Building Better Period-Tracking Interfaces

A long look at fertility and period-tracking apps raises questions around how developers can make personal data interfaces better and empower women to use data for their own good.

Tracking female reproductive cycles is, obviously, hardly new. It’s right there in the actual name of the thing — “menstruation,” from the Latin “menses,” aka “month.” Women have been tracking cycles from ancient times using everything from moon-based fertility rituals to paper (Kimberly-Clark’s Disney-produced film, The Story of Menstruation, from 1946 recommended a quaint little booklet for period-tracking called, quite delightfully, “Very Personally Yours”). And now, of course, we can use smartphone apps, digital basal thermometers, and, in theory anyway, data aggregation through the “quantified self” movement, to not only track this data but to correlate multiple overall health metrics and actually utilize it.

Tracking menstrual cycles can be beneficial whether it’s for health, fertility, or just wanting to ensure that you can wear white jeans without fear. We believe this kind of personal data tracking is an opportunity to give women the knowledge to take care of their health, fertility and white pantsuits, and even, according to some research, use our hormonal cycles in a positive way, leveraging our unique strengths rather than trying to conform to a male-oriented work cycle.

As data is used more and more to actively improve our health more and more—from our phones automatically tracking our steps to sensors that monitor blood glucose levels without needles and even portable EKGs for under $100—let’s not forget that roughly half the population will menstruate at some point in their lives. Which is why many iPhone-owning women were mystified when Apple failed to consider menstruation worthy of inclusion in the basic Health app until 2015. Don’t worry, it did previously include selenium levels, inhaler use, and perfusion index (shout-out to all you pulse oximeter owners out there!). Even now that it’s included, it feels as if it were designed by someone who has never had a period in their life (which may well be the case, since Apple’s workforce is roughly 70% male).

In this piece, we’d like to identify the pain points of current period-tracking app users, and consider what Apple and other developers could do to improve. How could we make these apps and interfaces better? How can we use emerging technology to enhance features, create ways to partially automate data input, and empower women to use this data for their own good?

It’s no secret that venture-capitalist funding — an industry that is 93% male — is inherently biased towards men and men’s needs. Even Ida Tin, the entrepreneur behind very popular period-tracking app Clue, struggled to get funding, as male investors found discussing periods “taboo.” For the development of effective products, this is a bad place to be.

A sense of taboo or “period shame” is anathema not only to funding, but to the essential principles of user-focused design: talking to actual women about their periods, watching how they use your design, and learning about their real needs — even if it’s not what you, or they, expected.

So, let’s talk. Let’s start from scratch. Let’s rethink why we track periods and what we want from that data. Then let’s see how we might apply user-centric design and emerging technology — from sophisticated wearables to AI and medical research — to improve them.

We sent an admittedly unscientific survey out to some of The Lodge’s and Wieden+Kennedy’s email lists and were deluged with responses. Our survey of 93 friends and colleagues showed that over 90% of women* currently track their menstrual cycle, or have in the past. A similar survey by the University of Washington showed a figure of 89%.

*Note: All of our respondents voluntarily identified as women; we understand that not all people who menstruate do so. We did not ask participants to disclose their sexual orientation, but we know that we had multiple sexual orientations among our participants.

According to our survey, the top two reasons for period tracking are “general health” (60%) and “being prepared (e.g., having pads and tampons)” (56.5%). Interestingly, the next two reasons, at 36.5% each, were for fertility/ovulation and contraception, followed by identifying health problems (27%) and pain management (22%). Other reasons (5% or less) included providing medical practitioners with information, as well as help in managing supplements, moods, and specific health issues such as endometriosis, hormonal acne, and migraines. These reasons tracked well against the UW study.

1. How might we personalize apps to take into account different cycles?

As we discussed, there are a variety of reasons why women track their menstrual cycles. Unfortunately, while there were a few women who loved their current apps (in particular Clue, Flo, and Eve), nearly all of them fell short — and the most common complaint was one that retrospectively seems pretty darn obvious: no two women, and no two cycles, are the same.

