Joe McKendrick at Forbes has posted a concise and thoughtful column about
Everyday Chaos, including four rules to guide your expectations about machine learning.

It’s great to see a pre-publication post so on track about what the book says and how it applies to business.

1 Comment »

At the PAPis conference Bas Nieland, CEO and Co-Founder of Datatrics, is talking about how to predict the color of shoes your customer is going to buy. The company tries to “make data science marketeer-proof for marketing teams of all sizes.” IT ties to create 360-degree customer profiles by bringing together info from all the data silos.

They use some machine learning to create these profiles. The profile includes the buying phase, the best time to present choices to a user, and the type of persuasion that will get them to take the desired action. [Yes, this makes me uncomfortable.]

It is structured around a core API that talks to mongoDB and MySQL. They provide “workbenches” that work with the customer’s data systems. They use BigML to operate on this data.

The outcome are models that can be used to make recommendations. They use visualizations so that marketeers can understand it. But the marketeers couldn’t figure out how to use even simplified visualizations. So they created visual decision trees. But still the marketeers couldn’t figure it out. So they turn the data into simple declarative phrases: which audience they should contact, in which channel, what content, and when. E.g.:

“To increase sales, çontact your customers in the buying phase with high engagement through FB with content about jeans on sale on Thursday, around 10 o’clock.”

They predict the increase in sales for each action, and quantify in dollars the size of the opportunity. They also classify responses by customer type and phase.

For a hotel chain, they connected 16,000 variables and 21M data points, that got reduced to 75 variables by BigML which created a predictive model that ended up getting the chain more customer conversions. E.g., if the model says someone is in the orientation phase, the Web site shows photos of recommend hotels. If in the decision phase, the user sees persuasive messages, e.g., “18 people have looked at this room today.” The messages themselves are chosen based on the customer’s profile.

Coming up: Chatbot integration. It’s a “real conversation” [with a bot with a photo of an atttractive white woman who is supposedly doing the chatting]

Take-aways: Start simple. Make ML very easy to understand. Make it actionable.

Q&A

Me: Is there a way built in for a customer to let your model know that it’s gotten her wrong. E.g., stop sending me pregnancy ads because I lost the baby.

Bas: No.

Me: Is that on the roadmap?

Bas: Yes. But not on a schedule. [I’m not proud of myself for this hostile question. I have been turning into an asshole over the past few years.]

Comments Off on [liveblog] Bas Nieland, Datatrix, on predicting customer behavior

Vinny Senguttuvan is Senior Data Scientist at METIS. Before that, he was at Facebook-based gaming company, High 5 Games, which had 10M users. His talk at PAPIs: “Predicting Customers.”

The main challenge: Most of the players play for free. Only 2% ever spend money on the site, buying extra money to play. (It’s not gambling because you never cash out). 2% of those 2% contribute the majority of the revenue.

All proposed changes go through A/B testing. E.g., should we change the “Buy credits” button from blue to red. This is classic hypothesis testing. So you put up both options and see which gets the best results. It’s important to remember that there’s a cost to the change, so the A-B preference needs to be substantial enough. But often the differences are marginal. So you can increase the sample size. This complicates the process. “A long list of changes means not enough time per change.” And you want to be sure that the change affects the paying customers positively, which means taking even longer.

When they don’t have enough samples, they can bring down the confidence level required to make the change. Or they could bias one side of the hypothesis. And you can assume the variables are independent and run simultaneous A-B tests on various variables. High 5 does all three. It’s not perfect but it works.

Second, there is a poularity metric by which they rank or classify their 100 games. They constantly add games — it went from 15 to 100 in two years. This continuously changes the ranking of the games. Plus, some are launched locked. This complicates things. Vinny’s boss came up with a model of an n-dimensional casino, but it was too complex. Instead, they take 2 simple approaches: 1. An average-weighted spin. 2. Bayesian. Both predicted well but had flaws, so they used a type of average of both.

Third: Survival analysis. They wanted to know how many users are still active a given time after they created their account, and when is a user at risk of discontinuing use. First, they grouped users into cohorts (people who joined within a couple of weeks of each other) and plotted survival rates over time. They also observed return rates of users after each additional day of absence. They also implement a Cox survival model. They found that newer users were more likely to decline in their use of the product; early users are more committed. This pattern is widespread. That means they have to continuously acquire new players. They also alert users when they reach the elbow of disuse.

