account that visibly runs while losing a serious share of its budget to intent-free searches, wrong settings and neglect.
Four holes: an unfed negative list, an open display network, forgotten cities and year-old ad copy.
Roughly a third of the budget; within a quarter it fell to a five percent band, with a small discovery share left deliberately.
No; the comparison ran on identical spend. The cut waste flowed to winners and enquiries reached 1.7 times.
Fixing the tracking; with double counts and missed enquiries, nothing could be measured honestly.
No; the same pattern repeats at every scale, and proportionally it hurts small budgets more.
Because the wasteful search terms are invisible to anyone who does not read Turkish; the account looks fine by default.
Through a routine: weekly term scans, monthly copy tests and quarterly reviews, tied to the calendar.
