Predicting human action has a variety of applications from human-robot collaboration and autonomous robot navigation to exploring abnormal situations in surveillance videos and activity-aware service algorithms for personal or health care purposes. As an example, in autonomous healthcare services, consider an agent- monitoring a patient’s activities, trying to predict if the patient is losing her/his balance. If the agent is capable of predicting the next action, it could determine whether s/he might fall and take an action to attempt to prevent it. DataSet for this problem can be downloaded from here.
The team was thrilled to see a 9x increase in use of the feature within the first week. It was satisfying to see a direct impact so quickly, but it was also gratifying that this campaign achieved the long-lasting impact on feature use we had hoped for. Three months after the reward offer ended we compared the use of the file-sharing feature to the 3 months prior to the campaign and found that there was an average 73% increase in use per user.
Optimize user engagement to maximize revenue. Strong, consistent user engagement over their lifecycle helps maximize monetization. Keep track of and optimize for continued engagement, as it is a predictor of future monetization and engagement.
Focus on user engagement on the first day. Higher retention and LTV is achieved from users who take a core action in their first day in the app. So, make sure you show people the value your app has to drive strong retention and LTV.
Understand your core engagement metric. Identify the engagement action in your app that correlates (or, ideally, predicts) LTV and spend. Consider how engagement is designed to tie to retention and conversion of your users.
Confirmshaming occurs when the design tries to make a user feel guilty by taking some action. This is regularly used to convince the user to sign up for a newsletter. Sometimes they try to bite you emotionally: “I’m not interested in getting more knowledge”, “I don’t care about children’s health”, “I would rather not keep up-to-date” etc.
…il and phone number verifications. Some of trophies can be achieved by numbers of Snaps users send. As they earn more trophies, the rest of the locks create curiosity gaps for heavy users; they go online and research the full list of possible accomplishments. Achieving full trophies has become a game for core fans, and many of them post their accomplishments accordingly.
Users can use Reddit Coins (formerly called credits) to award or accolade their favorite posts or comments. In other words, they get to have some power to influence the conversations that take place in the community. The recipients gain status, which motivates them to stay active on the platform.
… million views, we compared notes with the tracking efforts of the respective developers. Et voila: for every person who is clicking on the tracking link of an influencer video on YouTube, there are 4 more people who are not clicking but still download the game within four days after watching. We are calling this large group of users the ‘non-tracked majority’. The multiplier of 4x is good n…
This is the same process we go through when trying out voice search, Alexa, or Google Home. As soon as the technology exhibits an inescapable sign of life — verbal dialogue — we elevate our expectations of it. Now, it’s not just a tool. It’s on par with something that truly is alive. It’s on the hook for demonstrating manners, following social norms, and picking up on empathy. When it fails to live up to these heightened expectations, we judge it. We lose trust in it. All while rationally knowing what it really is, and what it can and cannot do.