I needed to know whether my frustration was a shared sentiment among WhatsApp users. So I conduced a user survey among a cross-section of WhatsApp users - millennials to home-makers to working professionals.

The survey focused on validating whether WhatsApp users found it difficult to search and locate media on the platform. These were the main findings:

Based on these findings, it was clear that:

I mapped out the business pains and goals as well as the user's pains and goals to help come up with a solution.

I needed to place some constraints on myself before I could start thinking of possible solutions:

Keeping all of this information in mind, I came up with three possible solutions:

Let's break down each of these solutions in detail. The first idea was a simple one and it worked on the principle that allowed the user to narrow down the search results as long as they knew "details" about the media message they were looking for. So what happens if the user (like me!) was not able to recall any details? Then, unfortunately they would not be able to locate it. With the second idea, it seems like an overkill for WhatsApp because at the end of the day, the WhatsApp user is not searching for their media message across the internet.

The third idea had the most promise and it piqued my interest. It had the advantage of being familiar amongst social media users which made it familiar and intuitive. At a first glance, this idea had legs but it needed to be fleshed out more so I could see how it could be integrated within the WhatsApp platform.

If you look at the tagging system from a social media lens it can be understood as having a very limited functionality. It is mainly used to help social media users locate public posts and not personal or private media.

Instance 1: you can use search for tags

Instance 2: Tags can also be used underneath posts

But interestingly, if you look at file management systems like the color coding system on the Mac systems (MacBook as well as the Mac desktops) there is a color coding tagging system used to help users tag their private files.

This in-depth understanding of how flexible tags can be will help push their utility within WhatsApp and make space for a possible multi-purpose feature.

Before jumping into the design process, I created a user flow of how I assumed the user would interact with this feature within WhatsApp and let it guide me through the design process

As mentioned earlier, it was important to see how the existing system on WhatsApp would accommodate this feature. So I sketched out different ways my users would interact with this feature

So far so good, right? The solution seemed to fit within the WhatsApp ecosystem quite seamlessly but there was a small hitch with this solution. The use of tags was supposed to help the user search for media messages not to organize going forward.

How do we use this system to help locate media messages that the users have received in the past? Am I expecting users to manually tag past media so that they can search for it later? That would be extremely tedious and quite honestly I don’t think users would even use this tagging feature if that was the case. So how can we help users avoid tagging past media messages manually?

That’s when I thought of AI/ML smart technology. Is there space to use an intelligent AI/ML engine to learn from the tagging going forward and provide auto-suggestions for the users to tag media that’s already in their media repository?

I explored this idea with the following sketches:

If WhatsApp is successfully able to sift through a user's media repository, identify the appropriate files, and suggest them to the user, they would successfully eliminate the ennui of manually tagging files.

My low-fi sketches gave me the clarity and confidence I needed to bring my design ideas to Figma and jump straight into creating the high-fidelity designs.
This project was conceptualized on an iOS device in dark mode - I used the following resources on the Figma community to help me recreate the WhatsApp screens:

These are a few of the key screens:

I conducted a remote moderated usability study with 5 participants. I recruited two individuals who were familiar with social media and three individuals who were not. They were asked to complete the following tasks:

Overall, my participants completed the tasks with a 100% success rate, and many expressed enthusiasm and curiosity on how the feature could be expanded upon.

I was quite pleased that my users were able to use my feature with ease and navigate through the tasks seamlessly.

However, there was a significant Aha! Moment: my initial assumption of how users would search for their tags was proved wrong. Referring to my original task flow, I assumed that my users would go through the search bar to find their media using tags however they all went through the “#Tags” button.
One user said “that's where I created the tags so I'm assuming that that is where I could go to find all my tags and associated media”. 

Another significant moment was seeing my participants react to the surprise AI/ML notification in real-time: WhatsApp AI photo-identification system has found more photos for you to review under your new tag "MyNephew".

All the participants explored this notification (without prompting) and reviewed the photos, and then added them to the tag they created in the first task. They were all pleasantly surprised and were quite impressed with the feature.
This surprised me because this AI/ML engine was more of an after-thought when I was solutionizing but it seems to be the secret sauce that will help with the adoption of this solution. 

This project is really exciting to me because WhatsApp is one of the biggest messaging platforms on the market and I believe this feature could add a whole new dimension to WhatsApp. I also believe that this feature could be expanded on and these are some of the ways I came up with:

I plan on spending some more time post-DesignLab to really explore some of these ideas.

Interested in reaching out and talking about my work? Contact me here!