|
April 29, 2019
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.
October 11, 2016
I’m at the PAPIs conference. The opening panel is about building intelligent apps with machine learning. The panelists are all representing companies. It’s Q&A with the audience; I will not be able to keep up well.
|
NOTE: Live-blogging. Getting things wrong. Missing points. Omitting key information. Introducing artificial choppiness. Over-emphasizing small matters. Paraphrasing badly. Not running a spellpchecker. Mangling other people’s ideas and words. You are warned, people. |
The moderator asks one of the panelists (Snejina Zacharia from Insurify) how AI can change a heavily regulated audience such as insurance. She replies that the insurance industry gets low marks for customer satisfaction, which is an opportunity. Also, they can leverage the existing platforms and build modern APIs on stop of them. Also, they can explore how to use AI in existing functions, e.g., chatbots, systems that let users just confirm their identification rather than enter all the data. They also let users pick from an AI-filtered list of carriers that are right for them. Also, personalization: predicting risk and adjusting the questionnaire based on the user’s responses.
Another panelist is working on mapping for a company that is not Google and that is owned by three car companies. So, when an Audi goes over a bump, and then a Mercedes goes over it, it will record the same data. On personalization: it’s ripe for change. People are talking about 100B devices being connected by 2020. People think that RFID tags didn’t live up to their early hype, but 10 billion RFID tags are going to be sold this year. These can provide highly personalized, higher relevant data. This will be the base for the next wave of apps. We need a standards body effort, and governments addressing privacy and security. Some standards bodies are working on it, e.g., Global Standards 1, which manages the barcodes standard.
Another panelist: Why is marketing such a good opportunity for AI and ML? Marketers used to have a specific skill set. It’s an art: writing, presenting, etc. Now they’re being challenged by tech and have to understand data. In fact, now they have to think like scientists: hypothesize, experiment, redo the hypothesis… And now marketers are responsible for revenue. Being a scientist responsible for predictable revenue is driving interest in AI and ML. This panelist’s company uses data about companies and people to segmentize following up on leads, etc. [Wrong place for a product pitch, IMO, which is a tad ironic, isn’t it?]
Another panelist: The question is: how can we use predictive intelligence to make our applications better? Layer input intelligence on top of input-programming-output. For this we need a platform that provides services and is easy to attach to existing processes.
Q: Should we develop cutting edge tech or use what Google, IBM, etc. offer?
A: It depends on whether you’re an early adopter or straggler. Regulated industries have to wait for more mature tech. But if your bread and butter is based on providing the latest and greatest, then you should use the latest tech.
A: It also depends on whether you’re doing a vertically integrated solution or something broader.
Q: What makes an app “smart”? Is it: Dynamic, with rapidly changing data?
A: Marketers use personas, e.g., a handful of types. They used to be written in stone, just about. Smart apps update the personas after ever campaign, every time you get new info about what’s going on in the market, etc.
Q: In B-to-C marketing, many companies have built the AI piece for advertising. Are you seeing any standardization or platforms on top of the advertising channels to manage the ads going out on them?
A: Yes, some companies focus on omni-channel marketing.
A: Companies are becoming service companies, not product companies. They no longer hand off to retailers.
A: It’s generally harder to automate non-digital channels. It’s harder to put a revenue number on, say, TV ads.
Comments Off on [liveblog] First panel: Building intelligent applications with machine learning
June 18, 2014
Jill Lepore has an excellent take-down in The New Yorker of Clay Christensen’s The Innovator’s Dilemma. Yet I am unconvinced.
I thought I was convinced when I read it. It’s a brilliantly done piece, examining Christensen’s evidence, questioning his methods, and drawing appropriate lessons, including wondering why we accepted the Innovator’s Dilemma for decades without critically examining it. (Christensen became so famous for it that his last name isn’t even flagged as a spelling error on my Mac.)
I got de-convinced by a discussion on a mailing list I’m on that points to some weaknesses in Lepore’s own argument, including her use of “cherry-picked” examples — a criticism she levels at Christensen — and her assumption that the continuity of companies, as opposed to their return on assets, is the right measure. As a person on the mailing list points out, John Hagel, John Seely Brown and Lang Davison take return on assets as a key metric in their book The Big Shift. And then someone else maintained that ROA is a poor measure of networked phenomena. That morphed into a discussion about the pragmatic value of truth: Does disruption provide a helpful framing for the New York Times as it considers its future?
The problem is that brains are truthy. They are designed to pay attention to things that seem to matter to us, bending our world around our concerns and interests. And brains are associative, so they make sense of the world — maybe even at the level of perception — by finding the relationships that seem to matter to us. In Heidegger’s terms, we are not indifferent knowing machines, but are creatures that care about what happens to us and to others. The brain is an unreliable narrator.