Fourth: Predictive lifetime value. Lifetime value = total revenue from a user over the entire time the the produced. This is significant because of costs: 10-15% of the rev goes into ads to acquire customers. Their 365 day prediction model should be a time series, but they needed results faster, so they flipped it into a regression problem, predicting the 365 day revenue based on the user’s first month data: how they spent, purchase count, days of play, player level achievement, and the date joined. [He talks about regression problems, but I can’t keep up.] At that point it cost $2 to acquire a customer from FB ad, and $6 from mobile apps. But when they tested, the mobile acquisitions were more profitable than those that came from through FB. It turned out that FB was counting as new users any player who hadn’t played in 30 days, and was re-charging them for it. [I hope I got that right.]

Fifth: Recommendation systems. Pandora notes the feature of songs and uses this to recommend similarities. YouTube makes recommendations made based on relations among users. Non-matrix factorization [I’m pretty sure he just made this up] gives you the ability to predict the score for a video that you know nothing about in terms of content. But what if the ratings are not clearly defined? At High 5, there are no explicit ratings. They calculated a rating based on how often a player plays it, how long the session, etc. And what do you do about missing values: use averages. But there are too many zeroes in the system, so they use sparse matrix solvers. Plus, there is a semi-order to the games, so they used some human input. [Useful for library Stackscores
?]

Comments Off on [liveblog] Vinny Senguttuvan on Predicting Customers

I’m in Talent Garden‘s largest branch, which is also its headquarters, in Milan. It’s a ridiculously large co-working space for startups, with an emphasis on openness. I’m enjoying sitting at a table with a few other people, none of whom I know and all of whom are speaking Italian.

I like co-working spaces enough that if I were looking for a place to work outside of my house, I’d consider joining one. It’s that or the local library. It depends on whether you find being around the young and the digital to be distracting, energizing, or both.

I find it energizing. Nevertheless, the segregation of the young from the old is a cultural and business loss.

Talent Garden ameliorates this by renting space to a handful of established companies (IBM, Cisco, and a bank, here in Milan) to provide mentoring, and so the old companies can get behind startups they find interesting. It’s a good model, although since I’m just here for the afternoon, I don’t know how much actual mingling occurs. Still, it’s a good idea.

I also like that Talent Garden explicitly tries to build community among its users. Not waving-in-the-hall community, but a community of shared space, shared events, and shared ideas. The American co-working space I’m most familiar with has public areas but assumes startups want to work in rooms with closed, solid doors. An open floor plan helps a startup culture to grow, which is perhaps more needed in Italy than in the US. Nevertheless, you can’t have too much community. Well, you can, but that’s easier to remediate than its opposite. (For a US shared space dedicated to building community, check out the treasured Civic Hall in NYC.)

(Note: Unlike the co-working space I’m most familiar with, TG does not provide a free, well-stocked kitchen. Just as well. Free kitchens cause my metabolism to think its faster than it is.)

I’m in Italy to participate in an Aspen Institute event in Venice over the weekend (poor poor me). I stopped in Milan to give a talk, which I internally have titled “Is the Internet Disappointed in Us?” It’s actually a monolog — no slides, no notes — about why my cohort thought the success of the Internet was inevitable, and why I am still optimistic about the Net. If you’re interested in having me in to speak with your group, let me know. Yeah, a plug.

And while I’m plugging, here’s some disclosure: Talent Garden is the venue for this talk, but no one is paying me for it.

5 Comments »

Robin Chase is giving a lunchtime talk at the Berkman Center.

There is a totally new organization paradigm that exists next to the Internet, she says. She calls it “Peers, Inc.” It changes how we shape the economy. It’s happening now. Her explanation will be in three parts:

Platforms

First, platforms for participation that leverage excess capacity. E.g., Facebook, Skype, Meetup, YouTube, MOOCs, open source, Blockchain, etc. For example, Skype is a telecoms company built on the excess capacity of its users systems. Working with excess capacity means sharing.

Bed-sharing (couchsurfing, AirBnB) uses excess beds. “It took four years for AirBnB to have more available beds than the largest hotel chain”It took four years for AirBnB to have more available beds than the largest hotel chain (InterContinental): 650,000. Couchsurfing has more than a couple of million.