“There is no option for if you don’t have a period most of the time (I’m on an IUD) and so it thinks my cycle is like 74+ days long. It’s got the data points I like but it can’t help me predict anything due to that problem.” [Clue]

In reference to the above comment, Clue (which is actually pretty comprehensive in its efforts) does actually allow you to track birth control, medication, and ailments, but it seems that isn’t accurate to the cycle. We tested this out: we indicated we had an IUD and entered some random days of bleeding, and the ovulation and “day of cycle” days immediately changed to indicate fertility overlapping with menstruation (see screenshot), so it’s easy to see why women are confused.

Clue screenshot — Apps get confused when nontypical data is entered

Even excluding adolescent women and women who are approaching menopause, women’s “normal” cycles can last anywhere from 20 to 40 days. This doesn’t take into account the effects of different methods of birth control, particularly IUDs, pregnancy itself (weirdly, one woman in the UW survey said after having a baby and breastfeeding, her app thought her cycle was 700 days long!), or other factors, including but not limited to surgeries, stress, illness, medical treatments, eating disorders, and athletic amenorrhea (when highly active women miss periods or stop menstruating completely).

Most of the women we surveyed said that their period-tracking app could not accurately account for any of these.

“I’ve seen apps that practically scream YOU’RE LATE but that’s really not helpful when you have issues that cause you to not be on a straight 28-day cycle.”

At a basic level, there needs to be a simple way to input key factors that affect your cycle and have this taken into account as your cycle is predicted.

At a deeper level, a company called Freda in the UK is already offering algorithmic solutions at a consumer level, using cycle information, which improves over time, to ship menstrual products. There is no reason this tech couldn’t be incorporated in a period-tracking app. (According to VentureBeat, Freda has secured zero funding. We couldn’t find information as to why this is but we suspect it’s something to do with those 93% male VCs).

Having said that, Freda’s site says they use an algorithm similar to that of fertility apps, which still need further research and improvement. Dr. Robert Setton of the Weill Cornell Medical College authored a paper on this, commenting to Time magazine, “How could you mess something like this up? It’s so simple.” With all due respect to Dr. Setton, apparently it’s not — and possibly because these algorithms are not currently taking into account all the factors we just mentioned above. The app Flo, however, had high marks from our users, but several App Store reviews report inaccuracy when faced with issues such as miscarriage and toggling between the “Pregnancy” mode and others.

Apple’s Health is probably the worst offender here. It treats periods like any other form of measured data, like inputting a glucose reading — but periods are not a finite measurement. For example, if you enter the start date of your cycle, but then realize later it was actually just spotting, you can’t change the entry. Even worse, if you then enter a start date on another day, both entries are noted as the start date of a period, making the tracking highly inaccurate — an extremely basic error, which should be easy to fix. Despite the fact that Apple apparently reached out to Ida Tin to hear her “perspective on what’s important and what data they should collect,” it doesn’t seem to have been taken into account.

Given this knowledge, how might we personalize apps to take into account different cycles?

  • Give users a way to input factors that could affect their cycle.
  • AI algorithms should be developed that take this user input into account,
  • Don’t make the entire structure of the app dependent on predictable cycles — take a more multifaceted approach.

2. How might we make apps more inclusive?

“It feels a little complicated for what I want to track — I wish there were a way to visually simplify the screen if I don’t want to track water consumption or sleep in that app.” [Clue]
“I feel the logging feature is limited; there should be better options for detailed reporting of symptoms.” [Flo]

As we discovered earlier, there are multiple reasons for tracking one’s menstrual cycle. Despite the fact that the top two are consistently “general health” and “being prepared,” most period-tracking apps are focused on fertility and conception. This does not take into account many other use cases: very young women who are not (yet) tracking fertility or even necessarily ovulating, perimenopausal women, women in same-sex relationships, and the general majority of women who are tracking for health reasons only. One of our respondents noted that she had had a hysterectomy, so a lot of the fertility focus is not relevant to her either.

Flo screenshot — You cannot turn off the “Chance to get pregnant” notice

Additionally, women who have fertility issues can find the forced reminder of their “fertility windows” painful:

“I also tried and really liked Dot, but when I did there was no way to stop it from telling you the best days to get pregnant. In my case that was actually a painful reminder so [I] settled on Clue where I can toggle on and off all the features.”