We now have access to an unfathomable sea of information that can contradict anything we settle on. That sea has been assembled by caring creatures and their minions, but it is so vast and global that it contains information beyond the caring and linking of any one of us. Every understanding can be subverted with a wink and a hand wave because all understanding simplifies a world that is resolutely and even necessarily complex. The universe outruns us.
Now we have machines that can look at masses of data and escape from our temptation to turn everything into a narrative. But those machines are limited by our decision about which data is worth gathering and connecting. There is hope in this direction, but it’s not clear whether we are capable of accepting the findings of machines that correlate without stories.
TL;DR: Our brains are truthy and the world is too big to make sense of. Not that that will stop us from trying.
[June 20:] Clay Christensen has cried foul in an interview.
December 17, 2013
Peter Cappelli has written an excellent post at HBR that falls simultaneously into the “Well, Duh” and “Needs to Be Said” bins: “It’s Not OK That Your Employees Can’t Afford to Eat.” Well, duh! It’s amazing that it even needs to be said. (Note that the Duh belongs not to Peter but to whomever needed to hear that.)
Let me put it differently. From my point of view, here are the two fundamental objectives for any business:
1. Enable your customers to lead lives that are a little bit better.
2. Enable your employees to lead good lives.
Profit is what you use to do both of those things.
October 6, 2013
“People don’t want to buy a quarter-inch drill. They want a quarter-inch hole.”
This gets quoted a lot by marketers. Usually it gets attributed to Theodore Levitt, an economist at Harvard Business School, but he quite explicitly [pdf] attributed it to Leo McGinneva, about whom I can find out nothing other than that he was a “businessman.”
This quote has the salutary effect of focusing marketers away from what they’re selling and on what customers are buying. So, I find it useful. But also irksome.
I’m irked first of all for the small reason that people don’t actually buy quarter-inch drills to drill quarter-inch holes. The buy a quarter-inch drill bit to drill a quarter-inch hole. A quarter-inch drill is a drill that accepts drill bits with a maximum of a quarter-inch shank. And, yes I know I’m being annoying.
The more important reason this formulation bothers me becomes clear if you use something other than a tool as your example. “People don’t want to buy a towel hook. They want a _____.” How do you fill in that blank without it being simply redundant: “They want a hook to hang a towel on.” It’s not just that it loses its rhetorical punch. Rather, it becomes clear that you have to go further into the customer’s value system to make sense of it. Why do they want a towel hook? Because they like dry towels? Because they want to impress their new in-laws? Because they repainted and the old towel hook is now the wrong color? Because they want a place to hang a dress so that the shower will naturally steam it? Because their shower rod is coming loose? Because their pet ferret is getting old — poor Ratface! He can barely see! — and is soiling towels left on the floor?
So, people don’t buy holes. They buy something that helps achieve a goal that is particular to them and is part of the larger set of interests and values that make them who they are. The hole example helps but doesn’t go far enough.
We all know this. So why does the “drill/holes” example keep coming up, and keep feeling like an insight? To me, this is evidence of just how much we take for granted the misalignment of the interests of businesses and customers — the great business tragedy of the Age of Massness.
But that’s a different story.
May 30, 2013
Greg Silverman [twitter:concentricabm], the CEO of Concentric, has a good post at CMS Wire about the democratization of market analysis. He makes what seems to me to be a true and important point: market researchers now have the tools to enable them to slice, dice, deconstruct, and otherly-construct data without having to rely upon centralized (and expensive) analytics firms. This, says Greg, changes not only the economics of research, but also the nature of the results:
The marketers’ relationships with their analytics providers are currently strained as a service-based, methodologically undisclosed and one-off delivery of insights. These providers and methods are pitted against a new generation of managers and executives who are “data natives” —professionals who rose to the top by having full control of their answering techniques, who like to be empowered and in charge of their own destinies, and who understand the world as a continuous, adaptive place that may have constantly changing answers. This new generation of leaders likes to identify tradeoffs and understand the “grayness” of insight rather than the clarity being marketed by the service providers.
He goes on to make an important point about the perils of optimization, which is what attracted the attention of Eric Bonabeau [twitter:bonabeau], whose tweet pointed me at the post.
The article’s first point, though, is interesting from the point of view of the networking of knowledge, because it’s not an example of the networking of knowledge. This new generation of market researchers are not relying on experts from the Central Authority, they are not looking for simple answers, and they’re comfortable with ambiguity, all of which are characteristics of networked knowledge. But, at least according to Greg’s post, they are not engaging with one another across company boundaries, sharing data, models, and insights. I’m going to guess that Greg would agree that there’s more of that going on than before. But not enough.
If the competitive interests of businesses are going to keep their researchers from sharing ideas and information in vigorous conversations with their peers and others, then businesses simply won’t be as smart as they could be. Openness optimizes knowledge system-wide, but by definition it doesn’t concentrate knowledge in the hands of a few. And this may form an inherent limit on how smart businesses can become.
