I’ve been quiet here on Searchblog these past few months, not because I’ve nothing to say, but because two major projects have consumed my time. The first, a media platform in development, is still operating mostly under the radar. I’ll have plenty to say about that, but at a later date. It’s the second where I could use your help now, a project we’re calling Mapping Data Flows. This is the research effort I’m spearheading with graduate students from Columbia’s School for International Public Affairs (SIPA) and Graduate School of Journalism. This is the project examining what I call our “Shadow Internet Constitution” driven by corporate Terms of Service.
Our project goal is simple: To visualize the Terms of Service and Data/Privacy Policies of the four largest companies in US consumer tech: Amazon, Apple, Facebook, and Google. We want this visualization to be interactive and compelling – when you approach it (it’ll be on the web), we hope it will help you really “see” what data, rights, and obligations both you and these companies have reserved. To do that, we’re busy turning unintelligible lines of text (hundreds of thousands of words, in aggregate) into code that can be queried, compared, and visualized. When I first imagined the project, I thought that wouldn’t be too difficult. I was wrong – but we’re making serious progress, and learning a lot along the way.
This is an edited version of a series of talks I first gave in New York over the past week, outlining my work at Columbia. Many thanks to Reinvent, Pete Leyden, Cap Gemini, Columbia University, Cossette/Vision7, and the New York Times for hosting and helping me.
Well, Walmart vs. Amazon is all about big business – a platform giant (Amazon) disrupting an OldBigCo (Walmart and its kin). Over the past two decades, Amazon bumped Walmart out of the race to a trillion-dollar market cap, and the OldCo from Bentonville had to reset and play the role of the upstart. The Token Act levels the playing field, forcing both to win where it really matters: In service to the customer.
But while BigCos are sexy and well known, it’s the small and medium-sized business ecosystem that determines whether or not we have an economy of mass flourishing. So let’s explore the Token Act from the point of view of a small business startup, in this case, a new neighborhood restaurant. I briefly touched upon this idea in my set up post, Don’t Break Up The Tech Oligarchs. Force Them To Share Instead. (If you haven’t already, you might want to read that post before this one, as I lay out the framework in which this scenario would play out.) What I envision below assumes the Token Act has passed, and we’re at least a year or two into its adoption by most major data players. Here we go…
Fresh off her $2,700 win from Walmart, Michelle decides she’s ready to lean into a lifelong dream: Starting a restaurant in her newly adopted neighborhood of Chelsea in New York City. Since moving to the area from California, she’s noticed two puzzling trends: First, a dearth of interesting mid- to high-end dinner spots walking distance from her new place, and second, what appears to be higher-than-average vacancy rates for the retail storefronts in the same general area. It appears to be a buyer’s market for retail restaurant space in Chelsea. So why aren’t new places launching? She read the Times’ piece on vacancies a few years ago (before the Token Act passed) and was left just as puzzled as before – seems like there’s no rhyme or reason to the market.
Michelle wants to start a high end American gastro pub – the kind of place she loved back when she lived in Northern California (she’s fond of Danny Meyers’ Gramercy Tavern, pictured above, but it’s a bit too far away from her new place). She has a strong hunch that such a place would be a hit in her new neighborhood, but she’s not sure her new neighbors will agree.
Now starting a restaurant requires a certain breed of insanity – they say the best way to make a small fortune in the business is to start with a large one. The truth is, launching restaurants has historically been a crap shoot – you might find the best talent, the best designer, and the best location – but if for some reason you don’t bring the je ne sai quois, the place will fail within months, leaving you and your partners millions of dollar poorer.
It’s that je ne sai quois that Michelle is determined to reveal. The tools she will leverage? The newly liberated resources of data tokens.
Before we continue, allow me to draw your attention back to the rise of search, indeed, the very era which begat Searchblog in the early 2000s. Google Adwords launched in 2000, and within a few years, the media world had been turned upside down by what I termed The Database of Intentions. As if by magic, people everywhere could suddenly ask new kinds of questions, finding themselves both surprised and delighted by the answers they received.
A Gates-Line compliant ecosystem quickly developed on top of this new platform, driven by an emerging industry of search engine marketing and optimization. SEO/SEM sprung into existence to help small and medium sized businesses take advantage of the Google platform – by 2006 the industry stood at nearly $10 billion in spend, growing more than 60 percent year on year. Adwords grew from zero to millions of advertisers by connecting to a long tail of small businesses that took advantage of an entirely new class of revealed information: The intents, desires, and needs of tens of millions of consumers, who relentlessly poured their queries into Google’s placid and unblinking search box.
Were you a limo service in the Bronx looking for new customers? It paid huge dividends to purchase Adwords like “car service bronx” and “best limo manhattan.” Were you a dry cleaner in West LA hoping to expand? Best be first in line when customers typed in “best cleaners Beverly Hills.” Selling heavy machinery to construction services in the midwest? If you don’t own keywords like “caterpillar dealer des moines” you’d lose, and quick, to whoever did optimize to phrases like that.
My point is simply this: Adwords was a freaking revolution, but it ain’t nothing compared to what will happen if we unleash data tokens on the world.
