Scentbird fragrance subscription: context, role, process

Fragrance recommender

How do I pick a fragrance online?

Selling fragrance online is like selling music on compact disks.

Imagine you can’t download it.

There are genres, notes, bands/brands, but you can’t say if you like it before you experience it. Maybe several times. But you know that your friend recommended you a lot of great stuff before because they tried it all. Not because of the notes or octaves or instrumentals or vocals, right? And they recommended different genres, too.

That’s where the recommender comes into play: a friend.

Users range from newbies to connoisseurs and most of them have to figure out how to pick a fragrance online. They rely on an algorithm to narrow down the options, but they want to make the final choice themselves.

User problem

Business goals

Increase LTV by providing better recommendations.

Scope of work + outcomes

Power-user interviews

Learn how they pick the next product online; 20 interviews.

Product card redesign

Add to queue metric performed ~50% better.

Product ratings

We needed more product ratings to train new model; the Orders history section redesign helped us to grow the number of ratings three times in the first month.

Data pipelines

We've built data pipelines to track each recommender's performance from the moment the product was displayed to the moment it was received and rated.

The new recommender algorithm

The new algorithm was implemented and tested against the previous during the A/B test.

Pick the right algorithm

Product properties

The previous recommender was based only on product properties (TF-IDF). I.e. if you like banana, it will recommend you more banana-based items.

Collaborative filtering

My goal was to provide recommendations that were based or real-life experience from other users to give people more variety of recommendations. I.e. if you like Tesla EV, you are likely to be interested in SpaceX as well, even though their properties are different. Same with fragrance.

Natural language processing

Another option was to try Natural Language Processing methods to extract topics from user-generated product reviews. I've built a prototype. After analyzing 400,000+ comments I found out that most of them were vague ("LOOOOOOVE IT!!!"-kind of vague) and the remainder was insufficient to extract representative (useful) information. At that moment we decided to postpone this path, although scraping other sources was really an option.

The ultimate idea was to combine all of them, but it was too much for an MVP.

Scentbird: reusable modules

New algorithm and UI were presented as a new section with inner navigation layer

We added the Recommendations section to the main navigation to make it more discoverable. Section accumulated pre-existing Quiz (which we modified a bit to reflect the new algorithm) and Orders history, where we displayed only fragrances.

Scentbird: recommender UI components

Core UI components

It looks simple at the surface, while most of the heavy lifting happened during the user research and working with data.

Challenges

Hypothesis: recommendations quality affect user retention

Core design stages

Data pipelines to track recommenders efficiency

Data pipelines: tracking stages

Context

There are a lot of recommenders sprinkled all over UI. Depending on the context, they were mostly smart database queries based on similar notes, brand names, collections, search suggestions, the fragrance of the month, etc.

There was a hypothesis that some of them caused more harm than good.

To measure the efficiency of each recommender and its influence on retention, we decided to build data pipelines, gather the data, and analyze it to make informed decisions in the future.

Product tracking stages

This data collection implies long waiting time since fragrances are shipped one item monthly (maximum three items per month on a highest plan).

Get 3× more ratings to improve the new model

We added a simple rating mechanism to the Orders history section. The number of ratings grew three times instantly, thanks to the existing user base.

Orders history: rating

First, we redesigned information cards so that they reflected different states of the order. And we gently introduced the rating mechanism. The idea was that the existing users will be able to quickly rate previously received products as they check their next delivery status.

Orders history: negative rating

Filtering negative reviews

E.g., shipping issues can cause negative reviews. Reasons other than the product itself were kept out of the model (as irrelevant).

Product card redesign

I learned from power users how they pick fragrances: they read a lot, digging into notes and reviews, searching for more information on other sites. To ease their process and to show more options to newbies, we altered a product card.

(Small product cards remained for backward compatibility.)

Original

Scentbird: original product cards

Altered

Scentbird: altered/redesigned product cards

Personalization and future algorithm improvements

We added like/dislike options to the recommendations feed to gather people's reactions as an initial measurement for any of our recommenders. These options are also a great tool to personalize future recommendations.

Dislike options were implemented first to learn what confuses users. Sometimes dislikes don't mean that the recommendation is wrong; it's just that the user has already tried the fragrance and doesn't want to see it in the feed again.

Product card

Dislike reasons

Dislike options were meant to eliminate allergic notes and to follow-up on fragrances people already tried (two different ways to improve recommendations and distinct fragrance dislikes from feed adjustments).

The Quiz

Previous Quiz was based on product properties.

This time we mapped positive ratings to quiz answers, and from this moment, we could also map quiz answers to other users' experiences. I.e., new users could take a quiz and get recommendations based on similar responses and positively rated products by other users.

Updated Quiz look

Updated Quiz look

Altered Quiz look

Concept: filter/quiz

The intention was to display this version on the homepage instead of the original Quiz (since it combines both elements). This section didn't make it into production, but it is worth mentioning since it looks more aligned with the new style. It was created in collaboration with Karine Arutyunyan.

Mobile layouts

Recommendations section mobile layouts and conditions

Recommendations section mobile layouts and conditions

Fragrance history section: mobile layouts and conditions

Fragrance history section: mobile layouts and conditions

Takeaways

The new algorithm will continue to use data to provide different recommendations — ones that real users found useful. Collaborative filtering, being automated and feature-independent, scales better for recommender services.

More about Scentbird: subscription context, retention strategy

Scentbird is a New-York-based fragrance subscription with over 300,000 active subscribers (as of January 2020). I focused on a digital retention strategy.

2018-20

Scentbird: Mobile navigation redesign

Information architecture improvements: 20% faster task completion rate

2019

We removed 50% of UI elements from mobile navigation without dropping any metrics and made it easier to understand for 80% of users.

Mobile navigation redesign
Scentbird:

Homepage redesign + onboarding: focus on a fragrance

2019

The new personalized modular system helped registered users to get the most of their subscription while increasing LTV.

Compare before and after