We invented big institutions to do things that we can’t do as individuals. E.g., large investments, projects that require intelligence in lots of different areas, standardized contracts. And there are things that individuals do better: customization, specialization, creativity, trust.

These two coexist, and the Net enables them to collaborate. She calls this Peers, Inc. (“Institutions and governments are also Inc’s in this world view.”) The Inc’s provide a platform for participation, and the individual provides creativity and specialization.

Robin “adores excess capacity” because it’s green and efficient. Excess capacity is something that’s already been paid for but contains unused value. How do you harness it? 1. You can slice it so only pay for what they use (e.g., ZipCar); this lets you avoid buying more car than you need. 2. You can aggregate (e.g., AirBnB, Waze). 3. Open up these assets, e.g., data.gov and GPS.

The Inc side builds platforms for participation. They organize lots of small parts. “They “Platforms give the power of the large to the small”give the power of the large to the small.” They can scale. She points to a French car-sharing company: BlaBlaCar. Four million people use it every month.

Peers bring diversity. E.g., smartphones and apps. Smartphones are far harder to build than the apps they enable. Over 2M apps have been developed since smartphones were invented in the past seven years. “We’ve seen more innovation than throughout all of human history” because people can build apps that are relevant to their own situations. App creators are free-riders on top of the $600 people spend on their smartphones.

2. Peers Inc give us new powers, which she thinks of as miracles.

“The most depressing thing I know is climate change.” By 2100, we’ll see a 4-6°C increase unless we take dramatic action. What does that feel like? “The last time we were minus 7°F was the last ice age.” Warming the planet that amount transformed the planet. We should expect the same level of change if we boost it another 7°F. By 2060, it will be really awful. So we have to address this.

“Banny Bannerjee says: “You can’t solve exponential problems with linear solutions.””Banny Bannerjee says: “You can’t solve exponential problems with linear solutions.”

The “miracles” give her some optimism:

a. “We can defy the laws of physics” by leveraging excess capacity. If she had proposed building 640,000 rooms in four years she would have been told that that’s not possible. But AirBnb did it by leveraging existing excess capacity.

b. “We can tap exponential learning.” Platforms can get millions of iterations in and can do a lot of learning. E.g., learning a language. A semester is 130 hours. Rosetta Stone teaches the same in 54 hours. But it’s expensive. “My new favorite company is DuoLingo.” They do a lot of A/B testing. They now can teach you a semester in 34 hours. They have 90M people using it. A year and a half ago DuoLingo opened up its processes: Russians learning Balinese, etc. Now 45M of the 90M are learning language pairs DuoLingo did not create. (DuoLingo makes money because they have humans translating sentences from organizations that pay them incrementally.)

c. “The right person will appear.” E.g., Obama raised the prospect of normalizing relationships with Cuba. Six months later, AirBnB had 2,000 listings there, thanks to the Net.

Her only hope for climate change is creating platforms that will address climate “at scale, speed, and locally adapated.” E.g., a platform for a house will remember to turn off the light when there’s been no movement. We’ll get smart cities through the Internet of Things. Distributed energy. Autonomous vehicles, which will arrive in force in the next 5-12 years. We’ll only need 10% of the cars because we’ll be sharing them. “Public transportation will be at the cost of a bus but the speed of cars”Public transportation will be at the cost of a bus but the speed of cars, transforming job opportunities. (But the Internet of Things means that everything is tracked.)

All of these miracles only happen because of both sides of Peers Inc.

3. “Everything that can become a platform will become one.” Old-style industrial capitalism put thick boundaries around companies. Today, what’s inside and outside is blurred.

Four reasons Robin is convinced we’re moving into the collaborative economy:

1. Shared networked assets always provide more value than closed assets

2. More networked minds are smarter than fewer proprietary minds.

3. “The benefits of shared open assets are always larger than the problems associated with open assets.” E.g., yes, some people put scratches in ZipCars, but the company nevertheless is doing very well.

4. What I get is great than what I give.

We are in a time of instability. “Peers Inc is the only structure that can experiment, iterate, evolve and adapt at the pace required.”Peers Inc is the only structure that can experiment, iterate, evolve and adapt at the pace required.

So, how can we structure things so we give up the least privacy necessary? “What is the least privacy loss that delivers a habitable climate”

QA

Q: For me it’s not privacy loss but who we’re losing our privacy to. What about platform accountability? Aren’t we pushing out power into more abstract systems that we cannot see or address?

