The explosion of digital platforms is upending traditional business models and industries. Selling products and services is no longer enough; forward-looking companies are using experiences and outcomes to beat the competition. So what can businesses do to avoid being disrupted?
Ray Wang, principal analyst and chief executive officer at Constellation Research, offers a strategy for succeeding in the digital era in his new book, Disrupting Digital Business: Create an Authentic Experience in the Peer-to-Peer Economy. 1to1 Media spoke with Wang about how to succeed in a P2P economy and why savvy marketers are moving away from deterministic data strategies toward probabilistic ones.
1to1 Media: So many companies claim to be disruptive that "disruption" has become a buzzword. How are you tackling this concept?
Ray Wang: The book is about this massive digital disruption that's been happening since 2000 where digital is enabling the creation of new business models. There are also different levels of innovation. There's incremental innovation, where you're creating the next version of something and just making it smaller or bigger or adding new features. Truly disruptive innovation is creating a brand new business model. For instance, the iPhone disrupted dozens of business models when it gave us a camera, flashlight, alarm clock, navigation system, and more in one device.
Where do companies stumble when it comes to being innovative and truly disruptive?
RW: The problem is this takes time and you have to be in a continuous state of research to get there. But there are certain things you can do right away, such as changing how you price things and how you distribute things. That brings us to a peer-to-peer economy where companies are removing the middleman through micro-lending or new ways of buying things. Tesla is disrupting the dealer networks and Etsy allows you to create a craft and distribute it to anyone around the world. What's interesting is they're selling directly to you, the consumer.
How can companies adopt a peer-to-peer mindset and move away from thinking in terms of B2B and B2C?
RW: Terms like B2B and B2C go away when you think in terms of context like our relationship to a product or experience, the time of day, location, how you're feeling-that all comes together. Right now I could combine Verizon or AT&T cellphone data, point of sale information, and satellite imagery and tell you what the rent should be in a neighborhood.
Every organization will be brokering their data. Data is a competitive weapon. And it's not about omnichannel, multichannel, or those other buzzwords. It's about delivering a continuity of experience, processes, screens, channels, journeys, etc. That's what customers expect and when you deliver on that. Then the whole notion of separate marketing, sales, and customer service departments goes away.
In your book you mention that probabilistic data strategies will overtake deterministic strategies. Why is that?
RW: It goes back to context. With probabilistic models I'm using as much data and context as I can so that I know with a certain degree of confidence that X is going to happen. For example I can gather enough data to see patterns about where you shop and how you use your device to know this is you.
The deterministic way would be that I only measure data through certain access points and I only track people from trusted cookies. Deterministic models use set rules whereas probabilistic models know that with a degree of confidence this is you based on your previous actions. We're also moving away from deterministic and probabilistic thinking because what we ultimately want to get to are intention-driven [tactics.]
What's an example of an intention-driven marketing tactic?
RW: Let's say I know this is you because of X and Y and I want to see if you'll do something differently depending on what I send you. For example, if you always get a Frappuccino at Starbucks, how can I get you to order a tall tea? Maybe it turns out that when you travel, you order differently. So I'm going to wait until you're at a Starbucks that you're not normally at, and make you an offer for the tall tea. With intention-driven systems, we know that what you normally get is a Frappuccino, but let's try to sneak in another offer to see if you bite and then we figure out why you took that offer.
What about the accuracy rate of probabilistic models compared to deterministic models?
RW: Accuracy is always going to be an issue [with probabilistic models]. If you're shooting for 99.9 percent accuracy, don't bother. But if you want 90 percent accuracy and your customers are okay with close enough, it's not an issue. The message could be as simple as we know you always get this, but would you be interested in trying something new?