Lean startup in the enterprise - the era of collaborative engagement (part 3/6)

This is the third entry in our six part series on the era of collaborative engagement that businesses are stepping into. The second entry focused on this being the era of B2C rather than B2B. In this piece we will begin to look at the implications of this shift, particularly as it pertains to how digital products in this new consumer-facing era are conceptualized and developed.

Lean startup:

In an environment where the primary concern is to optimize existing operations, we are generally dealing with a whole lot of knowns, and are building systems and processes to optimize those knowns. On the other hand, when you are developing new consumer facing products and services, the knowns - in particular consumer preferences and user behavior - are a lot less concrete, and what you need to build is not always clear up front. Hypothesis-driven experimentation, as prescribed by the lean startup approach, provides the best mechanism for tackling this problem. 

In his landmark book, The Lean Startup, Eric Ries provides the example of how Intuit uses hypothesis driven validation to build new products and features. Eric cites the example of SnapTax, software developed by Intuit, that initially started with the hypothesis that customers would prefer scanning their W2 to automatically feed the content into their tax returns instead of manually entering the information. However, when they began to test the hypothesis they quickly discovered that scanning could be at least as cumbersome as entering the information manually, and that what customers really wanted was the ability to take a picture of their W2 from their cell phones and have the information automatically populate their returns.

It's not about software:

What's critical to note though is that success of hypothesis driven testing has less to do with developing software than knowing what experiments to run and how to interpret the data collected from those experiments. Consider the example of Dropbox. To test the hypothesis that people were craving a simple to use storage solution that would enable users to synch files across their various devices, Dropbox created a video (see below) to test this hypothesis. Following the release of the video, Dropbox's beta sign ups shot up from a few thousand to almost 75K overnight, which immediately validated the core hypothesis surrounding the product. Yes, in Dropbox's case they did write code to actually develop the product in parallel to releasing this video, however technically this same hypothesis could have been tested without writing a single line or product code.

 

A different approach to managing IT-centric projects:

What the lean startup approach does is add a new stage to how IT-centric projects are managed and success measured. In more traditional IT-centric projects project generally evolve through three phases: 

  • requirements gathering
  • build
  • operate

When deploying consumer-facing products with more unknowns than knows, the requirements gathering stage changes to what can be thought of as the "search" stage. In this stage, like startups, enterprise project teams should focus on hypothesis driven testing to find the right product/product features to build or service/service options to provide. With this change, the success criteria and metrics used to quantify each stage start to look something like this:

Success Metrics.png

Our observation: 

Off late, we have spent a tremendous amount of time and energy researching consumer-facing mobile applications deployed by large companies, and the data we have uncovered is clearly pointing to very limited focus by companies on the search phase. For some perspective, consider these stats for Pepsico.

  • # of apps in Google Play across all brands: 25
  • total max global downloads on Google Play: 1.86M
  • total consumers engaged (assuming all unique consumers, and 50% of engaged users are on iOS vs. Android): 3.72M  

For a company with over 25 millions likes on Facebook, and many times that in terms of global consumers, less than 5M downloads across 51 apps (iOS & Android), for an average of less than 75K downloads per app, is a very low number. What's more interesting to note is that what we see with PepsiCo is not an anomaly, brand after brand from companies across verticals show the same trend. 

In conclusion: 

In our view getting the search phase right is absolutely critical when deploying new products and services. Adding the search phase to the traditional project management approach is the critical first step towards reversing this trend.