UX & Analysis process for Clue Insights Feature
INSIGHTS CARDS for CLUE
BACKGROUND:
Clue is one of the leading period, fertility and health tracking app. It offers people the possibility to monitor their period, receive reminders about period, PMS, ovulation and fertility and track their symptoms to discover the unique patterns in their menstrual cycles. 
Clue is based on science and empowers users to take control and understand their health through reliable information, analysis and a unique experience.

Clue is rated as the top free period tracker app by the journal Obstetrics & Gynecology, a publication of the American College of Obstetricians and Gynecologists (ACOG). 
GOAL
When analysing users usage and data, we discovered a majority of people use the app and track mostly around their period and less during the rest of their cycles.

The goal of this project was to figure out why people track mostly around the period and how could we prompt them to track during the rest of their cycles. In other words: what were the needs and expectations to meet to increase their satisfaction and regular usage.

Envisioned as a solution we could easily implement and iterate further, the project was required to focus on a simple mvp and be executed and delivered in a limited amount of time.   
DESCRIPTION
The main questions were the following:
- Why do people track mostly around the period and not so much during the rest of their cycles.
- What are the benefits of tracking, why do people track and what do they expect to achieve (expectations) and get (reward) by tracking cycle after cycle around their period.
- What can we offer to users to make them track during the rest of their cycles.
- What are the needs on their side and the solutions to meet on our side.
- How to measure the performance of the project and iterate it further.
FIRST INTERVIEWS
Together with Adeline Lee, the project manager, I started with interviews of two distinct groups of people. One group was composed of 5 Clue users, invited at the office and the other group consisted of 5 women I spontaneously asked for an interview at BetaHaus open space.

I asked both groups why and how do they monitor their cycles and why is it useful for them to keep track of their cycles events. I asked them to explain further the benefits they find in monitoring, what they learned and how they used the data.
Then, I asked about the symptoms and cycle events they focus on and why and when they track. Eventually I asked what other things would be useful for them to learn or understand about their cycles, focusing on period and non-period related information.

The reasons why I asked two groups were the following: 
- Clue users would give me answers related to their personal usage of the app and goals, based on their choice to stick to Clue app and the benefits the current version offers, as well as the specific needs they have that the app is not meeting yet.
- Random people would focus on their personal way to monitor their cycles, without knowing the current solutions Clue app offers. They would focus on basic and universal needs.

From the first series of interviews we identified different points:
- People focus mostly on their period and track symptoms around it to understand better what is happening with their bodies at this time of their cycle.
- People have less knowledge about the other events happening during their cycles, such as the fertility window or the PMS. Clue group seemed to have more knowledge than the other group. Both groups were eager to learn more about their bodies during the rest of their cycles but didn't know where to start.
- Different people have different routines: some would focus solely on body related symptoms such as pain, skin and hair quality or emotions, other would add lifestyle events to their tracking routine such as exercise, food cravings or sex. However, everyone we interviewed seemed to have the same need to know if their cycles and symptoms were normal and how to interpret them.
- People were seeking reliable information to understand better their symptoms. Clue people especially felt the need to receive such information during the rest of their cycles, in order to get some reward for tracking.
- All people said they would track more often in this case.
HYPOTHESIS & FURTHER TESTING
After analysing the insights gathered during the first interviews I came up with different directions to test.
I focused only on information. My goal was to identify what kind of information would be useful for people to receive while tracking out of their period time, and what would prompt them to track continuously during their cycles.
I also needed to understand when and how people would like to get the information.
I asked people to identify their 3 favourite cards (see image above) and tell me why they found them useful and how would they use this type of information.
In parallel, Adeline launched a survey to the Clue Beta base asking the same questions. 