And women who have had miscarriages are, in some apps, forced to toggle back from a “pregnancy” mode, which is not only emotionally impactful but also screws up their cycle tracking:

“I miscarried last month and had to switch back from pregnancy mode. While I logged all of my miscarriage bleeding as “period” in the app … now it’s telling me that my period should start now … miscarriage throws your cycle out of whack. It is very frustrating knowing that I’m going to have to go in and fix the period log every day and have the app tell me that I’m ‘late.’” [App Store review, Flo]

The best solution to this is user testing across a wide range of use cases, and giving the user control over what information is surfaced in the main interface (Clue does this to some extent, but fertility is still emphasized). Flo has attempted to split their app into three use cases from the first screen: “Track my cycle”, “I want to get pregnant” and “I’m pregnant”. But as one user points out, that can also be painfully exclusive:

“My biggest issue with it is that it is not inclusive for LGBT people. As a queer woman, I have no risk of accidental pregnancy and I do not need information on condoms … other than feeling entirely excluded and like my identity and my life do not matter whatsoever to the developers, this is a pretty good app for tracking your period.” [App Store review, Flo]

To make apps more inclusive:

  • Users should be able to exclude data they do not feel is relevant to them.
  • The interface should only surface supporting information that is relevant to the use case at hand.
  • Apps should be tested across a wide range of use cases, including edge cases like hysterectomy patients, menopausal women, transgender women and adolescents.

3. How might we make data actually useful?

“I would put in my symptoms each day, but not actually sure how these were tracked/factored. Unless I went back to look at what I had entered each day, not sure it was useful. My symptoms didn’t seem to amount to any real useful data.” [Clue]

This is a key issue with Apple Health — even though it only tracks period start and end date, as well as flow, there really is no meaningful correlation made between that data and other Health tracking metrics such as sleep, exercise, or anything else (also, there is no option to track pain or mental health). You can’t even view different months in comparison to each other without zooming out to a “year” calendar view, which reduces each period to a tiny dot with not even the start date visible unless it’s your latest cycle. This makes the data literally useless. (You also can’t enter data beyond the current month, even historical data, so we couldn’t enter enough data to show more than one month).

Apple Health’s “Year” view makes comparing your cycles over time fairly useless

Without at the very least aggregating the input data into charts that can be tracked over time or, even better, correlating the different inputs and providing context and further information, a period-tracking app is not much more than the “Very Personally Yours” booklet from 1946.

“It was easy to use, but as always it’s tricky to discover milestones in real data vs perfectly designed examples. Even with my data visualized and graphed it was still tricky sometimes to exactly know what was going on.” [Kindara app]

To be useful, any data visualization needs to take into account the user’s personal input, rather than expect a perfect 28-day cycle with no anomalies.

Additionally, correlating this with educated algorithms could produce valuable medical data and help identify health problems such as endometriosis and flag this for the user if relevant. For example, if you indicate that you have pain at the beginning of your cycle, or feel depressed every time you have low estrogen and progesterone, the algorithm could track these two metrics and understand the relationship between them, and explain what these symptoms could mean and what actions you could take.

The most common pain point for our survey respondents was a lack of meaning from tracking all their data — and this was especially the case when the user was very young and didn’t have much experience with periods, or approaching menopause and experiencing irregularities they hadn’t before.

“It’d be great if those logged symptoms, over time, pointed us to pertinent articles.” [Flo]
“I wish that it would tell me WHY those things were happening. I wish I knew *why* I get cramps the first three days consistently, or why my flow will be heavy and then have a steep drop off, etc. This is probably stuff you learn in health class, and I just don’t know, but we watched Intervention in my high school health class which isn’t helping me too much here.” [P Tracker Lite and Clue]

Clue’s cycle view is a useful visual alternative to the calendar view, and Flo does have a “Patterns of Your Body” graph, which helps to visually make sense of some of the data, but a comprehensive visual overview with inputs selected by the user (e.g., “compare mood, flow, and sleep”) could help identify more personal meaning behind symptoms.