A: I was on a panel at the Platform Cooperativism conference. I pointed out that these platforms are incredibly expensive. ““He who finances the platforms creates the rules of engagement.”He who finances the platforms creates the rules of engagement.” “I want these platforms created by a distributed, autonomous us.” We don’t have time to just hope this happens. “I have real anxiety.”

Q: [me] Suppose we build protocol instead platforms…

A: I’ve put all of that into the same bucket.

Q: Shareable cars disrupt ZipCar. There will be user agreements. How do we disrupt that?

A: “He who creates the data owns the data.” Autonomous vehicles have a middle space, e.g., around safety and learning issues. It’s in the deep public interest to have this data. But we need to make the privacy issues understandable and parseable by ordinary users so they can choose.

Q: Isn’t privacy gone already?

A: We can still do some structuring.

Q: Why does trust work over the Web, which is mostly anonymous?

A: Ebay was the first to figure out you need ratings and commentaries. We use other people as our proxies for trust.

Q: iRobot’s Roombas currently don’t upload what they’ve learned about the layout of your house. But Nest knows everything. What should the rules be?

A: That’s what I’m asking you. We have to figure this out.

Q: It’d be great if we had more choice about which pieces of info we give to platforms. Is there any work on standard ways of parceling out pieces of our identity?

A: I know people are working on this. “It comes back to the amount of money, time, and marketing it takes to push great ideas into market.”

Q: What are we doing to educate the younger generation about privacy?

A: Maybe you can push Harvard to do appropriate role modeling. Maybe students here could push for an icon system that tells us what data you’re taking from us, etc.

Q: [me] What would you tell a student about the dangers? And would you consider addressing this by putting restrictions on how the data is used, rather than on its collection?

A: How about doing some pilots to see what works? You have to inform people about the dangers as well regulating the industry.

Q: How will we embed public safety concerns into software for self-driving cars?

A: Self-driving cars will always follow the rules. No speeding. No parking in no-parking zones. All the existing rules will be embedded. So we’ll embed the appropriate behavior for ambulances, etc. No siren required. Also: The auto industry always brings up autonomous cars having to decide which person to kill in an accident. But why would you bring up this stupid case? One in a million trips this might happen? There are more deaths than that now. “Right now, 80% of cars are single occupancy. We need to put a high price on that.”Right now, 80% of cars are single occupancy. We need to put a high price on that.

Q: It sounds like you’re describing a train: get somewhere, not park…Why not public transportation?

A: We’ll see how it plays out. It’ll be a complex ecosystem. It’ll be decided city by city. More important than who owns it are: Will they be electric? Will it be 10x more expensive for single occupancy? Will we have pharmacy cars or liquor cars that deliver their wares without having a storefront? Who will design the software?

Q: Practically, how do you combat zoning for selfishness, e.g., my own one-person gas guzzler?

A: I don’t spend a lot of time on local issues. When I have, logic and data haven’t had much effect.

2 Comments »

Perhaps you’d like to post to your WordPress blog from an app that isn’t WordPress.

I know I do. I write most of my posts in an editor (javascript + php) I’ve worked on for over ten years. Someday I’ll clean it up and post it at GitHub so you can all have a laugh. Meanwhile, it intermittently loses its ability to post straight to my blog, so I have to copy and paste the text into the WP editor. But I fixed it again today. So, here’s a tutorial for people at my level of non-technicality. (I got huge help from a post at HurricaneSoftware. Thanks!)

First, make sure that the file xmlrpc.php is installed where you’ve installed your WordPress blog software. This file comes from WordPress itself, and it should be there automatically. Check the permissions; I think it should be 644 but I am terrible at permissions.

I run my homegrown editor from my Mac, using the Apache web server that MAMP supplies. That lets me write blog posts even when I’m not online. That means the directory from which I’m running my JavaScript and PHP is on my hard drive. I keep these files in  /Applications/MAMP/htdocs/blogdraft/. (Blogdraft is the name of the folder in which my code resides.) To the web server, the address looks like this: /localhost/blogdraft/.