Ok, back to Michelle and her new restaurant. Of course Michelle will leverage Adwords, and Facebook, and any other advertising service to help her new business grow. But none of those services can help her figure out her je ne sai quois – for that, she needs something entirely novel. She needs a new question machine. And the ecosystem that develops around data tokens will offer it.
Thanks to her Walmart experience, Michelle has become aware of the power of personal data. She’s also read up on the Token Act, the new law requiring all data players at scale to allow individuals to create machine-readable data tokens that can be exchanged for value as directed by the consumer. After doing a bit of research, she stumbles across a startup called OfferExchange, which manages “Token Offers” on behalf of anyone who might want to query TokenLand. OfferExchange is a spinout from ProtocolLabs, a pioneer in secure blockchain software platforms like Filecoin. It’s still early in TokenLand, so an at-scale Google of the space hasn’t emerged. OfferExchange works more like a bespoke yet platform-based research outfit – the firm has a sophisticated website and impressive client list. It uses Facebook, Twitter, LiveRamp, and Instagram to identify potential token-creating consumers, then solicits those individuals with offers of cash or other value in exchange for said tokens.
Michelle does a Crunchbase search for OfferExchange and sees it’s backed by Union Square Ventures and Benchmark, which gives her some comfort – those firms don’t fund fly-by-night hucksters. And OfferExchange site is impressive – in less than five minutes, it guides her through the construction of an elegant query. Here’s how the process works:
First, the site asks Michelle what her goal is. “Starting a restaurant in New York City,” she responds. The site reconstructs around her answer, showing suggested data repositories she might mine. “Restaurants, New York City,” reads the top layer of a directory-like page. Underneath are several categories, each populated with familiar company names:
Restaurant Reservation and Review Services
OpenTable Google Resy Yelp Eat24 Facebook (more)
Food Delivery Services
GrubHub Uber Eats PostMates InstaCart (more)
Uber Lyft Juno Via (more)
Real Estate Services (Commercial)
LoopNet DocuSign CompStak (more)
Foursquare Uber Lyft Google NinthDecimal (more)
American Express Visa Mastercard Apple Pay Diners Club (more)
And so on – if she wished, Michelle could dig into dozens of categories related to her initial “restaurant New York City” search.
Michelle’s imagination sparks – the kinds of queries she could ask of these services is mind blowing. She could limit her query to people who live within walking distance of her neighborhood, asking her *actual neighbors* for tokens that tell her what restaurants they eat at, when they eat there, the size of their checks, related reviews, abandoned reservations, the works. She might discover that folks like Indian takeout on Mondays, that they rarely spend more than $100 on a meal on Tuesdays, but that they splurge on the weekends. She could discover the percentage of diners in Chelsea who travel more than two miles by car service to eat out at a place similar to the one she has in mind, and what the size of the check might be when they do. She can also check historical average rents for restaurants in her zip code, over time, which will certainly help with negotiating her lease. The possibilities are endless.
Put another way, with OfferExchange’s services, Michelle can litigate the merde out of her je ne sai quois.
This post is getting long, so I’ll stop here and pull back for a spot of Thinking Out Loud. I could continue the story, imagining the process of the token offer Michelle would put out through OfferExchange’s platform, but suffice to say, she’d be willing to pay upwards of $5-20 per potential customer for their data. The marketing benefit alone – alerting potential customers in the neighborhood that she’s exploring a new restaurant in the area – is worth tens of thousands already. And of course, OfferExchange can connect anyone who offers their tokens to Michelle’s new project a discount on their first meal at the restaurant, should it actually launch. Cool!
But let’s stop there and consider what happens when local entrepreneurs have access to the information currently silo’d across thousands of walled garden services like Uber, LoopNet, Resy, and of course Facebook and Google. While better data won’t insure that Michelle’s restaurant will succeed, it certainly increases the odds that it won’t fail. And it will give both Michelle and her investors – local banks, savvy friends and family members – much more conviction that her new enterprise is viable. Take this local restaurant example and apply it to all manner of small business – dry cleaners, hardware stores, bike shops – and this newly liberated class of information enables an explosion of efficiency, investment, and, well, flourishing in what has become, over the past four decades, a stagnant SMB environment.
Is this Money Ball for SMB? Perhaps. And yes, I can imagine any number of downsides to this new data economy. But I also believe the benefits would far outweigh the downsides. Under the Token Act as I envision it, co-creators of the data – the services like Uber, OpenTable, or Facebook – have the right to charge a vig for the data being monetized. Sure, it’d be possible for an entrepreneur to steal customers via tokens, but I’m going to guess the economic value of allowing your customers to discover new use cases for their data will dwarf the downside of possibly losing those customers to a new competitor. Plus, this new competitive force will drive everyone to play at a higher level, focusing not on moats built on data silos, but instead on what really matters: A highly satisfied customer. That’s certainly Michelle’s goal, and the goal of every successful local business. Why shouldn’t it also be the goal of the data giants?