Comments Off on [2b2k] Can business intelligence get intelligent enough?
March 2, 2013
CNN.com has posted my op-ed about why where you work is not about the quality of your life so much as about the substance of it.
Judging from some of the reaction, I should emphasize that if the only way to save Yahoo were to require everyone to come to work every day, that would certainly be the right decision. But it seems clear to me that Marissa Mayer was sending a signal with this policy, for surely there are some people who were working productively from home. So, if the new policy is a signal and is not actually required to save Yahoo, then I think she has underestimated how disruptive a signal it is. [To late to stick in a spoiler notice? That was the essence of my op-ed.]
Also, CNN.com has stripped out the links, I’m pretty sure unintentionally. Here they are:
On dispersed coding teams [pdf]
A box full of 404s
October 1, 2012
My latest column in KMWorld is about why your business needs scholars. In fact, though, it’s about why the idea of scholarship is more helpful than focusing your thinking on knowledge.
Comments Off on [2b2k] Your business needs scholars
August 2, 2012
I wanted to replace the smashed screen of a white MacBook, and found what seemed like a very good price from Wegener. The new screen arrived very quickly, and was exactly as described. But when I started to strip down the MacBook, I discovered I had ordered the wrong screen. It’s surprisingly easy to do.
So, I sent an email to Wegener and quickly got a reply, followed by a phone call. The support person said they are happy to send me the right screen, for which I have to pay a little more because it’s a more expensive part. They’re sending it even before I return the old one. So far, the experience has been terrific: Quick responses, friendly people, good return policy.
Then they told me that in the carton for the replacement part I’ll find a postage paid mailing label. I reminded them that the problem was entirely my fault, and thus there’s no reason for them to pay for shipping. Yikes, that’s some good customer service! (I went ahead and returned the first screen on my own dime.)
It’s amazing how powerful an experience it is to be treated like a human being by a business.
Yo, Humbles, I already bought the product. You don’t have to entice me any more. On the other hand: You’ve made me love you even more, and you’ve helped some musicians spread their music just a little wider.
There’s been one major variable that I know of: I switched from a Braun electric toothbrush to a Philips SoniCare.Why? Because the Internet told me to. I believe that the correlation is not accidental (see what I did there?), but of course it is just one data point.
June 2, 2012
Is it just me, or are we in a period when new distribution models are burgeoning? For example:
1. Kickstarter, of course, but not just for startups trying to kickstart their business. For example, Amanda Palmer joined the Louis CK club a couple of days ago by raising more than a million bucks there for her new album. (She got my $5 :) As AFP has explained, she is able to get this type of support from her fans because she treats her fans honestly, frankly, with respect, and most of all, with trust.
2. At VODO, you can get your indie movie distributed via bittorrent. If it starts taking off, VODO may feature it. VODO also works with sponsors to support you. From my point of view as a user, I torrented “E11,” a movie about rock climbing, for free, or I could have paid $5 to stream it for 10 days with the ability to share the deal with two other people. VODO may be thinking that bittorrenting is scary enough to many people that they’ll prefer to get it the easy way by paying $5. VODO tells you where your money is going (70% goes to the artist), and treats us with respect and trust.
3. I love Humble Bundle as a way of distributing indie games. Periodically the site offers a bundled set of five games for as much as you want to pay. When you check out, you’re given sliders so you can divvy up the amount as you want among the game developers, including sending some or all to two designated charities. If you pay more than the average (currently $7.82), you get a sixth game. Each Bundle is available for two weeks. They’ve sold 331,000 bundles in the past three days, which Mr. Calculator says comes to $2,588,420. All the games are all un-copy-protected and run on PCs and Macs. Buying a Humble Bundle is a great experience. You’re treated with respect. You are trusted. You have an opportunity to do some good by buying these games. And that’s very cool, since usually sites trying to sell you stuff act as if buying that stuff is the most important thing in the world.
4. I’m hardly the first to notice that Steam has what may be the best distribution system around for mass market entertainment. They’re getting users to pay for $60 games that they otherwise might have pirated by making it so easy to buy them, and by seeming to be on the customer’s side. You buy your PC game at their site, download it from them, and start it up from there. They frequently run crazy sales on popular games for a couple of days, and the game makers report that there is enough price elasticity that they make out well. If I were Valve (the owners of Steam), I’d be branching out into the delivery of mainstream movies.
There’s of course much much more going on. But that’s my point: We seem to be figuring out how to manage digital distribution in new and successful ways. The common threads seem to be: Treat your customers with respect. Trust them. Make it easy for them to do what they want to do with the content. Have a sense of perspective about what you’re doing. Let the artists and the fans communicate. Be on your customers’ side.
Put them all together and what do you have? Treat us like people who care about the works we’re buying, the artists who made them, about one another, and about the world beyond the sale.
Joho the Blog uses WordPress blogging software.
Thank you, WordPress!
