Case

"Sure, but why?" – how AI and a unified rhetorical approach in advertising reduced product conversion costs by 73%

  • Brand
  • Sber
  • Advertised product
  • "Auto" is a section within the SberBank Online app where one can create a digital profile for a car and manage it in just a few clicks. Here one can: pay fines and toll roads, take out or renew insurance, schedule maintenance at verified dealerships and s
Sber is a category leader and offers car enthusiasts a comprehensive end‑to‑end ecosystem that addresses all their automotive‑related needs. Other category leaders – T‑Bank and Avito – are pursuing a similar path, building their own digital ecosystems. However, competition comes not only from them but also from traditional, "analogue" ways of buying, selling, and owning a car. The project’s target audience is car enthusiasts aged 25–50 (core: 30–45) from large and medium‑sized cities.

Problem

Spring and early summer mark the peak season for getting cars ready for the road. During this time, drivers take out compulsory auto insurance (OSAGO), schedule maintenance, and plan trips. Therefore, typical product‑focused messaging like "pay your fines in the app" tends to get lost in the noise. That's why we set a broader objective than simply listing the section's features. We needed to speak the audience's language, demonstrate the product's value through situations familiar to every driver – and develop effective creative and tools to make the most of this peak car‑preparation season.

Solution

The project was built around the recognizable verbal construct "Sure, but why?", which plays on familiar car‑ownership situations: insurance, maintenance, fines. Each scenario ended with a reference to the "Auto" section in SberBank Online, where the issue could be resolved more simply. Why make life harder when routine automotive tasks can be handled with less effort? On Avtoradio, a two‑week morning show featured a game called "Rhetorical Question" that engaged the audience with the section's services by debunking harmful car advice. Listeners called in and answered questions about popular tips by saying "Sure" (if they thought the tip was useful) or "But why?" (if they considered it a myth or harmful), and won prizes for correct answers. The campaign was reinforced with teasers, video spots, and integrations into the "Auto Market" segment. In digital, when creating display and performance inventory, we used AI: neural networks generated a series of posters and three short videos. The creatives played on specific car‑buying or ownership situations through grotesque exaggeration. The AI‑generated videos strengthened digital communication and gave the campaign an additional visual anchor. Moreover, before launch, the AI evaluated dozens of creative options and predicted their performance based on historical data. This allowed us to select concepts that were guaranteed to work and bypass the lengthy production cycle. The project was also supported across our own channels on Telegram, VK, and Max "For Those Who Drive," as well as in the brand media Rambler Auto. This helped retain the audience that came for expert content and guided them toward the product.

Results

All channels delivered a unified message and worked in synergy. Radio built initial interest, digital scaled reach and caught up with the audience, while our own media grounded the communication within the ecosystem audience, engaging users and reinforcing their brand loyalty. The campaign's total reach across all channels amounted to 45 million; the cost per visit to the "Auto" section dropped by 73% compared to the pre‑launch period. The promotion generated a sixfold increase in section views relative to the previous always‑on placement. Compared to the advertising traffic before the project launch, we achieved: • 4.3x more clicks to gas stations; • 3x more clicks to vehicle appraisal, trade‑in, and "Auto Deals of the Month" offers; • 2x more clicks to insurance policies; • 1.5x more clicks to new cars and used‑car listings. The actual cost per click on media advertising turned out to be 2x lower than forecast, matching the efficiency level of a conversion‑driven performance placement.