People preferred the following cards: 
- Insights about themselves, like their average period or cycle length and their data compared to global averages. They found this type of information useful and reassuring.
- Information directly connected to the category they just tracked and how it connects with their cycles. People enjoyed the perspective to receive something in direct context with the action they achieved (tracking a specific category) and saw it as a direct reward for their efforts and a useful piece of information to predict and understand their health patterns. They also enjoyed the idea to see what could be good for them to track together with what they just tracked.
- They were excited by "Good to know" tips in direct correlation with the cycle phases (period, follicular phase, fertile window, luteal phase, PMS). They pointed out the need here of actionable pieces of information.
- General and scientifically approved information in correlation with their cycle phases. People were eager to learn more from a reliable source and understand better their bodies, but they preferred short texts with sources than long texts

Other cards that were not performing well:
- "Quote of the day" from other users was judged fun but not really useful.
- Gamified approach seemed too light for a serious subject such as health.
- Comparison with other Clue users (xxx users also tracked this category today) was judged too broad and not personalised enough. However people mentioned that a comparison between their data and other Clue users data, if in correlation with the phase, would be useful.

In general, people were excited by the perspective to receive personalised information about the category they tracked and comparison to global averages (answering the "Am I normal?" question), something in context with a possible action to take and general but reliable pieces of information. They all agreed such insights would prompt them to track more, more often and more regularly.

After identifying the possible content of the cards, I tested how the Insights should be triggered and delivered to users, and how often. To proceed I prepared different clickable prototypes via Marvel and went for another round of testing.

From the testing sessions, I learned the following things:
- People want to receive the Insights as the result of an action they performed. Tracking was the natural answer (vs: when they open the app or go to Calendar or Cycle view screens)
- People don't want to be interrupted during their tracking session.
- 3 to 5 times a cycle is the preferred frequency.
- People want to chose when and read the insights when they are ready and not be forced to read them.
- To iterate further and refine Insights content, it would be good to ask people to indicate if the insight they received was useful for them.

I presented then the first outcomes and analysis to the team and we discussed further steps and iterations.
SOLUTION
For the next steps, I prepared again Marvel prototypes with the final flows and tested them internally and externally. I finally came up with final flows and final Insights cards while Adeline, project manager, took care of the logic (frequency and rules) together with our content person and PO.


The final cards were the following:
- Average cycle and period length of user with a comparison to typical ranges.
- Average cycle length variation of user with a comparison to typical ranges.
- A series of actionable tips, "Good to know", in direct correlation with what the user tracked and/or the phase of the cycle they are currently in.
- A comparison to Clue users averages who tracked the same category during the same phase.
- A general piece of information, from scientific literature, in correlation with the current categories tracked and/or cycle phase.

We decided to go for a simple mvp and therefore excluded 2 other cards to keep them for future iteration, because of their technical and logic complexity:
- A visualisation of user's average cycle with the frequency of one selected category over time.
- The "Good to track with" card.

We also added the possibility for user to indicate if the Insight was useful or not for them. As all the cards use a similar UI template, this would allow us to iterate in the future and replace one type of cards not performing so well by another card type without too much design and implementation efforts.

I designed the flow users will go through to receive the Insights cards in a simple and non-invasive way. Users would track and once they are done and close the tracking overlay, they would get a permanent but non-invasive notification inviting them to read a new Insight. They could then choose when it's a good time for them to proceed. 

We discussed with the team the current flow and agreed to keep it simple and leave for next iterations the possibility to turn off Insights or to consult past Insights.

Next step was for me to refine the UI with Sketch and define the animations and transitions via Principles. I exported then the main UX flows and individual screens via Zeplin and defined the animations and UX/UI specifications for/and together with the dev team.
OUTCOME
Insights project was a success: more than 70% of people judged the cards useful for them and the second iteration of the feature is in preparation. It also helped to increase retention and frequency of tracking.
UX & Analysis process for Clue Insights Feature
8
174
2
Published:

UX & Analysis process for Clue Insights Feature

UX & Analysis of a project - Insights cards for Clue
8
174
2
Published:

Tools