In short
- The account scored 47 out of 100 on an 80 check audit, with up to USD 2K a month in waste and every lead valued at zero.
- LanguageBird's leads convert on average 4.7 months after they arrive. Campaigns were being paused on 30 day reads of a half year decision.
- Reread on the real cycle, paid search revenue stands at USD 30K, and the client's own current read is a 4x ROAS.
Revenue hiding in plain sight
Some revenue hides in plain sight. LanguageBird's paid search had been producing enrollments for months. The reports couldn't see them because the reports were looking at the wrong window of time.
An online school where the buyer is a family
LanguageBird is an accredited online language school. Live, one to one instruction, a catalog running from Spanish and French to Mandarin, Korean and American Sign Language, students all over the United States.
The buyers are families, and families deciding on their kid's education don't click an ad on Tuesday and enroll on Wednesday. They research, they compare, they talk it over at dinner, they come back a month later.
What this meansThat habit sounds like a footnote. It's the whole case.
The audit: 47 out of 100
We ran the Google Ads account through our 80 check audit. First read: 47 out of 100. The specifics were rough.
- Between USD 1,500 and USD 2,000 a month, 13 to 18% of the budget, going to waste
- The top spending keyword burning over USD 1,100 a month with a quality score of 1 out of 10
- Six or more keywords cannibalizing each other across three to four campaigns at once
- Three zombie campaigns with past end dates, still enabled, still spending
- USD 500 to 700 a month bleeding into competitor brand searches
- Every lead entering the account recorded with a conversion value of zero
That last bullet did the most damage. With every lead worth USD 0 and enrollments closing months after the click, paid search looked like a cost center, and campaigns were getting paused mid flight based on 30 day reads of a decision families take half a year to make.
An account like this can be producing real revenue and still look like a failure. It just needs its measurement calibrated to the wrong buying cycle.
What this meansThe waste was real and fixable. The zero value on every lead was the error that hid the revenue.
First the cleanup, then the clock
The cleanup was standard surgery. Pause the quality score 1 bleeder, kill the zombies, give each keyword one home, build negative lists to stop the competitor bleed, move the recovered budget to campaigns earning it.
The clock was the interesting part. We rebuilt how revenue gets attributed, tying actual enrollments back to the lead that started them across the real decision cycle. The number that fell out: LanguageBird's leads convert, on average, 4.7 months after they first arrive.
What this meansBefore deciding a channel doesn't work, check whose calendar you're measuring it on.
What the account looks like now
The current numbers hold up under any read. Between April and late July, the account generated 402 leads at USD 72 each, off a 9.4% click through rate and an 8.1% conversion rate. A 9.4% CTR on search is the kind of number we double check before putting in a report. It held.
And the client's own current read on return: a ROAS of 4x. Their number, from their revenue data, not a platform estimate.
Growth work is running on top of the clean base: dedicated campaigns for Texas, California and Florida, where demand concentrates, and conversion values feeding the account so bidding can eventually optimize toward revenue instead of raw form fills. One more layer: we track LanguageBird's visibility across ChatGPT, Gemini and Perplexity, watching which pages the AI engines cite when families ask where to study a language.
What this meansThe 4.7 month figure sits in every performance conversation now. Campaigns get judged at the end of their incubation instead of the middle.
The takeaway
LanguageBird's paid search was profitable before we showed up. The difference is that now everyone can see it, the client confirms it at 4x, and the budget behaves accordingly.
- Audit what is actually measured before touching a single bid
- Give every lead a conversion value, even an estimated one
- Measure the real time from lead to enrollment, and judge campaigns on it
- Run paused campaigns again with a trial period as long as the buying cycle