To make the data actually useful:

  • Data should be correlated — at a minimum to track symptoms over time, and preferably to combine symptoms and produce relevant analysis.
  • Data should be correlated to flag potential health issues, pain management, or supplement tips.
  • If an algorithm parses the data and identifies a health risk, the app could suggest the appropriate next steps to take.

4. How might we reduce the burden of data entry?

This was an overwhelming request from our survey participants: a period app is useless if you don’t enter the data into it. Tracking your period and symptoms shouldn’t be yet another chore to worry about. Nearly every option (including manually using a calendar) and app had a comment about this in our survey:

“I generally dislike the manual process of it, I never end up remembering to key stuff in which is how I end up eventually never tracking it.”
“I used it several years ago, probably when period-tracking apps first came out. It was hard to use, and I had to click on a number of different areas just to log one day.” [P Tracker Lite]
“Some apps make it feel like a chore. Some apps that are simple and well designed don’t include or leave room for tracking unlisted symptoms or pain levels. Others ask so many questions it feels overwhelming.” [Flutter, Phendo]
“I wish it was better at alerting me to fill in period information. It’s so easy to forget! Then you lose track and can’t remember.” [Flo]
“Great UX but sometimes I get tripped up on filling out all the options and it isn’t as clear to mark done on the fields I care to fill out.” [Clue]
“I also wish it was easier to track which days I experience pain or headaches. It’s several clicks to enter that data and I’m lazy.” [Clue]

Firstly, the ability to simplify options by excluding those you are not interested in would help here.

Multiple options available in P Tracker Lite

Secondly, what if we rethought how we entered data into these apps in the first place? Fertility tracking bracelets do exist (at 89% accuracy), but what if we used existing wearables that were not so fertility focused? A simple interface that combines detecting physiological data with very simple manually entered information — a tap for bleeding with an icon for flow level, then a simple swipe to input your chosen data points.

Biometric sensors could also easily be adapted to feed data automatically to an app — small sensors that can be stuck to the body like a temporary tattoo — or even, as Ida Tin has suggested, molecular sensors that can be inserted into the body much the way birth control implants currently are.

On the simplest level, what if you were given a simple, natural language interface for each day you enter data that guides you through the pertinent metrics for that exact stage of your cycle? Even better, why go into the app at all? Just text your friendly period-tracker chat bot a comment like, “Just got my period,” and it could answer “Are you sure?” You could answer “No, it might just be spotting,” and the app would flag that data and check in later on. Stanford University has been trialling a mental health bot called Woebot, which takes a similar approach, and it’s friendly and fun — it even has hedgehog GIFs. (There’s a promising if very limited chatbot called Izzy, although it’s on Facebook Messenger, which is not secure, and another called Florence, which is much more clinical.)

In another approach, what if we could scan our pad/tampon box with our app, and that would automatically set the start date of our current period? Supplemental data could be added manually, but if your menstruation were entered automatically for that cycle, you would only need to tweak the data rather than enter every single day manually. The app could remind you when the end of your period is estimated to approach, so you can add in any missing data when you have time.

To reduce the burden of data entry:

  • Data entry should be minimized to the relevant information for that particular point in time.
  • We should rethink physical ways to enter data: physiological tracking, wearables, sensors.
  • We need a more intuitive way to enter data with gentle reminders at key times — a chatbot or machine-learning-powered interface that takes both automatic input and prompts for manual input.
  • Consider ways to automate the process: scanning physical objects like tampon boxes to initiate the period section of the cycle.

5. How might we ensure data is exportable?

“Lost all my data a few times — would be good to download also so I can take to doctor appointments (I use as a symptoms diary for endometriosis & adenomyosis).” [OvuView]
“I don’t want to switch and lose my data.” [Period Tracker by Monthly Cycles]
“I wish there were better ways to export my data and adjust if I changed when I use my contraception without having to start my data all over again.” [MyRing]

For several reasons, users need their data to be exportable. Firstly as a backup, because you can conceivably rack up quite a data set very quickly.

