Own Your Category With a Network-Effect Product

At the intersection of product marketing and product development is constructing a “data flywheel” business model 

I recently found myself speaking with a number of companies - in completely different industries - yet all expressing the same essential strategy: We’d like to make our product more valuable - and more difficult to duplicate - by making it constantly learn from our customers. 

I stopped to think about that as an increasingly common strategy…. And how (in contrast) I’d been accustomed to marketing products that are just a race against competitors who were adding similar features to mine. That old approach was an endless technology treadmill, constantly trying to outpace feature parity for leadership.  

My insight was that a *real* leader is a product (or service) that develops an unfair edge, a hard-to-duplicate sustainable advantage, that makes (and keeps) the offering a runaway success. The "network effect" of customer usage is how it keeps its edge over the competition.

This is how I rediscovered the notion of the “Data Flywheel”; The idea that the more relevant data you collect from users, the more you can build better learning/algorithms and continuously develop a better and more valuable product. And finally, this positive feedback is what helps you acquire even more users and defend against competitors. 

The concept of the data flywheel was originally conceived by Jim Collins, author of “From Good to Great” IMO, a seminal business book. Another great discussion of this model is from Christopher Lochhead’s podcast (#24) part of an excellent series of marketing discussions. This product-centric network effect isn’t a new concept, but definitely one I don’t (yet) see most companies pursue. With the acceptance and maturation of more Artificial Intelligence (AI) engines, I expect this trend to accelerate over the next few years.

Your customer data is value - and key to valuation 

I’ll begin with this contrast: If your business is purely transactional, if you have no ongoing relationship to customers, if your product is not constantly learning and improving from interactions, then this should be a warning sign. It's because your customers only present a single value-in-time (the moment of purchase) and that’s all. Worse, you'll eventually lose to a competitor and/or be disrupted by someone evolving faster than you.

Many in the financial industry believe that ongoing customer interactions and data have very *real* value. Consider the simplistic example of valuing a company for its customer list and their purchase history, not just for its existing cash flow. (A very good paper by Glue Reply is on models of valuation of data as an asset) Or, consider valuing a company for its customer database, not just for hardware sales… a great recent example is Google’s acquisition of Fitbit. Per Motley Fool, “...there's a lot of data in there, of course. Fitbit collects all this health data. They've been building this digital health platform, much like Apple. Both Fitbit and Apple want to help users be able to manage their health data on this platform". At the acquisition price of $2.1 billion, this represented a valuation of approximately $75/customer.... a bargain if you think about the follow-on products/services Google may be able to offer in the near future. 

I recommend both product marketing and product development need to re-think what information the company can (or should) consider collecting, using, and ultimately monetizing in the form of a continuously improving product/service... and one that can generate recurring revenue while doing so.

Where Product Marketing and Product Management might begin:

In some cases it might be obvious what information you could collect to begin to build an accretive “flywheel”. If not, I suggest bringing engineering, product management, customer success leaders, and product marketing together - for a facilitated brainstorm. 

Ideally you want to answer this: “If we just knew ____, we’d provide incredible new value and insight for our customers.'' Consider dwelling on questions such as...

1. How could we create a better customer “network effect”?
2. What types of benchmark customer data sets could we develop, offer, and use?
3. If we had it, what external data would we tap into? (Or, could we partner with others to develop it?)
4. Do our customers want to compare themselves to others? What questions do *they* ask that we could help answer?
5. What customer information do we already have that other business (even unrelated) want? how could we monetize it?
6. What continuously updated product feature or data would keep customers returning to use our product/service?
7. Could we apply AI/ML to detect customer use patterns for higher-value offerings?
8. During your session, it’s important to think outside your current business model, and even outside your current product category. Remember - you’re trying to pursue a potentially disruptive approach to an otherwise traditional product. It might lead you in some exciting directions.

Finally, you'll probably want to build a customer “data lake” (as I see it often referred to) a unique, highly-valuable collection of customer interactions, intentions and experiences. And, if used wisely, it’s what will make your product/service more competitive and sustainable. It could include customer use of your product, their interactions on your web properties, their profiles they (opt-in to) provide to you, and even related public-domain data. But choosing what data to capture is key...

A few examples to get you thinking (real and invented)

• A software security vendor - builds a database of past attack vectors across all of its customers, including a list of previously compromised login credentials. In turn, the security software provides customers with increasingly faster, more complete responses to security threats and attacks. 
• A network monitoring vendor - builds a customer database of network performance benchmarks and typical precursors to failures across various technical environments and architectures. It then constantly provides better predictive alerts to potential failures rather than waiting to react to existing failures.
• A vape pen provider - through its interactive app, assembles a database of feedback to/from customers about satisfaction ratings of oils, oil vendors, and optimal devices and temperature settings for each. It also correlates age and sex of users to preferred oils, to make suggestions for future use and purchases.
• Automobiles - Unlike traditional insurance vendors, Tesla can assemble a database of driving habits of individual drivers to create custom risk profiles allowing the company to often offer less-expensive automobile insurance than a typical broad-based actuarial tables would otherwise permit.
• Entertainment/content providers - Netflix, Amazon, and others constantly monitor viewing and purchasing habits and demographics to recommend additional products as well as follow-on entertainment suggestions for constant up-sell opportunities.
• Excercycles - Peloton has famously created a subscription service for its exercise hardware that gathers and constantly updates a database of user data and demographics, matches individuals for competitions, tracks improvement, and creates custom workouts.
• Custom Sports Helmet Vendor - a custom 3D-printed sports helmet manufacturer provides a database of customer-submitted custom designs and improvements, as well as user ages and sports levels - proactively recommending new/upgraded helmets for growing and advancing athletes.

I’d love to hear how some of you have pursued a “data flywheel” approach to your product or service - and I’m sure other readers would too.

For More Inspiration…. 

Here are some additional links/articles you might find relevant:

• Strategy+ Business: How to build disruptive strategic flywheels
• Forbes: How To Hack Your Way Into A Proprietary Data Set
• JP Morgan: Go deep on data and grow your e-commerce business
• Reply.com: The valuation of data as an asset: a consumption-based approach (PDF)

Thanks for reading!