VK Ads Case Study: Adult English Leads for 61–768 RUB
About the project in this case study
Alyona Yermolina is a strong expert in her niche. She came up with both a lead magnet and an intensive course. She’s been on the market for a long time, and that’s a big part of why this case worked.
At the same time, the “Adult English” project is relatively fresh and new for her practice.
Here’s what the cover of the VK public page looked like at the time.
And here’s the first screen of the website:
What was the goal
Bring in leads to GetCourse (a Russian LMS + course platform with payments and email automation) to sell the intensive.
The GetCourse link arrives by email after registration on the website https://englishalena.ru/ via a GetCourse widget.
So we needed as many people as possible to fill in the form, and then be detected as new and active registrations inside GetCourse.
The lead magnet is a couple of lessons on that same platform.
Task parameters
For the intensive to pay off, we forecast these metrics:
Click — no more than 30 RUB (Russian roubles).
Click-to-lead conversion — 23–30% and higher.
Then, if at least 10% convert into purchases at a 3,500 RUB ticket for the intensive, we recoup all traffic costs and get a small profit on top.
Alyona took an archived plan (a discontinued package), simply to test what we could do. Fee + ad budget added up to 35,000 RUB in her total spend.
The service was purchased on Wednesday, January 11, 2023. For three weeks, meaning everything ended on 01.02.2023.
What results we achieved
On the screenshot you see the peak value of registrations in GetCourse. Overall, it’s a slice from January 11 to February 02.
I read that activity as the result of paid traffic.
Some activity also came from Alyona’s parallel work, but it’s hard to isolate. You can still see that before January 18 there was a calm period, same as between January 22 and 24. Some users stayed active there, but there were no new registrations.
Those were the days when we were doing analytics and technically improving the ads to push the outcome.
In total: 165 new registrations.
94 active users. Here “active” includes both new and old registrations.
If we sum all active users and divide by all active days (17), we get 5.53 active users per day on average. The peak is more telling: 16 active users in a single day.
If we treat new registrations as leads, then we got 165 leads. If we treat “active users” as leads, then we can’t sum them the same way. In that interpretation, the “lead” is 16 leads (the peak daily activity).
There were 3 unsubscribes from all email campaigns during this period.
In total, averaged out, the click price was 31.64, slightly above the click KPI.
The cheapest click price (averaged within one campaign) was 10.47 RUB.
The most expensive click price was 82.54 RUB
That was a hypothesis test that didn’t work.
The most effective setup was 2 audiences with one creative. These were members of 2 English-learning groups with 3,000+ subscribers — direct competitors.
Meaning: they were subscribed to at least 2 competitor communities.
We also used keywords related to learning English inside VK (major Russian social network).
spoken English for beginners, English conversation club, spoken English lessons, spoken English from zero, spoken English courses, learn conversational English, speak English, English lessons for beginners
These produced clicks within the KPI range.
Here are the most effective combinations
We found them almost immediately.
We started with Test Gen 1, looking for the best creative
We also installed Yandex.Metrika (analytics tool by Yandex) on the site, and with link tagging we tracked which exact combinations drove the most clicks on the “Get the course” button.
We tested right away using an “audience carpet” (a broad grid of audience segments for fast comparative testing; RU slang). That’s why you see those numeric prefixes in the audience names — they’re for analysis via UTM tags.
This is how it looks in Metrika
To squeeze extra results, we also launched retargeting:
The most effective retargeting setup, surprisingly, was targeting people who had already clicked and visited the site or communities.
And yes, a side bonus of this activity: the VK public page grew by 98 subscribers.
It was around 28–30 at the start of our work.
Let’s sum the results up in a table:
If at least 10% of the 163 people still on the list buy the intensive, profit will be far above all costs. Roughly a 200% outcome.
So around half of the traffic converted into leads.
If “leads” are defined as the peak “active users” minus 3 unsubscribes, then only 4% converted into leads from traffic. In that model we would need a 60% purchase conversion rate just to break even.
That’s where the spread in lead cost from the title comes from. It’s unclear what exactly we should treat as a lead.
We lean toward treating it as 163.
Sure, it looks nicer for a case study :)) And there’s a more important reason.
Anyone who registers gets into the GetCourse email list. They might become active later, meet the product, or even buy the intensive without touching the two free lessons.
Email marketing shouldn’t be dismissed here.
SUMMARY
The target + retarget bundle is workable. Over time, the competitor-based audiences will need a refresh, because they burn out and exhaust themselves.
The keyword-based audience is homeostatic. It can “hold itself” without constant re-tuning. It depends on how much interest there is in English-related queries inside VK’s ecosystem.
The offers and creatives turned out to be attractive to people.
Here’s the retargeting post that helped close the loop: https://vk.com/wall-212210704_168
+ we attached a video from VK Clips (short video format inside VK, similar to TikTok/Reels)
And here’s the one that collected the primary clicks and first touches: https://vk.com/wall-212210704_165
+ we also attached a video from VK Clips
There are accumulating datasets (retargeting audiences) that will make clicks cheaper over time, like “Visited” and “Watched to the end”.
The public page itself has also accumulated audiences to build its own lookalike audience based on similarity to page subscribers.
So yes — we know exactly what to do with the traffic next.
What new audiences to try, and how to gradually strengthen the product’s marketing side, because one of the most relevant audiences here is the competitor audience.
The more you monitor competitors, study the market, and work on marketing side-by-side with traffic, the higher the result will be.
The question is whether Alyona will come back. I asked her to leave a review, but she hasn’t rushed. On the call where I walked her through the report, she spoke very positively about the work.
She said she didn’t expect this level of professionalism.
I also added cheat sheets to the report so Alyona could independently increase the odds of sales when it’s time for the intensive.
Credits:
Our team, The Quiet Orbit: https://vk.com/bo_target
Me: https://vk.com/the_redbeard
Rita, as the project manager at the time: https://vk.com/hikaru_snow
Elena, as the targeting specialist, until the end of January 2023: https://vk.com/snegnaja_zima
Alyona Yermolina: https://vk.com/alena_english_sp and https://englishalena.ru/
If you want to reach out for a service, you can message me here: https://t.me/bo_martian