Thursday, July 28, 2016

Developing a Mobile Analytics Strategy

Would you invest in a stock if there were no way to track its valuation?  Owning the stock would be a blind venture:  no way to know if you should buy, sell—or if you’ve already lost your shirt.

Similarly, to make confident investments in a mobile products portfolio you need to establish clear measurement of progress against business objectives, proactively track performance, and create actionable insights to guide future product development.  

Mobile analytics enable you to execute on all these critical tasks.  And, according to MIT research, top performing companies tend to leverage analytics 5 times more than competitors.

Here are a few examples of how mobile analytics data enabled companies to directly improve business results:
  • Comcast used analytics to improve its recruiting mobile app’s user experience and saw a dramatic impact on its overall recruiting results, with mobile applicants increasing by over 20% overall and mobile candidates completing their applications at a rate of over 50%
  • used insights gained from Flurry’s mobile analytics to increase in-app purchases per user by 25%
  • Airbnb improved conversion rates 5x using Mixpanel’s mobile analytics
Creating a mobile analytics strategy would seem to be a obvious part of every mobile product development effort.  However, it is often an after-thought at best.  Why?  

The simplest answer is that delivery teams focus on what’s needed to get an app to market, not how to manage business results.  But—much like managing a stock portfolio—it’s imperative to plan from the start how to measure, track and improve a mobile product over time.

Now, mobile analytics can be a complex topic.  There are a variety of vendor solutions, many of which are web-oriented solutions that have recently bolted on mobile capabilities rather than being truly fine-tuned for mobile.  (On the other hand, the traditional web analytics solutions more readily offer cross-channel insights, especially if your choice of mobile analytics tool is not integrated.)  In any case, among the mobile-focused vendors, while there is a lot of functional overlap, products are often targeted for specific business models or types of analysis.  

So, how to start?  Here we’ll explore three steps to creating and implementing a mobile analytics strategy.

Initially, there are three main categories of information you should focus on supporting:
  • Business Measures of Success/KPIs.  Each product has a unique set of business outcome or success measures—these KPIs can include business results, technology performance, user behavior, and so on.  Whatever the business objective, make sure that the data required to calculate each KPI is accurately captured.  Also, know that KPIs may change over time as you gain insight and product matures.  Your objective is to ensure that all stakeholders can track the KPIs by which products’ contribution to the business is measured.
  • Technology and Operational Monitoring.  Track technical aspects of product performance; for example, response time for a key operation, which devices are being used, app crash details, network connectivity, etc.  Your objective is to proactively surface defects and performance issues—before they are posted in a negative review or result in app deletion—that prevent users from enjoying an optimal experience.
  • Engagement, Tracking User Behavior.  Prepare to listen for users’ signals about how and when they use the app, favored features, interactions and paths through the app.  Your objective is to connect the data points to surface insights on how to improve the aspects of user experience that are important to overall success.
In short, the data you track should support creation of information that either alerts you to issues to address or opportunities to explore.  This information will be historical (older than 24 hours), operational (today) or real-time.  

Plan to start simply:  Start small, test data accuracy and relevance, and iterate often—just like you’re doing with your mobile product as a whole.  Then assess the data and information you’re getting (or not getting) and optimize data being tracked—or even which tools are being used—to ensure you’re  generating actionable information.

Next, over time and as your analytics capabilities become more relevant and accurate, plan to generate insights from users’ cross-channel experiences, not just mobile.

Finally, ensure that everyone—including the development team—has access to the data so that they can suggest ways contribute to product improvement.

Select Tools
When should you select mobile analytics tools?  The answer is as soon as your product feature roadmap and supporting technology stack is established.  

As noted above, there are many mobile analytics tools available.  What’s more, mobile analytics tools are relatively new and immature.  So, what approach and which tool(s) to choose? 

Selection of the right analytics tools for your specific application can be a complicated task as vendors tend to be focused on various aspects of analytics; for example, app store analytics vs. crash data vs. defined user event analytics.

Here are some general principles to guide you:
  • Start with what’s free and basic—a combination of Google Analytics and Flurry.  These two popular tools, along with the app store tools (if your app is in the public app stores), will provide a strong foundation.  Also, there are open source options, such as Countly, available.
  • Understand your product’s category.  Don’t pick an analytics tool geared for mCommerce if your app is more focused on content delivery.  Also, assume that over time a blend of tools will be necessary to meet your specific needs.
  • Understand your product roadmap.  Pick analytics tools that will grow with you—and that you can still afford—as you add features, users, platforms and/or countries.
  • Look for features that enable you to clearly comprehend and quantify user behavior.  Two such features are   session playback, which lets you watch a simulation of actual users actions in your app, and heatmaps, which help you visualize which parts of your applications are highly utilized.
  • Ensure that app architecture is designed to enable you to easily use multiple and/or switch mobile analytics tools as needed.
  • Consider how you will track user activity across devices.  Tools like Google’s Universal Analytics track user paths across devices.
As you refine your selection of tools and approach, don’t forget to leverage enterprise data from operations, customer support, marketing or other relevant sources to build unique insights into your business and its customers.

Beginning to collect product analytics data when your app goes live is missing a very valuable opportunity:  beta and/or pilot data.  A pre-release period using an early adopter group or a savvy user set (for example, through PreApps) can yield a significant amount of actionable analytic data.  

You may discover, for example, that real-world users don’t use your app the way you envisioned and you need to make navigational or usability changes, drop or add features, optimize technology performance or re-prioritize your roadmap—valuable insights that will make your general availability release more successful.

Analytics data is often your first alert—before user reviews or even a phone call from customer support—that something is amiss.  Consequently, when your app goes live establish a regular schedule to review analytics data.  If your app is customer/consumer facing you may choose to review salient data daily with a full weekly review, including carefully sifting through data to discern trends and user signals.  Make sure to provide all stakeholders with regular KPI reporting, as well as actionable insights you’ve developed.

Finally, make sure that the information generated from your mobile analytics is appropriately factored into continued product roadmap planning.  Objective data should play a significant role in feature prioritization (what to keep, improve or drop), evaluation of underlying technology performance, user experience design effectiveness, and quality to name just a few areas of influence.

Closing Thoughts
We believe that all enterprise mobile apps—whether for the employees, partners or customers/consumers—can significantly benefit from the insights enabled by mobile analytics.  We'd recommend that everyone plan to deliver durable value by developing and implementing a clear mobile analytics strategy for each mobile product in their portfolio.

No comments:

Post a Comment