An organisational red flag

Developing a simple filtering optimisation for WhatsApp, a well-used app with a surpilow NPS score.

The spark

I sat down last night to reply to 2 messages. This morning, I realised that I had completely neglected to reply to one of the two arguably very simple tasks I set out to complete. Why? I ask myself. WhatsApp had my engagement, but the 2 tasks were competing for attention with many less important and less time-sensitive messages.

Picking up your device for one thing and then spending one’s time dealing with something entirely different is not a new phenomenon. For an organisation like WhatsApp, engaged users generate data. The relatively new ‘updates’ tab in WhatsApp suggests that they are trying to generate a market for advertisers, but does a user struggling to complete a task really generate good data? Online user feedback is poor and WhatsApp also has a surprisingly low NPS score for a tool that so many people have actively opted into. Perhaps it’s a case of social proof gone wrong — should we all be on the more user-centred Signal instead?

What’s the problem?

WhatsApp is the DeFacto messaging App for myself and millions around the world.

I use WhatsApp to plan events, share moments, memes and much more. It contains a sea of personalised media and text-based information within its chats, groups and communities, so what happens when I want to recover a very specific bit of information?

Users open WhatsApp on average 20 times a day, it has a super simple + familiar UI, digging out logisitical information isn’t always difficult — popping a 1 or 2 word entry into the search field, locating information and simply copying or forwarding it on.

I use WhatsApp’s search function as a quick hack for accessing information that I only partially recollect. For example, I might type the name of the street that my sister lives on to find a message where she previously sent her full address — Hey presto, I can easily forward the full address to a friend who wants to a birthday card to my sister!

Thinking about my task distraction alongside my little search hack got me thinking about other search and filtering opportunities within the App. There are communities, which are public-facing groupings around topic areas or interests, you can also search for businesses. There is also a filter function next to the search bar on the home page that simply allows you to search all messages or displays only unread messages.

I propose developing an opportunity for WhatsApp where users may organise their contacts and chat content in a more nuanced form than by infinitely scrolling the homepage or by searching via exact text match / filtering read status.

Firstly, I addressed my assumption around the problem space — where better place to start than in WhatsApp? I set up a broadcast channel reaching out to some of my long-suffering friends and family most reliable testers.

With some guerrilla research and testing, I will validate or reject the case for this potential optimisation — or perhaps find an alternative opportunity along the way.

First Stop

Using some of the research tools available at Optimal Workshop, I built a quick and dirty study. Asking respondents a combination of open and closed questions as well as setting up a hybrid card sorting task.

Every respondent was a DAU of WhatsApp, despite my NPS findings being positive (20), all bar one respondent gave neutral feedback. From online research, my sample group seems to reflect the wider picture of WhatsApp: being the most popular messaging platform in many large European countries, while at the same time having a distinctly average NPS. What we can assert at a high level is that there is room for improvement somewhere in WhatsApp’s UX. Is the poor NPS score from something WhatsApp has done, or something it hasn’t?

To get a rounded picture, I started by asking some general questions about their experience with the App, this would help me understand the types of chat that users are engaged with (work, social, family etc.), the types of task users undertake in the app, as well as the context of these tasks.

The results from generalised questions helped to build a background to my core questioning around the topic of searching and filtering, it also give me insight into how to weigh the IA of any new search / filtering optimisations made to the UX.

I asked specific questions around the job of searching for information within chats / groups.

It appears that WhatsApp is increasingly used for work and similarly settings where pain-free access to accurate information is important. All respondents also regularly used WhatsApp to find practical information relating to work, life admin or similar. User’s of WhatsApp expect a more granular functionality than we might be happy to live with if it was simply an AdHoc cat meme repository.

Users either didn’t use the search / filter functions at all or they did but generally found functionality lacking. So the question for me was; how might I help users usefully organise their growing database of chats, groups and all of the information therein?

Based on my preliminary survey, I set about building an Empathy Map that aggregated the feedback gained.

This empathy map helped me develop a user persona to use as a platform to start working though human-centred design process.

A context chameleon

Building up an empathy map and basic persona threw up some contradictory statements, Initially I thought about how I could fine tune the persona so that it pulled in one clear direction, but then I reflected on the fact that for many users, contrasting needs or wants are a reality, even on an individual basis.

This factor can cause conflict for users, so having the opportunity to compartmentalise a user’s competing needs would allow WhatsApp to give users only what they need at the time they need it. Creating shorter, more positive interactions with the app.

Before diving into potential solutions, I drew up a concise How Might We matrix to help further define both the user + organisational needs in this scenario.

The challenge of designing an optimisation around the various and sometimes conflicting needs of a given user could cause major issues when considering the absolute inalienable prioritisation of consistent + simple UI, brand guidelines and more.

I set about looking for minor adjustments that could potentially add major functional and experiential gains that would have minimal impact.

Finding a simple, user-friendly solution for organising messages is the goal. Gradually adding the opportunity for more complex categorisation will give peak users extra functionality while not alienating edge users.

I had a look around the platform to see what measures were in place that offered similar solutions to this user goal — I also wanted to scout locations for the potential optimisation.

Speaking to my testers, we discussed their organisational strategies for messages and emails — such as tagging, starring, smart mailboxes, system folders etc.

Being able to organise messages quickly was important to testers — they are often already in task flow that is unrelated to the message that they categorise. By organising messages there is an inherent implication that the information in the message will be needed later. This means users are not completing tasks relating to a message in the first instance (who does?) or that they will later repeat tasks associated with the message.

Red Flags

In the end, I settled on a solution that would have a very minor shift in the UI, using a well-understood concept of colour-coded flags — this light touch would suit both user and organisational goals. Increasing functionality stratifying it in a way that keeps it simple for those who prefer it that way but at the same time allowing peak users to customise their flags which could then easily be filtered through the pre-existing filtering UI element.

Mark as ‘Unread’ function is retained without, giving users a choice in whether they adopt the new system or. Alternative routes to flagging could be incorporated in ‘holding down’ message menu, but this design has been left out for brevity.

The How Might We Matrix threw up an interesting further optimisation that links to the concept of Dunbar’s number. Social Apps like WhatsApp give us the tools to stay in touch, but they can also create a connectivity overload. For happier and healthier interpersonal relationships, our connectivity needs better management. Perhaps a story for another day….