Yandex Lavka is a grocery delivery service, part of the Yandex Go ecosystem, which offers a variety of on-demand services such as ride-hailing, food delivery, and logistics solutions. Launched in 2019, Yandex Lavka operates as a hyperlocal delivery service that provides groceries and everyday goods within 15–30 minutes. With its rapid expansion, Yandex Lavka has grown to become a key player in the online grocery market in Russia and other countries.

Yandex Lavka is integrated into Yandex Go, the umbrella platform for various services, providing users with a seamless experience across transportation, food delivery, and more. As Lavka’s catalog grew significantly — from 2,000 SKUs in 2020 to 10,000 SKUs in 2024 — one of the key challenges was to motivate users to explore and purchase from new categories, as they tended to stick to familiar products.

Despite the rapid expansion of Yandex Lavka’s product catalog, user behavior showed a tendency to remain the same — most users repeatedly purchased the same items, rarely venturing into new product categories. Even though Lavka introduced a wider range of products, the diversity of user purchases didn’t grow proportionally.

The primary problem we aimed to solve: Users tend to purchase the same products and are hesitant to explore new categories, even as the catalog grows. As a result, the platform faced a missed opportunity to drive higher revenue by encouraging more varied purchases.

Our goal was to design a solution that would increase the diversity of purchases by encouraging users to try products from categories they hadn’t previously explored. This would ultimately lead to an increase in the average order value and a boost in overall revenue for Yandex Lavka.

To find the best way to encourage users to explore new categories, we tested several product ideas through user interviews and surveys, and later through a pilot test. The goal was to understand what would drive users to step out of their habitual buying patterns and try new products.

We explored several product ideas, including:

• Personalized discounts on categories where the user hadn’t made any purchases yet.

• A “New Items Box” program where users receive a curated box of new products.

• Free samples with every purchase to entice users to try new products.

• Interactive recipes featuring new products to encourage purchases through inspiration.

• A "Buy Together with This Item" recommendation system for complementary products.

We conducted in-depth interviews with 10 regular users of Yandex Lavka, focusing on their shopping habits, attitudes toward new products, and preferences for the proposed solutions.

Key insights from the interviews:

1. Users generally stick to the same products, driven by habit and convenience, and are often reluctant to experiment with new items or categories.

2. Discounts were seen as the most effective motivator to try something new, as they lower the perceived risk of unfamiliar products.

3. The idea of allowing users to choose categories for discounts was particularly well-received. Users appreciated having control over what products they could explore with added benefits.

4. The New Items Box sparked some interest, but users expressed concerns that they might receive items they didn’t need or want.

5. Samples and recipes were less popular, as users didn’t find them particularly relevant to their shopping habits.

Following the interviews, we conducted an online survey with 100+ Yandex Lavka users to validate the findings from the interviews and gather broader user insights. The survey asked users how often they buy new products, what would motivate them to explore new categories, and which proposed solutions were the most appealing.

Key results from the survey:

1. 60% of respondents selected personalized discounts on categories as the most appealing solution.

2. The New Items Box ranked second in popularity, but users were wary of receiving products they might not like or need.

3. Free samples and recipes were less popular, indicating that these solutions didn’t resonate as strongly with users’ shopping behaviors.

4. Users highlighted the value of control—the ability to select which categories to receive discounts for made the offer more attractive and lowered the barrier to exploring new products.

Based on the results of the interviews and surveys, we identified personalized discounts on new categories as the most promising solution. We would proceeded with a pilot test to validate the effectiveness of this idea in a real-world setting.

Pilot description:

Users were presented with an option to choose 2 categories from a list of new categories they hadn’t purchased from before. They would receive a 5% discount on products from those categories for the rest of the month.

Key metrics for evaluation:

1. The percentage of users who took advantage of the personalized discount offer.

2. The increase in average order value for users who chose new categories.

3. The growth in purchases from the new categories.

After conducting user interviews, surveys, and usability test, we concluded that the personalized discounts on new categories idea was the most effective solution. Users responded positively to having the ability to choose their discount categories, which gave them control and an incentive to try new products.

This solution proved to be the easiest to implement and yielded the best possible results compared to other ideas like the New Items Box or free samples.

The next step would be to scale the personalized discount feature across the entire Yandex Lavka user base. To do this, we planed to:

• Optimize the user experience for selecting categories, making the process seamless and intuitive.

• Enhance notification systems to remind users of their discounts and encourage them to take advantage.

• Monitor key metrics such as average order value and purchases from new categories, continually refining the discount system to maximize profitability.

• 25% projected increase in purchases from new categories.

• 12% projected increase in average order value for users who used personalized discounts.

• High engagement: 70% of users would selected new categories and used the discounts.