"Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Marc Grabowski, CEO at Iris Mobile.
Retailers need to build mobile phone applications. Users have shifted their digital time allocation to mobile devices, making an application the only way they can capture the data that will help identify and target these customers with the most personalized information appropriate for their buying patterns.
Failing to build a next-generation consumer app is a concession that an organization is behind the times and does not understand the shifts in user behavior or advances in technology functionality.
Unfortunately many application development cycles begin with an overzealous CMO, who declares, “We need to build an app!” Once retailers are committed to building an application, they have some serious considerations around purpose, insights and costs.
These are questions any data-driven marketer should answer before launching such a project. And while the application may be a vanity play for a CMO, the team tasked with its creation must ask the hard questions and own a point of view on these topics. Ultimately, they must drive the project to a successful outcome, whether it’s to deliver the app or share the data advising against such a deployment.
Purpose And Goals
The primary purposes of an application are to drive direct digital sales, act as a store locator and gather data from customers. An application’s most often forgotten aspect is data capture to identify customer interest on platforms outside of the application.
If the primary goal is to drive digital sales, it’s important to consult with others, including colleagues or competitors, to avoid the pitfalls of incorrect forecasting. Good user flows are needed to engage customers and move them through a purchase funnel to acquire products. This type of application should make the purchase flow so intuitive and fun that consumers want to visit the app to discover new products that can improve an element of their lives.
Simply to create a store locator is the worst motivation for building an application. Deploying a pedestrian store locator module to capture the “real estate” on somebody’s smartphone completely misses the point and will ultimately alienate consumers because their time has been wasted downloading an app for a task that could be done using the mobile web or Google Maps. If this is your starting point for an app, rethink the strategy when you are ready to do something that will really help your consumer.
By this point, you should intend to use the app to gather data and build a first-party mobile data asset. If data is the main priority, you will need to identify specific data gaps and paths to acquire missing data and make it actionable. You will need to build or partner with an analytics company to implement an SDK. Before going down this path, work with the IT team to understand their time frames to implement this or other SDKs, as you will need to add partners in the future.
A few data points that can be built with an application include device ID, which can be used for display targeting on exchanges, geographic location, which provides context for in-app notifications or beaconing, and a digital loyalty cardholder, which allows you to kill paper loyalty cards and the costs associated with this antiquated system.
No matter the overarching goal, it is worth considering whether you will be doing a single application across all brands and products or if you will be required to build (and promote) multiple disparate instances. The latter adds complexity as well as budget but is the right path if you have distinct brand identities.
Applications need to be measurable and improve the overall business model. What insights will the data provide around customer behavior and what will you do differently once you have evaluated these data? Raw metrics can be gathered on the fly but actionable insights must be pre-planned or risk monitoring the wrong behaviors.
Monitor the occurrences of leakage in your funnel and percent of users who drop in each stage. Be clear on the attributable windows measuring lifetime value to determine which cohorts of users are the most valuable to the strategy. If users visit once and never return, tweak the user experience. If this behavior is pervasive, it shows there was enough work done on the front end of user testing. Similarly, track the incremental value for each return visit, from a direct buying and data value perspective, and when diminishing returns are reached.
User group testing and beta launches will help avoid these failures and provide a smoother launch.
The last and potentially deciding consideration involves budget. As many marketers have learned, simply building an application is no guarantee that people will find it. Unless it goes viral early because your name is Kim Kardashian, significant budget must be dedicated to its adoption.
Think beyond the cost to build the application and also consider the budget needed to drive installations and re-engagement over the first year. Consider projected costs per app installation and forecast the number of installations and downloads required to break even, along with the number of return users included in the initial number of installations. To go the extra mile, do a sensitivity study on supply partners to assess the quantity of traffic that will be received from your most efficient sources and the point when they will become oversaturated.
There are also application upkeep and engineering costs. A transactional app can cost up to $2 million, while a marketing app could run up to $500,000, according to Forrester. As much as half of costs are related to front-end development, with the rest stemming from creative, design, back-end development and testing.
Count on a couple updates per year. More than 80% of respondents in an AnyPresence survey said they update apps at least twice per year, although nearly a third update apps once a month. After four years, most apps need a complete overhaul.
Once a full inventory of goals, predicted insights and costs has been undertaken, application development can begin. Look at competitors in the same category or vertical and determine who has successfully implemented a great app that should be emulated. Still confused? Consider poaching a rock star from a competitor and have them oversee the deployment of your app.
Follow Marc Grabowski (@MarcTGrabowski), Iris Mobile (@Iris_Mobile) and AdExchanger (@adexchanger) on Twitter.