Secondly, users should be able to share a standardized data set (such as a spreadsheet or well-formatted graph) with their medical provider. For women who choose to share their data with partners, it would also make sense to have a version that can be exported and shared, so you don’t need to give anyone access to your app.

In these last two cases, the user should be able to select what data is shared and protect the rest.

To ensure lasting data:

  • Users should be able to back up their data privately
  • Users should be able to choose which data is exported in a usable format to share with partners or medical professionals

6. How might we improve the look and feel of these apps?

“It’s horridly ugly and not intuitive.” [Period Diary]
Period Diary screenshots — “Ugly,” but at least the language is friendly and it has a supportive forum
“It’s ugly as all hell, but it works great.” [P Tracker]
Period Tracker screenshots — Note: “ugly” and with “pink flowers,” but we did think it was cute that the flowers bloom over time
“Most of the others were using pink/purple and flowers/cute animals. News flash: You can have a period and have good taste (+ be a grown up who doesn’t like overtly stereotypical feminine things hah!).” [Clue]
Clue screenshots — The cycle view is useful, but the reminder…not so much

Nearly everyone complained about “ugly,” “not intuitive” apps, and the stereotypes of pink flowers and cute animals. On the other hand, while most users preferred clean interfaces, if they were too clean they were considered “clinical.”

Apple homescreen icons for Period Tracker apps

In many cases, reproductive health is infantilized, and often sexualized even if the app is not for fertility tracking. Some emphasize the “generic” 28-day cycle right on the app icon.

Obviously there are cultural and personal differences at play here, and the most important thing is that information is meaningfully presented. However, as you can see in the iconography above, flowers and hearts are overrepresented, as symbolic as they may be, along with cute characters (some of which are sperm) and a lot of pink, which left many of the users we surveyed feeling patronized. These are apps you’re going to potentially use daily or at the very least monthly, so aesthetics are important in maintaining use.

This gets our unofficial award for Worst. Icon. Ever. [Period Logs app]

Some users reported changing the name of their app from “Period Tracker” to “Tracker” on their home screen, and others preferred using calendar apps, paper, or their “Notes” app because it was more private, so discretion should also be considered in the design of the apps themselves and their home screen icons, as well as the format of reminders and notifications. Having said that, normalizing periods is also an important and a valid consideration, but ultimately, some women would prefer to not have their phone home screen broadcast where they are in their cycle.

Additionally, tone of voice is important, and while most apps are friendly, some feel judgmental or patronizing, alarmist or babyish.

“Heaven forbid you admit to Glow that you’re feeling ‘moody,’ ‘stressed,’ and that you had a couple drinks — Glow bombarded me with so many lectures, I thought I was going to get grounded and sent to my room to think about the consequences!” [Glow app review on Bust]

The way the app “speaks” to users is an important part in continuing engagement and ensuring women feel that the app is helping them in a positive way —that is, being a source of information without being judgmental.

To improve the look and feel of these apps:

  • Apps should be clean but not too clinical.
  • The option of discretion needs to be taken into account on home screen icons, home screen app names, and reminders.
  • Even if specifically aimed at teens, design shouldn’t infantilize women’s health.
  • Tone of voice should be friendly and supportive, not judgmental, intimidating, or patronizing.

7. How might we optimize reminders?

“When I was trying to conceive it wasn’t intuitive [at] telling you when you were [ovulating]. It seems convoluted and the app itself isn’t user-friendly (at least the free version isn’t).” [Period Tracker]
“I like how easy it is but it doesn’t always send me a reminder like a few days before and I wish it would consistently.” [Clue]
“Like that it tracks symptoms, wish it gave me more effective reminders/notifications [Clue]

If fertility and being prepared are two big use cases for these apps, it makes sense that reminders need to be reliable and consistent. As discussed above though, they need to be discreet. In addition to reminders alone, the data visualization within the app should make it obvious if you are ovulating, if relevant. The Clue app does a pretty good job of this aspect; however, as a user above said, we found the reminders to be unreliable — and they only appear within the app. If a user has to remember to check the app two days before her period, it seems to defeat the purpose of reminders.