Seven or so years ago, a famous VC penned a manifesto of sorts. Writing at a time the world was still skeptical of the dominance to which his industry has now ascended (to think, such a time existed, and so few years ago!), Marc Andreessen had a message for the doubters, the naysayers, and the Wall St. analysts who were (credibly!) claiming that his investments amounted to not much more than a bubble:
Seven years later, no one can dispute Andreessen’s prescience. The man was right: If you had purchased a basket of his favorite stocks back then – he name-checked Apple, Amazon, and Facebook directly – you’d be up at least 10X, if not more. Software, it seems, has indeed eaten the world, and those smart (and rich) enough to put money into technology, as Andreessen has been, have done very, very well for themselves.
Of course, not many people have in fact been that smart (or that rich). As of last year, ten percent of investors own 84 percent of the stock market, and that ratio only gets worse as time goes by. Most of our society simply isn’t benefiting from this trend of software eating the world. In fact, most of them live in the very world that software ate.
We – us, all of us – have turned the world to data. Some of us – the founders of software companies, the funders of those founders, the cheerleaders who run the capital markets – took that data and used it to change the world. Along the way, the world didn’t disappear like some unfortunate animal distending a python’s midsection. No, the world remains.
Who are we now that we’ve been eaten? What have we become?
These are questions, it turns out, that almost none of technology’s leadership have deeply pondered. It certainly never came up in Andreessen’s manifesto. And it’s manifestly evident in the behavior of our most treasured technology founders. They are puzzled by these newfound demands from United States senators and European socialists. Don’t they understand that regulation is damage to be routed around?
But the world is not just software. The world is physics, it’s crying babies and shit on the sidewalk, it’s opioids and ecstasy, it’s car crashes and Senate hearings, lovers and philosophers, lost opportunities and spinning planets around untold stars. The world is still real. Software hasn’t eaten it as much as bound it in a spell, temporarily I hope, while we figure out what comes next.
Software – data, code, algorithms, processing – software has dressed the world in new infrastructure. But this is a conversation, not a process of digestion. It is a conversation between the physical and the digital, a synthesis we must master if we are to avoid terrible fates, and continue to embrace fantastic ones.
I spent today at the first-ever Boing Boing Ingenuity, a two-day hackathon cum “vaudeville show” – truly Boing Boing brought to life. It made me so proud to see the essence of conversational marketing at work – a major brand adding deep value to a community, an independent publisher realizing its dream of celebrating its voice and community through a unique event that built up online over many months.
Here’s the Twitter stream. It was really great. And the main insight I took away was this: Brands will soon have no choice but to become data (because we are all becoming data, after all). A car creates tons of data every mile it is driven, for example. Faced with this fact, how might a car brand respond? It could see that data as its private asset, put up fences around it, and make that data really difficult for the driver (and society at large) to access. Or, it could act like Ford did, and tack in the direction of openness.
Ford has created a platform called OpenXC that opens APIs into 50 or so data streams coming out of your car. On the first day of Ingenuity, teams of makers, hackers, and regular folks came up with amazing ideas that leveraged Ford’s innovative platform. For example, one team built an app that senses when a pet or child is in your car, then monitors the car’s internal temperature. If it gets too hot, this app can lower the windows, turn on the AC, and text the car’s owner. I mean, how cool is that?!
Boing Boing’s editors presented the winners on stage on Sunday during Ingenuity, a day long celebration of, well, the weird and wonderful people and ideas that make Boing Boing, Boing Boing. There was a hack that turned driving data into music – if you drive aggressively and waste resources, the music gets aggravating. If you drive well, it gets soothing. Another hack interpreted braking, steering, and other information into new signals for other drivers – imagine a taillight flashing “Thank you!’ when someone lets you merge into oncoming traffic, for example. Yet another hack took all that data and turned it into a “cost per trip” dashboard that gamifies driving and encourages you to drive in a way that saves money.
These kinds of innovations can only occur in an ecosystem of openness. As our society tips toward one based on data, our collective decisions around how that data can be used will determine what kind of a culture we live in. And what I observed at Ingenuity strengthened my belief that companies that lean into an open approach to data will win. There will soon be streams of data coming from all manner of products – appliances, clothing, sporting goods, you name it. Wouldn’t you rather live in a world where you can export the data from your son’s football helmet to a new app that monitors force and impact against a cohort of high school players around the country? Or would a better world be one where Riddell Inc. owns and controls that data?
Way back in 2008 I wrote a piece about Facebook and data called It’s Time For Services on The Web to Compete On More Than Data. My point was this: winning on a strategy of data lockdown is a short-term play. What matters is the service you provide on top of that data. For companies like Ford, the key won’t be to lock in customer data and try to be the best at leveraging your proprietary insights. It’ll be allowing your customers to take that data out, remix it into a robust ecosystem, and feed it back to your company and products, so they can get better. Companies will compete on how they best leverage a customer’s data, not on whether or not they’ve locked those customers’ data assets in (are you listening, cable companies?!).
Of course, a true test of this optimistic scenario will come when GM, Toyota, or other car companies join Ford in offering a data platform, and a long-time Ford customer buys a Chevy. Will Ford let that customer take their data over to GM? Time will tell, but I know where I come down: Openness and portability will win in the end.