The operative part of this is your PHP file. Create an empty text file and name it, let’s say, postViaXmlrcp.php. For a first pass, it should look like this—and the brunt of this comes straight from HurricaneSoftware:

This PHP script relies upon another one, so you have to load it. “Require_once” will do so, and it will remember that it has done so during a session so you won’t waste computer resources reloading it every time you run this script.

You can get this script here. Right click on that link and choose “Save file as…” or however your browser puts it. Put it in the same directory as your PHP script. Make sure you name it “IXR_Librabry.php.inc.” Set its permissions. (See above.) Then leave it alone. 

These lines read data that you’ve sent from the JavaScript that we haven’t written yet. It assigns them to some pretty obviously-named PHP variables.

Notice that we’re doing nothing with the $customfields variable. That’s because I don’t know what to do with it. I would have just deleted that line, but it scares me. And yet fascinates me.

htmlentities is a PHP function that makes sure that your HTML with all of its weird characters arrive without being translated into something more literal and wrong. Line 12 tells it which character encoding to use. I could have decided on one for you, but instead I’m just using whichever one you already use. We have already established I’m a coward, right?

We are eventually going to be sending all of the content information to  WordPress via XMLRPC. This section packs an array (“$content”) with the information XMLRPC needs, attached to the keywords it loves. If you want to argue about it, take it up with XMLRPC.

We now create a new client for the IXR script you downloaded. It wants to know where your xmlrpc.php file is, which should be where the rest of your WordPress files are folders are.

Fill in your WordPress username and password.

Now we’re making another array. This one includes the prior array ($content) as well as your username and password. And note the comment. Setting to “draft” is very useful when you’re playing around with these scripts.

This does the deed. No one knows how.  If it fails, it’ll pop up the error messages and kill it before it spawns evil. Otherwise, it sends back the message that it posted successfully.

You can test this PHP script by running it in your browser. If you’re running a local web server, you’d enter something like this: /localhost/blogdraft/postViaXmlrcp.php. (That’s assuming you put it in a folder called “blogdraft,” of course.) Check with MAMP or whatever you’re using for your web server for details.

But running this as-is won’t work because it’s expecting the content to be sent to it from the JavaScript we still haven’t written. So, comment out lines #7-11, and insert something like these:

Replace the categories (“cat1”, “cat2”) with the names of categories that you actually use. Also, change “true” to “false” in line #29 so you’ll just produce drafts, not actually publish anything yet.

Now when you run this PHP file in your browser ( /localhost/yourLocalFolder/postViaXmlrcp.php), if should create a draft post. Check via the “All posts” page at your WP administration page to see if the draft got created.

When it’s working, comment out the four lines immediately above and uncomment lines #7-11.

The JavaScript

I’m going to pretend that you have some HTML page that has a text box where you can enter the content of your post, a similar box for entering the title, one for entering tags separated by commas, and checkboxes that list the categories you use. I’ll also assume that you use jQuery. So, your HTML might look soomething like this:

So, roughly, here’s what’s happening:

This loads jQuery from Google. Of course you could keep a local copy and include it that way.

The postIt function begins by using jQuery to fetch the values entered into the three text areas. (Just in case you don’t know jQuery, “$(“#titlebox”) gets the element with the ID of “titlebox.” And if you don’t want to use jQuery, you can get the same result with: var titlecontent = document.getElementById(‘titlebox’).value.

Creating an array of categories takes a little more work. Line #17 creates an array (“checks”) of all of the checkboxes in the div with the id “categories.” Lines #21-27 look at each of the checkboxes in that array. If line #23 sees that a particular checkbox has in fact been checked, then it puts the value of that checkbox into the created on line #19. (You want the value to be exactly the same as the name of the category in your WordPress installation. Also, remember that the checkbox’s value is not necessarily the same as the text displayed to the user.)

Now we call the PHP script that we created above. We do this via AJAX, using the jQuery syntax, which is much simpler than the native JavaScript way of doing it. Lines #30-40  specify the communication the JavaScript will have with the PHP file.

Line #30: The “POST” here has nothing to do with posting a blog post. It’s stating what sort of transaction we’re about to have with the PHP script.

Line #31: This is the path to the PHP file we’re going to run. If it’s in the same directory as this HTML file, you don’t have to monkey with a path name.

Line #32: We’re going to pass data to the PHP script in the JSON way of expressing data.

