Anna VoroninaHi! We're here with AdHand, a reseller of Telegram Ads and the BrainStat digital studio. Today we're looking at an interesting case in which the BrainStat team promoted an expensive course on crypto investments and managed to achieve a return on investment two months after launching advertising campaigns in Telegram.
The case study includes not only numbers and strategy, but also Q&A blocks with Mikhail Krotov, CMO of BrainStat.
You'll learn firsthand:
- How to distinguish fake channels from live ones,
- Why using emoji in creatives didn't work,
- Which UTM tags help you dig deeper into analytics,
- And a few other tactics you can apply to your launches.
About the project
BrainStat worked with a client who was selling an educational course on crypto investments worth 180,000 roubles. The peculiarity of the product was the teaching of modern methods with an emphasis on saving time. Students could start investing with a minimum sum of 1,000 roubles, without having to immerse themselves in complex analytical calculations or spend all day in front of a monitor.
The product funnel consisted of three stages:
- Free webinar of 90 minutes as an entry point
- Tripwire for 2500-2700 rubles to check the solvency of the audience
- Basic course for 180,000 rubles.
Between the purchase of tripwire and the flagship product, the client spent 3-4 weeks warming up through tutorials and personal consultations. Conversion from tripwire to the flagship course was about 20%.
The target audience is men 25-45 years old, interested in cryptocurrencies and investments, looking for a ready-made strategy without complicated calculations and with minimal time costs.
Crypto Niche Challenges
The BrainStat team faced the typical challenges of the crypto segment.
The first challenge is the huge number of fake channels with inflated metrics. A channel may show 50,000 subscribers, but real post views do not exceed 200-300, which gives an ER of only 0.4%.
The second challenge was an overheated audience that saw thousands of offers for crypto education. Standard creatives with an expert's photo and appeals to "make money on crypto" showed a CTR of only 2-3%.
The third challenge is the high cost of engagement, which in the crypto niche is 2-3 times higher than other educational projects.
- How have you learned to recognize fake channels? Are there any "red flags" to watch out for?
Use Telemetr and pay attention to the following details:
History of changes in the name and description of the channel,
channel administration contact (if the contact is the same in several channels, it may be a mesh),
the dynamics of subscriptions and unsubscriptions,
advertising placements in the channel.
Telemetr also marks "red" the channels that have been bumped up.
Strategy for selecting channels for advertising placement
Faced with the problem of low quality traffic due to fake channels with inflated indicators, BrainStat radically revised their approach to selecting Telegram channels for placement.
Transition from quantity to quality
Instead of mass purchases, the agency implemented a strict channel filtering system in order to get quality traffic instead of clicks from bots.
Channel Selection Criteria:
- Minimum Engagement Rate (ER): 30%
- Optimal ER: 50% and above
- Manual check of each channel
- Using Telegram channel directories: TGStat and Telemetr
- Evaluation of activity in comments
- Search for channels and bots in which competitors advertise: spy-service Telegium.
Audience segmentation
The team structured channel outreach according to their audience's interest in the product:
Hot segment - subscribers of competitors and specialized cryptocurrency education channels. This audience is already familiar with educational products in the niche and is ready to buy.
Warm segment - subscribers of trading channels in general. The audience is interested in financial instruments, but requires additional warming up to move to the crypto segment.
Additional segments included narrower areas: cryptocurrency arbitrage, mining, crypto exchanges and general investment communities. These sources were used to expand coverage and find non-obvious entry points.
- How do you find channels for "additional segments" like mining or arbitrage? Are there any secret sources?
Telegram has built-in recommendations: "Similar Channels" and "Similar Bots", and its own channel and bot parser.
Creative strategy vs. templates
The BrainStat team abandoned standard solutions and relied on non-standard visual solutions. The main focus was on working with images, with text playing a secondary role.
The breakthrough was memes based on popular culture. The creative with "The Matrix" used the iconic scene of choosing pills: the blue one meant "live from paycheck to paycheck", the red one - "Make money on crypto". The CTR of this creative reached 12%, which is four times higher than the niche average.
The creative with "Harry Potter" worked well - the scene in Gringotts bank with the gold and the caption "Me when I got into crypto".
They also tested contrasting images - an office worker at a computer versus a crypto investor with a laptop on the beach, emphasizing the freedom of time.
The neural network artwork created in Soro underwent mandatory post-processing to remove the "neural network effect" and make the images as natural as possible.
Emoji in creatives performed the worst and were excluded from campaigns.
Technical realization
The launch took place with minimal budgets - 0.7-1 euros per campaign. If there were clicks without registrations, an additional budget of 1-2 euros was given to test the hypothesis. The logic of decision-making was clear: if there were registrations, the campaign was scaled to similar channels; if there were no results after 300-400 impressions, the campaign was switched off.
The launch strategy envisioned two phases:
- first testing in "Channels" format with shows in selected channels
- duplication of successful campaigns in the "Users" format with targeting of subscribers of effective channels.
All analytics were based on detailed markup with channel tags, creatives, purchase format and placement date. Budget allocation: 80-85% to scale proven bundles, 15-20% to continuously test new channels and creatives.
- Show me an example of a real UTM tag you use? How deep do you drill down into analytics?
In the UTM tag we set the place of advertising, the username of the channel, the channel of attraction - TG Ads, the number of the offerer and the presence of a picture, emoji. In the future we analyze and optimize advertising on all parameters, making priority on the cost per lead and payback.
Campaign results
The dynamics of the cost per lead showed a steady improvement. The starting 11 euros per registration decreased to 9 euros after one month of operation (18% decrease), and reached 7 euros in the second month (36% decrease from the initial figure). The expected target is to bring the figure to 5 euros.
The project reached the break-even point in the second month of operation. Already in April, the first sales of the flagship product costing 180,000 roubles were received. The CTR of the creatives increased from the initial 3% to 10% thanks to the creative approach.
The conversion funnel stabilized at the following metrics: 9-10% CTR, 10% conversion from clicks to registrations, 10% conversion from registrations to tripwire, 20% conversion from tripwire to the main product.
Key Insights of the Project
The main conclusion is that in overheated competitive niches, creativity, not standards, wins. Memes, irony and recognizable cultural references catch the eye and stop scrolling better than any claims of "expertise". The audience is tired of direct advertising and is more willing to respond to emotions than to dry arguments.
The second point: the quality of traffic sources is more important than their volume. Especially in the crypto segment, where mass purchases on sites with questionable metrics quickly burn up the budget. Manual selection of channels with a live, engaged audience is a prerequisite for success.
The third lesson: a systematic approach and constant testing of new solutions yields results even in the most difficult niches. Active use of neural networks to generate creatives has accelerated production and made it possible to quickly find visual formats that "shoot".
- If you were starting this project over again, what would you do differently?
I think it's okay. We made the right decision - to run ads with a picture and a bigger budget. Thanks to this, we were able to achieve high clickability and make the cost per lead cheaper than it would have been in ordinary text ads
Current status and plans
After the April launch, the project continues to develop. There is an active rotation of new segments, creatives and constant testing. The basic bundles are gradually decreasing in effectiveness, which requires a constant search for new channels and visual solutions.
This BrainStat case shows that even in the most competitive niches you can achieve results with the right strategic approach, quality analytics and a willingness to constantly experiment.
BrainStat team regularly publishes more insights on Telegram advertising in its Telegram channel.
More useful articles
/w=1080,quality=90,fit=scale-down)
/w=1080,quality=90,fit=scale-down)