Most of the apps mentioned used native notifications, but that can be quite intrusive if you’re, say, in a meeting. What if your app sent you subtle, funny, or encouraging texts or a private voicemail? Or automatically added calendar reminders throughout the month (with privacy settings you choose)? Or Slacked you fun GIFs, depending on what the reminder was for? Or updated your shopping list ahead of time so you aren’t caught without menstrual products?

Again, Woebot is the best example we could find of a friendly, private, and conversational approach to health care.

To optimize reminders:

  • Find more natural, discreet ways to communicate with users.
  • Make reminders more fun and supportive, instead of scary and clinical.
  • Make sure reminders are timely and reliable.

8. How might we take a more holistic approach?

“I wish the apps out there took a more holistic approach and focused on creating something that could work for all women based on their needs.”
“I wish it would aggregate the symptoms I log so I could see a calendar view of like ‘expect this at this day’ — based on MY data, not just what generally other women experience. It’d also be cool to get notifications like ‘hey you usually get 10,000 zits during the next 4 days, so try a face mask!’ or something cute like that.” [Flo]

This is where AI could be a very interesting solution, building on all of the points above. A period is not an isolated event; it is connected to so many other symptoms. What if your period app understood this and intelligently analyzed your data to produce smart recommendations? Even better, what if it did this as a friendly AI-powered assistant? It could tell you when to try a face mask, order you a customized box of tampons, let you know at which time of the month you’re most likely to feel anxious or energized, remind you about birth control, and check in on how you’re feeling to subtly remind you to enter some data, assuming you haven’t installed your subcutaneous microsensor by now. SpotOn does sometimes send a notification asking how you’re doing, but it seems to occur at random times, and that fact alone doesn’t motivate a user to enter any information. And Hormone Horoscope does attempt to track the effect of hormones, but from what we could tell, there are no holistic and customized apps that convey this information in a natural way based on your personal data.

For a more holistic approach:

  • Machine learning and integrating a wide range of input could build a friendly, supportive assistant.

9. How might we use aggregated data to provide more context BUT also preserve your privacy?

“For me personally it took forever for it to become regular, this was before Google and being able to look up such things on the internet. Having to go to doctors and talk about that was scary at the time, knowing that what is ‘normal’ for one is not ‘normal’ for another, and it is scary to talk about so having some experience to read up could be nice.”
“I want to know what works for people who are similar to me.”
“Would be interested in seeing a large dataset of women’s experience with their cycles.”

A lot of our respondents said they would be interested in knowing what other women go through; even though we’re all different, it helps to know that there is such a variety of experience out there, and the cultural taboo around periods means a lot of women don’t get access to that information, particularly teens. Some of the apps do have forums, which is a nice touch.

Of concern are reports that a lot of these apps sell their users’ most private data or expose their users to hacking risks. Some submit their data anonymously to universities or government organizations. Flo also displays some aggregate data based on statistics, but most apps could use that data for good. Why not aggregate anonymous data to benefit users?

If you choose, you should be able to contribute your data to a pool to see aggregated data on other women’s cycles, and this data could be used to improve the accuracy of algorithms and health issue identification, provide relevant information from people with similar symptoms or age ranges, and help to identify edge cases to improve apps for all users.

To leverage data to provide context while protecting privacy:

  • Aggregated data could be used to improve period apps, not just contribute to medical research at institutions.
  • Hacking and selling of private information is a concern, so any data sharing should be strictly opt-in and in encrypted storage.

Conclusion

Overall, there’s a lot of potential to combine the best features of these apps and add machine learning, aggregated data, chatbot-driven interfaces, physical computing, and more meaningful data visualization to make these apps better for users. Apart from some cultural taboos, this will also require increased funding for apps that are targeted mostly at women — and that is, of course, another story.

Illustrations by Ryan Peltier.

In this Reproductive Care series of W+K Lodge Frontiers, our team of designers and engineers explores how brands can create better products and experiences for women, children, and their partners. Join us as we consider what it means to deliver not just utility but care along the reproductive journey. If you’re interested in collaborating on any projects or learning more about W+K Lodge, email us at frontiers@wklodge.com.