Line #33: This creates the JSON we’re going to send. It’s all within curly brackets. There are four phrases, separated by commas. Each phrase consists of a keyword (which you can think of as being like a variable) and a value. We are free to make up whatever keywords we want, so long as those are the keywords we use in the PHP file to fetch the data that they label; see lines #7-10 in the PHP script above.

Line #44: If there is an error in the PHP, it will send back some information. There is also an equivalent “success:” function available. But I’m doing something wrong, because even when the PHP works and the blog gets posted, I still get an error message. If you go back to Line #33 of the PHP, you’ll see that if the PHP succeeds, it sends the message “Article Posted Successfully.” For reasons I don’t understand, that message shows up in the “error:” function of the AJAX. So, I check the message. If it has the word “Successfully” in it, the script alerts the user that the post has been posted. If it does not, on line #39 it posts an error message.

That’s it. If it doesn’t work, it’s because you’re doing something wrong,  starting with listening to me. Obviously I can’t help you since I don’t even know how this thing works.

Good luck!

3 Comments »

Reddit is in flames. I can only see one way out of it that preserves the site’s unique value.

I say this as an old man who loves Reddit despite being way outside its main demographic. Of course there are outrageously objectionable subreddits—topical discussion boards—but you don’t have to visit those. Reddit at its best is wonderful. Inspiring, even. It is a self-regulated set of communities that is capable of great collective insight, humor, and kindness. (At its worst, it is one of the nightmares of the Internet.)

Because Reddit is so large, with 169M unique visitors each month, it is impossible to generalize accurately about what went on yesterday and is continuing today. Nevertheless, the precipitating cause was the termination of the employment of Victoria Taylor for reasons Reddit and she have not disclosed. Victoria was not only the wildly popular enabler of Reddit’s wildly popular AMA‘s (“Ask Me Anything”), she was the only Reddit employee visible to most redditors (Reddit users).

Victoria’s sudden dismissal was taken by many as a sign of the increasing misalignment of Reddit’s business goals and the culture of its communities. Reddit, it is feared, is going commercial. The volunteer moderators (“mods”) of some of the large subreddits have also complained that their requests for support over the past months have gone unanswered.

In protest, many of the large subreddits and a long list of smaller ones have gone private and are thus dark to most of the world. This will have some financial effect on Reddit, but it is better understood as a political protest, applying the technique used successfully in 2012 when Reddit, Wikipedia, and other major sites went dark to protest the SOPA/PIPA bills that would have limited Internet freedom. It is an assertion-by-deprivation of the cultural value of these subreddits.

It is, I believe, a mistake to view this uproar primarily in terms of economics or business. This is an attempt by a community to stay a community despite perceived attempts by the business underneath it to commercialize it. Up until now, Reddit the Company has understood the importance of accepting and promoting its community’s values. Advertising is unobtrusive, some of which lets users comment on the ad itself. Reddit makes money also from its users buying “Reddit gold” to bestow upon comments they find particularly valuable. Reddit gold has no monetary value, so users are consciously paying Reddit money for the privilege of paying another user a visible compliment. And Reddit has sternly defended the free speech of its users even when that speech is, well, horrible—although the management did controversially shut down some shaming and hating sites a few weeks ago.

Reddit is in bad shape today. The meme-making forces of sarcasm it’s famous for have been turned inwards.The most loyal users are feeling betrayed. Some of the communities that have driven Reddit forward as a cultural force are feeling abused. It’s hard to come back from that.

A big part of the problem is that Victoria, the face of Reddit to its own community, was accepted as “One of us! One of us!” as redditors sometime self-mockingly invoke the movie Freaks. Indeed, she embodied many of the virtues of Reddit at its best: curious, accepting, welcoming, helpful, funny. Many redditors saw themselves reflected in her.

Victoria was thereby an important part of Reddit’s support of what I call “The Gettysburg Principles“: She helped Reddit seem to be by, for, and of us. Now the face of Reddit is Ellen Pao, the interim CEO who is largely derided and detested at Reddit because she seems to be “One of them! One of them!”— a Silicon Valley player.

If we view this first and foremost as a problem in maintaining a community rather than strictly as a revenue issue, then I can only see one way forward: Pao should get off her executive horse, engage with the community in public, and show that she’s a redditor too. Alexis Ohanian, the co-founder, also should step forward with his best redditor face on. Alexis when free of corporate pressures is a redditor through and through.

