Amplitude had the privilege of sitting down with ManoMano’s Head of Product, Simon Berthet-Bondet, to discuss how his team uses Amplitude to drive successful cross-sell for their business, resulting in a 15% increase in sales.
Thank you so much for being with us today Simon! Can you tell us a little bit about yourself and ManoMano?
I am the Head of Product at ManoMano, responsible for building a customer experience that helps users understand DIY and gardening. ManoMano is the largest marketplace for products and services in DIY, gardening, and home improvement in Europe. We have a team of about 20 product managers, approximately 100 developers, and a team of data scientists.
We’ve seen tremendous growth thanks to our marketplace model—it’s really what sets us apart in the industry. We work with 5,000 sellers, offering over 16 million products in our catalog, available to our 7 million active customers across Europe.
In parallel to our consumer platform, we have also launched a platform dedicated to professionals and independent workers. In France, for instance, nearly 1 in 4 professionals has created an account on our ManoMano Professional platform.
Why is cross-selling so important to your business?
In ecommerce, profitability often hinges on the ability to recommend and sell complementary products: if we can increase the number of products they buy, we can increase the profitability of our business.
Cross-selling also allows us to become a provider for your entire DIY project, not just a one-off product sale. The reality is, DIY projects and gardening such as painting a wall or rebuilding your terrace, often involve many different tools and equipment. Our goal is to ensure customers have everything they need for their entire project. This makes cross-selling key.
How is your team set up to drive cross-selling?
Our focus on cross-selling started about 18 months ago, and from the beginning, we made sure that it was part of our OKRs so that the entire business was aligned.
We have dedicated feature teams: one product manager, four to five developers, and two data scientists for our cross-selling project.
What KPIs did you decide to focus on?
Our primary KPI is the number of items per basket. If we can increase the number of items per basket, we can directly impact revenue. We also look at secondary KPIs, such as the type of items added to the basket and whether customers are reaching their free delivery threshold. All of our KPIs are analyzed by Amplitude, which automatically enables everyone on our team to be aligned.
We break down our KPIs into different segments: we look at Countries and B2C vs B2B Businesses, for example. We will then go into a much more detailed segmentation to look at types of audiences such as electricians vs plumbers vs carpenters:
With Amplitude, we can see whether someone is looking for a range of different tools for their project, instead of looking just for one tool. We can actually see the journey that someone will take, the different URLs that they are visiting, and start to analyze their behavior across both the ManoMano website and App:
With Amplitude, we’ve discovered areas that can truly impact our business and profitability. Specifically for B2B customers, we saw that this audience often purchases the same product in large quantities. Thanks to these insights, we’ve implemented a pricing strategy that caters to this specific behavior.
With such a large catalog, how do you understand the affinity across different products?
We started with manual recommendations by collaborating with our category managers to identify products that could be sold together. However, given the sheer volume of products, this approach clearly had limitations. We engaged our data scientists to develop machine learning algorithms that analyze past customer purchases to recommend products for future customers. We even delved into deep learning algorithms to predict compatibility based on product information and user data in real time.
Providing relevant products is great but in this environment, especially digital purchases for DIY and gardening, you often need a lot of advice and guidance on the type of tools needed. We’ve started automatically creating content using Generative AI. Below is an example of “Wireless Gardening” references, that include tips and tricks alongside the actual product recommendations:
How has A/B testing impacted your cross-selling strategy?
A/B testing is critical for us, as it helps us assess the incremental value of our developments. When we launch new features in our product there are build costs, run costs, and a communicative cost that we impose on our users. It’s only acceptable to bring these features to market if it brings value to these users. So every time we develop something new, we try to measure it through an A/B test.
We have run 150 to 200 in-house A/B tests, and we don’t need to rely on data analysts to get these tests up and running. Everything is available for us within Amplitude directly—it provides us with templates for consistency and dashboards for transparency. Today, data is in the hands of the product managers.
In terms of results, we drove a 15% increase in incremental sales measured through 17 winning A/B tests.
What is your advice for someone looking to build a cross-sell strategy?
For me, there are 3 fundamentals.
- Make sure your catalog data is clean. Catalog data is one of the most important things if you are in a marketplace model. We don’t have full control over the data being added to our platform and therefore we have a lot of work to do to ensure the data is consistent. Having a clean set of catalog data, with the right taxonomy in place, will allow for a successful cross-sell strategy.
- Make sure you can track user data and are continuously driving insights from it. Cross-selling can only work if we can correctly track user behaviors. This is where Amplitude provides the richest set of data for us to work with and provides us with the deepest level of insights. Thanks to Amplitude, we can understand whether our recommendations are properly displayed, clicked on, and sold.
- Strike the right balance between precision and coverage. With such a large amount of references within the catalog, it is impossible to be 100% precise every single time. It is important to optimize your algorithms to find the right trade-off between coverage and precision when displaying your product recommendations.
What's next for ManoMano?
Our next steps involve guiding customers through their entire DIY or gardening project, not just recommending one-off products. In our domain, this is the real challenge—if you have a sink to change, do you understand the steps you need to take and the tools you need to use? When shopping online, you can’t just ask for help from an expert as you would in a store. So beyond just product recommendations, we want to be able to provide customers with guidance and education, through a fluid and seamless online customer experience. We’re testing conversational experiences through our website and application, as well as Generative AI to allow us to create some of this educational content for our customers.