Driving Online Growth and Customer Satisfaction with Data-Powered Machine Learning

By pinpointing the features that enable consumer purchases, we’re rolling out growth bundles for Storefront Pro — our premium white label e-commerce platform — and driving growth for retailers.

Strengthening customer connections

When retailers partner with Instacart, they’re looking for features that can deepen their connection with their customers and propel their brands forward. In the past, we’ve helped retailers accomplish this by identifying our best-performing features across the Instacart Marketplace and compiling those features in our white label e-commerce platform, Storefront Pro. With Storefront Pro, retailers can access exclusive performance metrics and maintain control over their brand’s appearance on their storefronts.

Our efforts in combing the Instacart Marketplace for the best features are ongoing. But, in the spirit of innovation, we also decided to reexamine how we deliver value and strengthen retailer-customer connections through Storefront Pro specifically.

Experiments in innovation for Storefront Pro

By narrowing our focus to Storefront Pro exclusively, we were able to devise a plan for scientifically evaluating feature performance and elevating the platform’s capabilities. With this approach, the data we found allowed us to maximize actionable outcomes — and roll out features that were proven to succeed. 

We developed a new machine-learning (ML) model that could:

• Identify the features driving growth on Storefront Pro
• Provide key insights to how we build growth bundles
• Inform rollout methods for efficient performance monitoring 

Let’s explore how we built this model and our development approach – and see how the results can help grow your brand.

Building the ML model

We started our Storefront Pro experimental ML build in 2022. Our primary goal was focusing on the customer funnel for high-performing customer segments and what features enabled those customers to make purchases. For example, would a customer be more likely to add an item to their cart if they found it while browsing, or if they found it during a specific search? 

After some experimentation, we uncovered a new strategy. Instead of focusing on adding particular items to one order, we looked for the features in their current order that led to a customer placing their next order. With this adjustment, our ML model identified features that customers were actively using to drive their future purchasing decisions.

We capped our study of repeat orders at 28 days to keep the data sample relevant. 

Equipped with this new strategy, we moved forward with building the ML model — and with the model complete, we could now identify the most promising features and move on to creating our growth bundles.

Creating measurable growth bundles

Our functioning ML model demonstrated not only which features influenced the customer life cycle, but how those features were influential. For example, the model specified when features performed well for first-time customers, and when those same features didn’t resonate as strongly with repeat purchasers.

By tapping into these insights, we created growth bundles that aligned with the entire customer lifecycle. For extra context on how we’d bundle these features for customer use, and to ensure our approach was aligned with the unique needs of retailers on Storefront Pro, we brought in our customer success team and retail partners for their feedback and suggestions.

We quality assured these growth bundles for several months, partnered with retailers for our trial run, and packaged features into bundles as suggested by the ML model. With this method, we were able to more rigorously test our hypotheses of what would be successful and have confidence that our findings could accurately inform future growth bundle rollouts.

Achieving results with repeatable methods

With our bundles, we hoped to see measurable success in several areas that enable sustainable retailer growth, including steady traffic, existing customer retention, and new customer activation.

Let’s take a closer look at what some of our growth bundles are accomplishing:

• Assisting customers in making informed purchasing decisions based on their previous purchases
• Streamlining the customer’s experience by removing non-converting pages
• Promoting Instacart+ to maximize convenience and value
• Increasing personalization — and conversion — with relevant product recommendations in an unobtrusive way

The results from our experiments were positive — and promising. With our new approach, we can give Storefront Pro features that are backed by measurable, repeatable data. Want to know the specs? Reach out — we’ll be happy to share.

At Instacart, we’re always pursuing growth by experimenting with the ways we deliver value, especially to Storefront Pro. With our new growth bundle strategy informed by machine learning, we now work with retailers to roll out features they can depend on to maximize growth — and customer satisfaction. 

To get the best Instacart has to offer with Storefront Pro and our team of data scientists, contact your Business Development Representative or Customer Success Manager. Not sure who to contact? Send us an inquiry to get started!

Instacart

Author

Instacart is the leading grocery technology company in North America, partnering with more than 1,400 national, regional, and local retail banners to deliver from more than 80,000 stores across more than 14,000 cities in North America. To read more Instacart posts, you can browse the company blog or search by keyword using the search bar at the top of the page.