There is still an opportunity for Reddit to show that it understands the source of all its value: communities trusted to run themselves, and a strong sense of shared cultures.

3 Comments »

There are a lot of things wrong with how Starbucks implemented its “Race Together” program for which it deserves the mockery it’s been getting. Whether it was intended to stimulate discussions with busy baristas (“So, you want that with nonfat milk and we shouldn’t fill it to the brim. Right? What’s it like being white? Did you say ‘Nicky’ or ‘Mickey’?”) or among customers who in my experience have never struck up a conversation with another customer that was not met by a cold stare or a faked incoming text, it was unlikely to achieve its intended result. (Schultz seems to indicate it was to be a barista-to-customer conversation; see 0:20 in the John Oliver clip linked to “mockery” above.) Likewise, the overwhelming male whiteness of the Starbuck’s leadership team was an embarrassment waiting to happen. The apparent use of only white hands holding cups in the marketing campaign was inconceivably stupid (and yet still better than this).

Yet there’s much that Starbucks deserves praise for more than just its recognition that racial issues permeate our American culture and yet are more often papered over than discussed frankly.

• They trusted their on-the-line employees to speak for themselves, and inevitably for the corporation as well, rather than relying on a handful of tightly constrained and highly compensated mouthpieces.

• They have held a series of open forums for their employees at corporate events, encouraging honest conversation.

• They did not supply talking points for their employees to mouth. That’s pretty awesome. On the other hand, they seem also to have provided no preparation for their baristas, as if anyone can figure out how to open up a productive conversation about race in America. The made-up phrase “racetogether” really isn’t enough to get a conversation going and off to a good start. (Michelle Norris’ Race Card Project might have provided a better way of opening conversations.)

Starbucks got lots wrong. Too bad. But not only was it trying to do something right, it did so in some admirable ways. Starbucks deserves the sarcasm but not just sarcasm.

[Disclosure: No, Starbucks isn’t paying me to say any of this. Plus I hate their coffee. (The fact that I feel the need to put in this disclaimer is evidence of the systemic damage wrought by “native ads” and unscrupulous marketers.)]

1 Comment »

The project with Doc that I mentioned is a new set of clues, following on The Cluetrain Manifesto from 16 years ago.

The clues are designed as an open source publishing project: The text is in the public domain, and we’re making the clues available at Github in various computer-friendly formats, including JSON, OPML and XML.

We launched this morning and a happy hell has broken loose. So I’m just going to posts some links for now. In fact, I’m copying and pasting from an email by Doc:

• @DaveWiner’s listicle version of New Clues (at listicle.io, one of his many creations).
• Dave’s blog post on New Clues, and working with Doc and me.
• New Clues on Medium’s Backchannel (courtesy of @StevenLevy).
• @Cluetrain on Twitter. Also @dweinberger and @dsearls
• The New Clues group page on Facebook
• Halley Suitt’s interview with Doc and me at the Globe’s BetaBoston
• Doc’s blog.

Gotta run…

5 Comments »

The Harvard Business School Digital Initiative [twitter:digHBS] — led by none other than Berkman‘s Dr. Colin Maclay — has launched its blog. The Digital Initiative is about helping HBS explore the many ways the Net is affecting (or not affecting) business. From my point of view, it’s also an opportunity to represent, and advocate for, Net values within HBS.[1] (Disclosure: I am officially affiliated with the Initiative as an unremunerated advisor. Colin is a dear friend.[2])

The new blog is off to a good start:

• Digital News, Devices, and Design Thinking in India by Hasit Shah
• Getting Your Car Stolen in Mexico and Other Fun Stories About Disruption” by John Winsor

I also have a post there titled “Generative Business and the Power of What We Ignore.” Here’s how it starts:

“I ignore them. That’s my conscious decision.”

So replied CV Harquail to a question from HBS professor Karim Lakhani about the effect of bad actors in the model of “generative business” she was explaining in a recent talk sponsored by the Digital Initiative.

Karim’s question addressed an issue that more than one of us around the table were curious about. Given CV’s talk, the question was perhaps inevitable.

CV’s response was not inevitable. It was, in fact, surprising. And it seems to me to have been not only entirely appropriate, but also brave… [more]

[1] I understand that the Net doesn’t really have values. It’s shorthand.
[2] I’m very happy to say that more than half of the advisors are women.

2 Comments »