Wolt Delivers Hyperlocal Experiences With Geospatial Analytics

Whether you’re craving sushi or sending flowers to a special someone, Wolt delivers food, gifts and more to people across Europe and Asia. Having grown rapidly since its 2015 launch, the Finnish company joined forces with DoorDash in 2021 and is now present in 27 countries. 

Wolt’s local commerce platform is powered by a three-sided marketplace that connects shoppers, merchants and couriers. Geospatial data is essential to Wolt’s strategy of providing hyperlocal experiences for millions of consumers. Wolt relies on Snowflake’s Data Cloud to power geospatial analytics that support a high density of merchants and venues, intra-country expansion and new country launches.

Surrounding customers with high-quality merchants is key to Wolt's success — but that can be difficult to accomplish when operating in more than 500 cities across 27 countries. “The real challenge comes down to building infrastructure that supports all these countries at scale while also operating at the level of detail to hyperlocally solve for great selection,” says Sharat Ramamani, Analytics Lead at Wolt.

Arbitrarily defining markets without considering population and merchant density can lead to negative experiences for consumers. “We believe that whether you’re the GM of a country or a sales rep, you should understand our condition and positioning in the market — and make decisions from that,” Ramamani says. “Prior to our geospatial solutions, we were a bit more limited in this regard.”

Seeking a better way to calculate population coverage, Ramamani used an open source geographic information system (GIS) to determine Wolt’s “catchment” areas based on merchant delivery radii. Layering on Wolt’s proprietary data set, which is based on the H3 hexagonal hierarchical geospatial index, provided a precise representation of Wolt’s population coverage by country. “H3 is really valuable because it standardizes the world into hexagons of the same size,” Ramamani says. 

Ramamani’s proof of concept (POC) bolstered ad hoc analysis, but architecting a performant, repeatable geospatial analytics solution required a modern data environment. As a long-time Snowflake customer, Wolt turned to the Data Cloud.

Snowflake’s native support for geospatial features accelerated Wolt’s transition from POC to a scalable geospatial analytics solution. “We already have individual data sets for all Wolt venue radii and Wolt cities loaded in Snowflake,” Ramamani says. “So, we use Snowflake’s existing geospatial capabilities to quickly determine coverage areas at scale for all regions within our 27 countries.” 

Wolt relies on Snowflake’s SQL functions for H3 to efficiently fill in coverage areas with H3 cells and, in turn, assign population data to specific markets. According to Ramamani, “Our population data contains over 2 million rows for the whole world. Now it’s just a simple join of our coverage area data set — with every hexagon having an H3 cell ID — and the population data set.”

Market managers used Wolt’s geospatial solutions to understand merchant density and coverage gaps. Intra-country expansion managers now have tooling to objectively compare potential markets and can launch cities weeks faster than before. 

International expansion teams can use geospatial analytics to inform decisions that impact Wolt’s global presence. According to Ramamani, “We launched Austria last year, and we could use our hex-level population data to identify hubs of population density. That helps us make decisions because density is often an opportunity.”

Consumers might be the biggest beneficiaries of Wolt’s geospatial analytics solution. Hyperlocal insights provide Wolt’s teams with increased market awareness that can improve the customer experience. For example, Wolt’s sales leaders use geospatial analytics to identify pockets of healthy and unhealthy coverage. Allocating sales resources to underserved regions — down to a 200-meter area — helps Wolt increase the selection available to consumers in their immediate vicinity.