Every company is becoming a data company. Data-Powered Apps delves into how product teams are infusing insights into applications and services to build products that will delight users and stand the test of time.
Applications of all kinds now face increasing pressure to offer analytics and stats to users on both sides of the fence: Subscribers and members want to see their personal analytics dashboards; business teams want to see data on those members as well as app performance. The trend of converting data into profitable insights is now an industry in itself.
As every company becomes a data company, users of all kinds are expecting analytics of some kind in the applications they use. Companies are recognizing that these insights have become basic table stakes for remaining competitive in the business world. The smart ones are finding new ways to monetize their data, either by embedding analytics into apps and services that existing users will pay for or using them to grow their audience and expand into new markets.
Let’s look at ways that some of the most-used apps, and the companies behind them, have already implemented data and analytics and what they’re doing with them.
Zoom and LinkedIn: Connecting people, collecting data
While Zoom’s core business may be videoconferencing, the data Zoom gathers for and about its users is equally critical to its success. The emergence of AI and machine learning-based predictive analytics will lead to insights from the data that will also prove important to the app’s competitiveness.
Customer conferencing usage stats as well as information around subscriber behaviors are both extremely valuable. The most obvious examples of the kind of data Zoom collects, which users can see themselves: Lengths of sessions, frequency of calls, etc. But more involved types of data that could benefit the company and eventually turn into new analytics products include:
- Tracking user location
- Sampling camera images
- Detecting other running apps
LinkedIn is another app/service that thrives on user data. Professionals of all stripes want to know how many LinkedIn visitors viewed their profile, how effective their recent profile update was, and more. With the massive proliferation of AI and machine-learning algorithms crunching numbers like these, social media users expect to be served with impressively detailed personal stats.
Increasingly, members and subscribers also expect to see predictive analytics brought to bear on their data, along with a presentation of insights in a dashboard with sophisticated visualizations. The trend of apps offering data and analytics to improve their clients’ experience is so important that the success of an app is becoming determined in greater and greater part by the effectiveness of its analytics.
Some of the most common analytics and outcomes expected today:
- Visitor count and a history chart of visitor frequency (new sessions and return visitors)
- How content updates affected relevance
- Whether a predicted preference yielded an actual purchase
- Which performance metrics led to achieved goals
Applying the method of using performance metrics to increase target achievement, for example, leads to a nearly infinite number of use cases across apps such as finance, fitness, communications, and education. Let’s look at five examples of apps/companies that are presenting user data to help drive engagement, differentiate themselves in a crowded space, and make themselves vital to their users’ lives.
Data-driven apps: Giving users the information they crave
As you’d expect from the section above, LinkedIn is a heavy collector and presenter of data. The oldest and still most-used professional networking site, it has a long history, and countless users have provided a wealth of data. Whole careers are built today using LinkedIn data to drive engagement, find top talent, influence ideas, and more.
The kinds of data-related questions that LinkedIn users of all kinds ask include: Does my content offer value/is it of interest? Did the recent updates to my resume lead to increased contacts from recruiters? Am I getting traction with my target audiences?
The data LinkedIn offers its users can and should be used to optimize the benefit users get from the platform. Higher levels of data and analytics often lead us to tougher levels of competition. LinkedIn offers the following types of data to help users put their best foot forward to employers and manage their professional personae:
- Update post analytics: Evaluate the effectiveness of your content updates, including photos and videos
- Followers and visitors analytics: Dive deep into your visitor demographics
- Talent Brand analytics: Companies that use LinkedIn Career Pages have access to Talent Brand analytics, giving users a picture of how their talent is talking about the company. Those running a Pipeline Builder campaign can see constantly-updated stats on the kinds of hiring leads they’re developing.
The COVID-19 pandemic made conducting online meetings a vital component of the modern business world. With countless meetings conducted each day, it was only natural that users would want to see the data behind their own sessions. Zoom’s Dashboard tab shows an array of user activities, including graphs of stats around users, meetings, and Zoom Rooms. Users can fine-tune which data is shown and how it is displayed; extending the potential, Zoom provides features to export user data for use in other apps.
Exporting user data is progressive, because Zoom knows that users will want to get maximum value from their conference data, possibly even applying ML to glean insights. Whatever your product or service, your users are creating data every day. Just letting them see how they’re using your product, as Zoom’s done, can be a powerful first step in adding data to your product as you create news ways for them to get more out of it within your platform.
Another communications app that experienced a surge in usage during the pandemic, Slack also recognizes the importance of providing its members with data and the potential for analytics intervention. Slack’s analytics dashboard includes information to help understand how a given workspace is used by a group. Charts show peak message levels as well as who sends how many. Everything can be broken down granularly to create a useful productivity picture.
Slack even accelerates business communication by organizing public channels by metrics like name, member count, and number of messages sent. Importantly, Slack built the potential for data and analytics into its platform as organic components. Inevitably, this is integral to Slack’s success today.
Shifting to the realm of fitness, Strava is a great example of how app developers can manage data internally so as to fluently make it accessible to all channels — users and engineers. Stava is a leading app for tracking athletic performance and goal achievement, an app where user data was vitally important even before the rise of big data and ML. Strava’s user performance data is more obvious than that of LinkedIn, so let’s look at how Strava developers built a model based on data. According to Strava’s engineering blog:
“Our data outgrew what could be handled on a single server, forcing us to partition it across multiple servers. As a result, any query which required a join between data on different servers could no longer be expressed purely in SQL. Not a problem for engineers, but a huge barrier for business analysts and other data-savvy, but non-technical staff.”
This is where Redshift, a popular Amazon data warehouse solution, was implemented to architect a data solution at Strava. Every night at Strava, application data from the previous day is replicated to a Redshift cluster. Strava features are instrumented through its logging infrastructure. When an athlete views performance data, an activity or a live data feed on Strava, this triggers Strava’s Kafka event streaming to log the data. After streaming to Redshift, the data can be queried and piped to app dashboards. Now, all athlete performance data is ready for insight mining.
Rosetta Stone is one of the world’s most popular language learning systems. Tracking a goal can be an important part of accomplishing it, so building insights into Rosetta Stone was a natural fit. The company embedded a robust learning management system called Catalyst to handle this.
Catalyst includes an innovative smart dashboard that quickly displays learning target achievement metrics in a way that’s easy to understand. Giving students frequent updates in an interface like this helps them understand the progress they’re making and stay engaged with their learning journey. One innovative use of analytics in language learning is identifying topics that pose particular challenges to students. As with Zoom, the ability to export performance data to external ML apps could also lead to improved educational methods.
The impending data reality
Though adding data to apps is still in its early stages, a variety of metrics have already been identified as critical components of success for apps today. Increasingly, enterprises are finding that it’s smartest to build data and analytics into their software from inception. This shift means that data science skills will become even more important to developers.
The upside of embedding data and analytics into apps (for both companies and users) cannot be understated. And the alternative is as grim as that thought was rosy: Apps and services that do not start adding data and insights will fail.
Eitan Sofer is a seasoned Sisenser, having spent the last 13 years building and shaping our core analytics product, focusing on user experience and platform engineering. Today, he runs the Embedded Analytics product line which powers thousands of customers and businesses, making them insights-driven. Eitan is also an avid music fan and surfer.