One of Germany's leading retailers, dm-drogerie markt (usually abbreviated to simply dm) has provided a range of cosmetics, healthcare items, household products, and health food and drinks for nearly half a decade. 

Founded in 1973 by Götz Werner as a drugstore in Karlsruhe, Germany, the retailer has grown to 3,700 stores in more than a dozen European countries, including Austria, Hungary, the Czech Republic, Croatia, Slovakia, Romania, Serbia, Slovenia, Bulgaria, Bosnia-Herzegovina, Italy, North Macedonia and Poland. It’s recognized for its flat hierarchical structure, employing more than 60,000 people, and its focus on employee well-being and social commitment over pure financial returns

The company’s subsidiary dmTECH is responsible for providing IT solutions, including the e-commerce and search capabilities of dm stores and retail website. With more than 900 employees, dmTECH aims to adopt solutions that support dm’s customers and employees in the best possible way, taking advantage of data-driven approaches and advanced technologies in web and app development in an agile environment.

Moving from a monolithic platform to more flexible search

Traditionally a brick-and-mortar retailer, dm’s online business is still relatively new, but omnichannel retailing has become a hot topic for the company, and even more important during the recent global pandemic. 

In 2021, dmTECH saw an opportunity to improve the search experience for its customers and internal teams but was challenged by the limitations of its existing platform. With that in mind, it sought to launch a new search project. 

The company was using the search capabilities built into its large e-commerce platform. But the greater movement towards omnichannel retail and a goal of ultimately adopting an API-first approach to search — to avoid architecture issues, provide more flexibility and reduce internal dependencies — meant its requirements for search had evolved.

"From a technical perspective, our existing search solution was part of a monolithic platform with limited extensibility. We were planning to further improve search and defined a four-step search pyramid — from technical base configuration to guided search to self-learning search, and ultimately to personalized search. We knew this would be hard to do with existing systems.”

In addition, synonyms for search terms were primarily maintained by business users while, with their existing system, requests were handled and checked by developers, creating bottlenecks. 

dmTECH wanted to enable marketing and business users to work with and optimize search themselves. With the previous system, options in the WYSIWYG-editor were limited and it lacked debugging. A more flexible self-service solution was needed. 

The company turned to an external search consulting partner for evaluation and decision-making support when looking for an ideal solution to meet its new needs, and Algolia was placed on the shortlist for testing along with 4 other vendors.

Improving Search and Suggestions with Algolia

The retailer tested the business value of adopting Algolia against its existing solution via A/B testing for two months, and selected the solution based on its measurable impact on business metrics, API-first approach, flexibility, transparency, and the ability for both business users and developers to act autonomously. 

The company went live with Algolia on its German web store in July 2021. This online launch was soon followed by launching Algolia on its mobile app, then its Austrian store, and across other countries by the end of the year. 

By 2022, Algolia Search and Suggest functionalities were implemented across all mobile apps and all its online shops, for product search, category pages, and some query-based content such as sliders. 

Using Algolia, they provide merchandising rules, business ranking, and the configuration of facets displayed in the online shop. 

The Algolia team provided consulting and support during and after the implementation. “During implementation and pre-sales, the Algolia team was highly involved and provided fast and competent answers to our questions and requirements,” says Florian Plag, IT Consultant, dmTECH GmbH. 

Customers are provided with suggestions dynamically calculated based on customer search, with predefined filters including dropdowns, price and, for some products, color. Search-based content, like category pages, brand pages or product sliders and grids, can be created and edited using dm’s CMS (Content Management Systems), with the slider being automatically generated when an editor enters a query.

In the short months since deploying Algolia, dm has already seen positive results: a 2% improvement in conversion rates during its first tests of Algolia Search, and a 1.17% improvement in click rates after introducing Algolia Suggests. The company has also recognized a noticeable improvement in speed and performance. 

Algolia has proven to be a significant boost to both dm’s technology and business teams, notes Denise Schäfer, Technical Lead Development at dmTECH‘s search team. Its API availability and the availability of API clients in different programming languages provided dm with the extensibility and flexibility that dm was seeking. 

“Developers found Algolia’s tutorials to be a great help, since it can be used as a blueprint, for example for sorting based on ranking or automatic boosting and generic rules,” Florian Plag says, adding that Algolia automatically recognizes relevant brands or categories in search terms and boosts them in results accordingly.

“Business users value the possibility with Algolia to easily follow and understand how the ranking is generated and how it is impacted by their changes, including the relevancy, comprehensive sorting, accessibility of custom rankings, and many options they have to create and configure rules.” - Florian Plag, IT Consultant, dmTECH GmbH

Bottlenecks begone: Testing, rules, debugging and more

With the onboarding of each new country, implementing new search capabilities became easier and the benefits to business users and developers became faster to achieve, reducing past bottlenecks. 

Today, because of the ease of working with Algolia, more employees are involved in Search, and have an opportunity to take advantage of the solution, says Heiner Eisenmann, e-Business Development, dm-drogerie markt. 

“We’ve tied Algolia to conduct A/B-Testing on various custom rankings,” he says. “We still do, and due to the configuration possibilities, there’s a lot of opportunity to fine-tune the base rankings according to our business rules.”

“We’ve tied Algolia to conduct A/B-Testing on various custom rankings,” he says. “We still do, and due to the configuration possibilities, there’s a lot of opportunity to fine-tune the base rankings according to our business rules.” Heiner Eisenmann, e-Business Development, dm-drogerie markt 

And dm has a lot of business rules implemented on both the web and app-based stores; for example, providing alternative products if a customer searches for a particular brand that’s not carried or out of stock, or products based on article numbers, or categories of products being replaced with a more general term in search queries. And the list goes on. (For instance, if a customer is looking for a specific brand of cough syrup, a predefined rule turns their query into a category search for cough syrup if the product is unavailable.) 

The Algolia dashboard is a team favorite tool, according to Eisenmann, allowing business users to define and maintain rules without developer intervention, while Algolia Analytics provides a fast analysis of user behavior for optimization. 

Using Algolia, dm’s team can easily debug the functionality of each index to ensure relevancy for the business users creating and maintaining them, allowing non-technical users to systematically check if and how the rules they put in place impact the customer experience. Internationally, Algolia has aided with business departments’ work in pinning products and creating hero categories to promote specific items.

The company’s tech team, dmTECH, is continuously working on improving both ranking and customer-relevant features, as well as rolling out features internationally. 

The team still has regular meetings with its Algolia customer success manager to discuss upcoming opportunities and questions, and specific topics, like deep dives on rankings or personalization. It is currently exploring the adoption of Algolia’s AI-based reranking and personalization and undergoing testing.