Why doesn’t Google sell a Google “Pro” version to all you addicts out there? A paid version that indexes more webpages, provides faster results, gives you more search filters and less web spam than the “Basic” version? With a billion users globally, wouldn’t it make sense to give the best features and functionalities to those most willing to pay for it? To understand that question, we’ll need to understand something about the nature of web businesses, users, scale, and monetization.
Pre-internet information technology businesses correlated to the DIKW hierarchy: data – information – knowledge – wisdom. In the 20th century, IT businesses provided the hardware to store, the tools to gather, the software to visualize and report, and the consulting services to understand, the inputs and outputs of human experience. The atom, the core irreducible mote, underlying the DIKW hierarchy is the stimulus outside of ourselves which we perceive with our senses, thus creating data capable of being recorded.
The web enables, and therefore requires, a very different structure. With the rise of the web we see businesses based around a different atom, a different mote: that of human behavior. Without our recognizing it, the global communications and technology network that has been our obsession for the past two decades has changed from being about things, to being about people.
Even more compellingly, it has changed from being constituted of data on the nature of things, to data generated by the nature of people.
The web hierarchy is:
Information – Intentions – Inferences – Interaction
If the pre-internet IT hierarchy is composed of bits that correlate to atoms, the web IT hierarchy is composed of behaviors that represent users:
- Users seek information on the internet, thereby
- Developing and revealing their intentions, thereby
- Enabling inferences that are valuable and non-intuitive, thereby
- Allowing for interactions or communities of which the user was previously unaware
This last step, the provision of discovery to users, has been an obvious, productive, and compelling application of the web since at least the founding of Firefly in 1995. (If you roll that way, here is Suck’s awesome 1995 review of the launch.)
We use Google to seek information, which in Google’s case is composed of requests that let us know who the Yankees play next Tuesday, how much butter to put in the cookies, where to eat in Luang Prabang, or the molecular weight of tryptophan. We use other sites to seek other types of information – on jobs, on stocks, the definitions of words. In either case, in each act of information-seeking we do not just learn about the information, but reveal something about ourselves.
John Battelle defined Google as a database of intentions: “The Database of Intentions is simply this: The aggregate results of every search ever entered, every result list ever tendered, and every path taken as a result.”
Beyond the search engines, users generate data about their actions at ecommerce sites, in dating applications, in the job search, in fact, every click, search, form, survey, poll, and mouse-over tells us something more about the user and their intentions. What they think they are doing tells us something about what they think, but also tells us a great deal more.
Intentions may or may not represent the conscious understanding of the user – for anybody who’s sat through usability labs and heard the subject say “I don’t know why I clicked, just seemed like that’s where I should go”, you know what I mean.
Based on their apparent intentions and expressed preferences, web businesses can make inferences about users.
As I am writing this, I’m listening to my Pandora Station named “Smog” after the band I seeded it with. Looking through the resultant list of songs to which I’ve given the “thumbs up” I see seven new bands that I’ve discovered in the past two weeks that I’d never heard of before and of whom I am now tremendously fond.
Surprising, non-trivial, productive inferences may not be unique to the web, but surprising, non-trivial, productive inferences at scale, are.
Allowing interaction and community
And those surprising inferences will create a user with the desire to act: to buy music, or join a discussion, or purchase an appliance, or go to a concert, or attend a Meetup. The things we can teach users about themselves allow them to participate in a broader set of activities, and understand the context of why they are doing so. It is not just generating behavior, or trial, it is generating self-enlightened behavior that is the most compelling aspect of this web hierarchy.
The IIII business hierarchy
The DIKW hierarchy had a corresponding business hierarchy: hardware for storing, tools for collecting, software for analyzing, and consulting for understanding. Each successive step in the hierarchy uses the bits provided by the prior step as the raw material from which it creates value.
Similarly, the IIII hierarchy has a corresponding web business hierarchy, as shown in bold below. And where the DIKW hierarchy passes bits from one level to the next, the IIII hierarchy passes user behavior from one level of the hierarchy to the next.
If access is your product, information is your revenue
If information is your product, intentions are your revenue.
If intentions are your product, inferences are your revenue.
If inferences are your product, interaction / community is your monetization.
If interaction / community is your product, time-space is your monetization.
If time-space is your product, trial is your monetization.
(In italics are the pre- and post-web-hierarchy business models, representing 20th century informations services business on one hand, and online-enabled offline businesses such as Meetup and Groupon on the other, respectively. I will discuss these in another post.)
Because each level of the hierarchy generates the “raw material” for the next level, web business have the interesting characteristic that you cannot monetize your core product – that is, the part of your product that generates user behavior, engagement, and attention for your site or app in the first place. For any particular product or service, one level in the hierarchy must be unfettered, cost-free, and generate as much user behavior as possible, in order for the company to monetize it effectively at the next level of the hierarchy. (In another post, I’ll discuss the scale – the perhaps surprisingly massive scale – of behavior required to generate monetization.)
And it is thus precisely because the user behavior on one level of the IIII hierarchy is the raw material for the monetization on the next level of the hierarchy that Google cannot have a “Google Pro” version. And it is precisely because the web is composed of human behavior that a new hierarchy is required to explain it.