ipsy Referral Program


Mid 2016, we launched a new referral program to a sub-set of subscribers where they were given a limited number of invitations to send to their friends, where their friends and family can receive a one month subscription to ipsy for free. This referral program is still in test mode as we continue to refine the feature.


The goal of the new referral program is to grow the subscriber base, with the expected outcome being a steady organic growth rate, where about 1 in 2 friends or family members referred would sign up and stay with us beyond their first free month. To validate the idea, I sent a survey to to our existing subscribers to gauge interest, and over 75% expressed at least moderate interest, while 53% indicated that they would be very interested to send a free Glam Bag.

I started by mapping a simplified user flow outlining the major use cases. In the beginning, we mostly followed Blue Apron’s model, where active subscribers need to enter the name and email of their friends or family, and we will send out an invitation to the friend on the subscribers behalf. Through the invitation, the friends could then claim their free Glam Bag by signing up and personalizing their first bag.

In addition to the flow diagramming, I also explored initial ideas by sketching and whiteboarding, then narrowing down and wireframed the most clear and easy to use ones. Early design reviews here helped gather initial feedback from stakeholders.

In terms of flow, discoverability is very important for the feature as we are encouraging the subscriber to send a free Glam Bag to their friends. There are 4 entry points into the feature, making sure that every subscriber has the chance to send invites. On the feature’s main screen, we also want to highlight scarcity by listing the number of invitations left. This not only rewards the existing users by giving them this feature as a perk for being with us, but also reduces the potential for fraud. The main goal for the interface is simplicity and ease of use. The input forms are short and simple with a noticeable call to action, while the sent list provides the user feedback for sending their invitations.

Visually, we know that product and photography has been performing very well on social channels. The campaign email and main hero image on the referral side apply that learning. I also followed the standards of our UI kit so that the feature reuses our existing components and looks consistent with the rest of the application. This also sped up development timeline. More information can be retrieved through a link on the main screen to reduce the amount of text, further bringing focus to the actionable items on the screen.

On the redemption side, a reminder was put in to guide the invited friend along the subscribe flow, which is the same as our existing subscribe flow. The profile photo of their invited friend is included to add a familiar touch. On the subscribe page itself, I made certain that the pricing reflects the discount of the Free Bag, and highlighted the fact that ipsy is a subscription service to set expectations.

The feature was released mid-2016 and the outcome after initial launch was not as successful as expected. During phase 2 of the project, we experimented with changing elements of the flow and pages to enhance the feature. The first thing we tried to improve on was discoverability. We noticed that not a lot of current subscribers are engaging with the feature and sending out invites. Therefore, we tried a more aggressive way of introducing the feature via a modal and saw a significant increase in engagement. Additionally, we also tested out a few different hero images to make the feature more exciting. Visually, consistency was important to us so we made sure that all assets are updated across the feature, including emails.

We noticed that a lot of invitations remain unopened. To tackle this problem, we designed two additional enhancements. The first is to replenish the number of invitations a subscriber would receive. Previously given only three, now the subscribers who have the privilege of sending invitations can enjoy three additional invitations once all three are sent. This ensures that the subscriber can send out invitations to a variety of their friends, some of whom may be more interested in our service than others. The second enhancement build is the ability for the subscriber to resend an invite to their friends and after just in case if the friend missed the invitation the first time around.


Resending is a simple 1-tap process, and the state of the invitation tile changes depending on interaction. We limited resending to be available only once, that way the friend or family member of the referrer would not be spammed. Subscribers can only resend the invitation if it has not been claimed after the first 72 hours, giving their friend and family plenty of time to join.

On the third iteration of the feature, after getting qualitative feedback from subscribers, we decided to implement an easy way for people to invite their friends and family via contacts from their emails. Since not a lot of people know their friends’ emails by heart, the contact importer allows them to quickly select and send invitations to their friends if they log into their email account.


We wanted to personalize the experience for our subscribers and not overwhelm or distract them with choice. Therefore, if a subscriber registered with us via Gmail, we would only offer them the contact importer allowing them to import their Gmail contacts, assuming that the subscriber is using their primary email with their ipsy account. The same goes for Hotmail and Yahoo. These three email providers covered the majority of our user base. If the subscriber created their ipsy account with an email address other than Gmail, Hotmail, and Yahoo, or if we cannot recognize which provider they use, they will be given a choice between the three.

As we continue to monitor the data, we will also be iterating to ensure that the expectations of the friends who recently subscribed via the free Glam Bags program is met. Currently we are seeing a fairly high free to paid conversion and a comparable 3-month retention.


Learnings from designing this product:

  •  Build an MVP, release it to a small cohort first and don’t stop iterating!
  •  Design small incremental experiments to detect changes in the KPI.
  •  Be comprehensive, look for areas of opportunity in both the referral and referee flows.
  •  Sometimes, people can be dishonest, send free Glam Bags to themselves, design solutions to prevent this from happening without it being a bad experience.
  •  Keep in mind the entire lifecycle of communication with the customer, including promotional and transactional emails that compliment the experience of the application.

Product: Simon Tisminezky, Nipun Sachdev, Jen Faenza

Engineering: Gulsah Kandemir, Nadya Voitik, Alexa Dorsey, et.al

Analytics: Jeremiah Gaw

Marketing: Jessi Calloway, Zoya Popoff, Ann Parden

Support: Katy DiNatale

<< Previous




Product Design, Visual Design

Next >>

©  Sophie Ying Su, 2017