Privacy has always been a concern with teenagers, especially in context of internet messengers. To address the need, hike provides a solution “Hidden Mode” to it’s users. Data shows millions of messages daily passing in hidden mode, which validates that, it did address a need. Then what was the problem with it?
To understand it let’s get a know-how of hidden mode. Varied approaches are taken to maintain chat privacy in IMs, 3 major approaches are seen:
1. A security lock over the app Ex: Applock
2. Deleting particular content after a time period Ex: Snapchat,Telegram etc
3. Hike’s Hidden mode
Hidden mode 1.0 allowed users to hide their chats and protect them with a password. At the same time it allowed users to access their non-hidden chats without entering any password. Also, it keeps information safe for users to view it later. It follows a 3 step process, Tap on ‘hi’ at the top, then a dialogue appears to enter your set password. On entering the correct password, your hidden chat appears.
Primary research
Looking deeper in data, we found some astonishing stats. From users who initiated hidden mode, 80% dropped off during the setup and those who completed the setup were highly active recurring hike users, data showed us that hidden mode users we just 10% of total hike DAUs, but they were responsible for 20% of total messaging volume.
Quantitative data tells you one perspective of problem, to know the rest we listen to our users. After reading through a bunch of mails, feedbacks on play store and reviews on various portals, it was seen that:
1. Users were unable to discover/initiate hidden mode.
2. There was friction in process of hiding chats.
3. Users were missing out on messages from there ‘hidden’ contacts,
4. Unable to differentiate between hidden/non-hidden chats inside HM.
5. Users were unable to change their password and reset HM.
6. Some enhancements were requested like, giving option to have a PIN as password for hidden mode.
After a Q&D primary research, stakeholder interviews were conducted and scope of project was finalised taking into account business goals and to fix following problems:
1. Hidden mode discovery.
2. Making hidden mode easy to setup.
3. Improving the experience of regular hidden mode usage.
Investigating pain-points — The why
After having knowledge of user pain points, we sketched up user journey maps for each task to understand reasons behind them. User journey was segmented into discovery, setup and regular usage of hidden mode. On drawing empathy at every touch point we were able to infer following reasons.
Hidden mode discovery and ease of hiding chats
To initiate hidden mode users had to tap on hike logo. After a successful hidden mode setup, user had to again, tap on hike’s logo to enter hidden mode and then hide their chats. We, evaluated this process though user journey mapping and trying to empathise with our target users at every point.
We saw that this was not an intuitive approach to hide a chat. This solution also lacked in flexibility to initiate hidden mode or hide a chat when there is a motivation.
So we asked ourselves why, when and where would a user hide a chat. From the brainstorming session we came to conclusion that a decision to make some entity hidden is due to either:
1. A person is to be kept secret.
2. Confidential information (here messages) to be kept secret.
When a user can makes a decision to hide such information in hike?
1. During a conversation.
2. Wanting to start a conversation/person which users thinks is now needed to be secret.
Where in hike can a user make this decision:
1. On conversation list.
2. Inside a chat.
Hidden mode setup drop-off
Setting up hidden mode was an elongated process.
On android platform, which can be categorised in 3 broad steps:
Trigger hidden mode->Hide a chat->Setup password.
On windows platform the flow was:
Trigger hidden mode->Setup password ->Hide a chat.
Windows flow had better conversions. On analysing user journey, it was seen that:
1. Redundant information increased the steps to complete setup.
2. Setup information was not in correct priority.
Regular HM usage pain-points:
- When a message from hidden mode contact is received, notification was seen on the user’s phone. The problem was, these notification did not display either the identity of the sender not the content of message, and when a user selected them there was no feedback inside hike to indicate the message. Result of which users were missing out on messages from their hidden contacts.
2. From the reviews it was seen that users were unable to reset hidden mode and change their password. Options to do both these tasks were provided in privacy settings. User could access these from app settings or by tapping “change password from settings >privacy” text in pop-up (as shown below in figure). Problem was this text didn’t provide sufficient affordance for user to understand that it was tap able. Also settings for hidden mode, “Reset hidden mode” and “Change password” were placed at bottom in privacy settings and were difficult to find.
To validate the user preference, an open card sort followed by a closed card sort study was conducted to categorise and prioritise information inside privacy settings. It was seen that users prioritised all the hidden mode settings above the other privacy settings and categorised it under one label.
Solution — The How
After knowing these reasons, it was time to get dirty and debate on ideas for flows which solved these pinpoints.
Next, bringing these ideas to life, we started sketching wireframes.
To get more closer to the real flows, prototypes were built, tested and iterated. These prototypes were then showcased to stakeholders for feedback. With minor tweaks the final flow and UI was finalised.
Some high fidelity prototypes to iterate on interactions: https://dl.dropboxusercontent.com/u/315449090/context1.framer/index.html
After, thousands of lines of code and weeks of regressive testing hidden mode 2.0 was ready to be shipped.
Results:
- Hidden mode discovery and regular usage of hidden mode
For a distinct set of users activating on new hike versions (with hidden mode 2.0) vs on older hike versions (with old hidden mode), observed for one week
• Stealth Adoption was 167% higher for new builds(until EOM).
• By D11 after activation, the % of DAUs using stealth was 96% higher for new builds.
- By D11, the hidden message contribution to Daily H2H volume was 48% higher for new builds.
2. Ease of hiding chats/setting up hidden mode
Nearly 3 out of 4 mark hidden events were attributed to new ways of hiding chats, with Long Tap and hiding chats making up 43% and selecting from 3 Dot menu item making up 30% share.
Long Tap: Hiding directly from Long Tap from main screen but outside hidden mode.
3 Dot: Tapping on 3 dot menu item from inside chat but outside hidden mode.
